PAEDIATRIC FORENSIC

MEDICINE AND
PATHOLOGY

: dited by

Anthony Busuttil
Em eritus Regius Professor of Forensic Medicine, University of Edinburgh; and Medical Director,
Fo rensic Medical Services, NHS Lothian, Edinburgh, UK

Jean W Keeling
_::ormerly Consultant Paediatric Pathologist, Royal Hospital for Sick Children,
::diilburgh, UK

i, ARNOLD
HODDER
?ART OF HACHETTE LIVRE UK
 I CONTENTS
I

Contributors xiii
Preface xiv

Acknowledgements xv
ist of abbreviations used xvi

1 Clinical assessment in suspected child abuse 1

Helen Hammond
Introduction 1

Alerting signs 2

Types of abuse 3

The need for comprehensive assessment 4

Significant harm 4

The interagency context (flow chart of process) 5

Joint working and the complementalY skills of paediatricians 6

and forensic specialists

Legislation 7

Consent and confidentiality 8

The process - joint paediatric/forensic examination 8

Documentation and report writing 16

Interpretation of the findings 17

Formulating an opinion 21

Ongoing health care 21

Involvement in ongoing legal and child -care processes 22

References 22

2 Investigation of suspected sexual abuse 24

Jacqueline YQ Mok
Introduction 24

The colposcope in the medical examination 27

Forensic evidence 27

Skills and experience required 28

Consistent vocabulalY 29

Normal female genital anatomy 29

Perianal findings 36

Acute, healing and healed anogenital trauma 37

Female genital findings in sexual abuse 38

Signs of anal abuse 39

Conditions that mimic abuse 40

Screening for sexually transmitted infections 41

Interpretation of clinical and laboratOlY findings 42

SummalY 43

References 44

 vi I Contents

3 Radiology of child abuse 47

Maeve McPhillips

Role of the radiologist 47

Radiological investigations
48
Skeletal injuries
51

Head injury 60

Visceral injuries 68
Soft-tissue injury 69
Differential diagnosis 69
Conclusion 73
References 73

4 Haematological abnormalities that can simulate abuse 76

Angela Thomas
Introduction
76
Primary haemostasis 78

Secondary haemostasis
79
Laboratory tests
81

Measurements of primary haemostasis 82

Evaluation of a bleeding patient 82

Patterns of abnormal results

86

Normal coagulation screen with a normal platelet count 90

Abnormalities of platelet number or morphology 94

Coagulation defects 96

The neonate
97

Drugs associated with bleeding 98

Bone marrow failure syndromes 99

Systemic disease associated with a bleeding tendency 100

Activation of coagulation 101

Conclusion 101

References 102
5 Biochemical investigations on post-mortem specimens 106

Denis R Benjamin
Introduction 106
General evaluation 107
Hypoxia 109
Inflammation
109
Anaphylaxis 109

Infection
110

Dehydration and electrolytes 110

Time of death (post-mortem interval) 111

Endocrine disorders
112

Genetic metabolic disorders presenting as sudden unexpected death 114

Technical considerations at the time of autopsy 117

References 120

6 Ocular involvement in non-accidental injury 125

Harry Willshaw
Introduction
125
Scope of ocular and adnexal injury 125

,
' .-' -----. .
 Contents I vii

Fundus haemorrhages 128

References 134

7 The death scene following the sudden death of a child 137

Anthony Busuttil
Introduction 137

Scene management 137

The crime scene manager 138

Sequence of events at the death scene 139

Unclothing the body 139

A good look around 139

Sudden infant death syndrome or non-sudden infant death syndrome 140

External petechiae 140

Bruising 140

Abandoned neonates 140

Deaths from trauma 141

Dyadic and multiple deaths 141

Sudden deaths of older children 142

Sensitivity and stress of the investigation 142

Inquests and inquiries 143

References 143

8 Post-mortem examination in babies and children 145

Jean W Keeling

Introduction 145

Death scene investigation 145

Rectal temperature 146

Medical and family history 146

Other important information 146

Radiological examination 146

Photography 147

Microbiological samples 148

Toxicological investigations 149

Biochemical and metabolic investigations 149

Weights and measurements 150

External examination 150

Estimating blood loss 152

Dissection (infants and older children) 152

Examination of the brain, spinal cord and eye 156

Examination of the newly born 158

Histological samples 162

Retention of organs 163

Exchange of information and multidisciplinary review 163

References 164

9 Pathology of neurological abnormality in early life 166

Waney Squier
Introduction 166

Clinical manifestations of early brain damage: cerebral palsy 167

Timing of injuries by histology 167

 viii I Contents

Acquired intra-uterine damage 169

Birth-related injury 173

Stroke in the developing brain ]76

Metabolic disorders 177

Infections 177

References 178

10 Fetal and perinatal death 180

Jean W Keeling

Introduction 180

Definitions 180

The law 181

Background information 182

Concealed pregnancy 182

Unattended delivery 182

Was the baby born alive? 183

Is the baby of sufficient maturity to survive? 187

Is there evidence of prolonged or difficult labour? 187

Are there any significant injuries? ]88

Fetal death foJlowing maternal injury 190

Is there a natural cause for death? 193

Can I give a cause of death? 193

Should the intrapartum still birth be a medicolegal autopsy? 194

References 195

11 Sudden unexpected death in infancy: sudden infant death 198

syndrome or something else?

Jean W Keeling

Introduction 198

The definition of SIDS 199

Epidemiology 201

Sleeping environment 203

Pathological findings in sum 205

Death certification 2]8

References 219

12 Sudden natural death in infants and children 225

Dick Variend
Introduction 225

Cardiovascular causes of sudden death 226

X-linked hypohidrotic (anhidrotic) ectodermal dysplasia 235

Intracranial haemorrhage, neoplasms and malformations 236

Gastrointestinal causes 239

Fatal anaphylaxis 239

Sickle cell disease 240

Haemorrhage as a cause of sudden death 240

Respiratory causes of sudden death 240

Epilepsy and sudden death 242

Deaths from acute asthma 243

' -

.
------­ ....
.
 Contents I ix

Diabetes mellitus
243
Genetic metabolic disorders
244
Other bacterial infections
247
Deaths related to obstetric events and premature birth
248
Miscellaneous causes of sudden natural death
248
Sudden unexplained death in older children
249
Sudden natural death in the early neonatal period
249
Sudden death associated with 'intermediate' pathology
249
References
250

13 Recent advances in paediatric toxicology 256

Patrice Mangin and Christian Giroud
Scope of the problem 256
Specificity of paediatric toxicology 258
Techniques used in drug testing 259
Special techniques for analysis of volatile substances 262
Alternative specimens for drug testing 262
Pitfalls and limitations of drug screens 267
Specific applications 271
The importance of paediatric toxicology in specific cases 274
Conclusions and future considerations in forensic paediatric toxicology 274
References 275

14 Head and neck injuries 282

Robert A Minns and TY Milly Lo
Definition
282
Epidemiology
283
Non-accidental head injury
294
Traumatic birth injury
300
Primary mechanisms of injury to the brain
302
Secondary mechanisms of brain injury
307
Injury to the cervical spinal cord
311
Genetic influence on recovery from traumatic brain injury
312
References
313

15 Heat-induced injury or death 318

Anthony Busuttil
Introduction
318
House fire deaths
318
The pathologist's role
319
References
327

16 Asphyxial deaths in children 329

Anthony Busuttil
Petechiae
329
Scene of death
330
Traumatic asphyxia in children
330
Entrapment asphyxia
330
Foreign body inhalation
330
 xI Contents

Plastic bag asphyxia 331

Overlaying and wedging 331

Strangulation 331

Hanging by a ligature 332

Drowning and near drowning 332

Imposed airways obstruction 332

Abuse of inhalants (solvent abuse) 333

Reverse suspension 333

Chemical asphyxia 333

Prevention 333

References 334

17 Accidental injuries in children 336

Anthony Busuttil
Overview of paediatric trauma 336

Bicycle helmets 338

Falls 339

Playground injuries 340

Sports injuries on snow and ice 340

Riding injuries 341

Agricultural injuries 341

Prevention 341

Older children and substance abuse 341

Accidental poisoning 342

Hypersensitivity 342

References 342

18 Drowning and near drowning 345

John Pearn
Introduction 345

The causes of childhood drowning: a perspective 345

The drowning medium 346

The pathophysiology of drowning 349

Forensic immersion syndromes 351

References 359

19 Sudden death of children in hospital 362

Jem Berry
Introduction 362

Definition and frequency 362

Deaths due to natural disease 363

Deaths due to failure to monitor 365

Therapeutic misadventures 366

Deaths due to dmg treatment 366

Deaths due to medical devices and procedures 368

Deaths in the dental chair 371

Sudden death in newborn babies 371

Accidents 372

Suicide 373

Filicide and homicide in hospital 373

,~------~-- .
,
.
 Contents I xi

Investigation of sudden unexpected death of children in hospital 375

References 377

20 Road traffic accidents in children 385

Anthony Busuttil
Road traffic fatalities 385

Investigation of a fatal road traffic collision 387

Vehicular collisions 390

Other supervening problems in collisions 391

Pedestrian injuries 392

Child cyclists 392

Diffuse axonal injury 393

Whiplash injuries 393

Injuries to children in utero 393

Other vehicular accidents 393

References 394

21 Forensic DNA profiling in cases involving children 395

Alex M Graham and David J Harrison

Introduction 395

Inheritance of genetic material 395

Forensic DNA analysis: history and techniques 398

Sample collection and processing 402

DNA evidence and child sexual offence 403

Y chromosome short tandem repeat typing 405

Mixed samples 406

Additional sample problems and solutions 407

Mitochondrial DNA 407

Paternity testing 408

Identification of body remains and missing persons 411

Identification of the 'abandoned baby' or fetal material and avenues 413

for identifying the source of an unknown profile

DNA databases 414

References 414

22 The dentist's role in child abuse and neglect 420

David Whittaker
Introduction 420

Dental neglect 422

Facial and oral pathology 422

The dentist accused of child abuse 425

Bite marks 425

References 432

23 Paediatric dental identification 435

G Howard Moody

Introduction 435

Comparison 435

Facial reconstruction and dental profiling 440

 xii I Contents

Age estimation 441

References 444

24 The expert witness and expert testimony 447

Anthony Busuttil
Introduction 447

Mission statement of the expert 448

Claim to expertise 449

Professional witnesses 449

Opinions 450

Yes or no? 450

Admissibility of expert evidence 450

Communications from the expert witness 452

Declaration by the expert in the report 452

In the witness stand or box 453

Pre-trial communication 453

Conflict of interest 454

Rules of evidence 454

Conclusion 454

Recent developments 454

References 455

Appendix A: Child protection examination forms 457

Appendix B: Tables of standard measurements 471

Index 485

~ - - - - - - - - - - - - - - - - -- -~

.
-------

.
 I CONTRIBUTORS I

Denis R Benja min Maeve McPhillips

Department of Laboratories Department of Rad iology

Cook Children's Medica l Cen ter Royal Hospital for Sick Children

Fo rt Worth, TX, USA Edinburgh, UK

Jem Berry Robert A Minns

Formerly Professor in Paediatric Pathology Professor of Paed iatric Neurol ogy
Directorate of Pathol ogy Dep artment of Child Life and Hea lth
St Michael's Hospital The University of Edinburgh
Bristol, UK Edinburgh, UK
Anthony Busutt il Jacqueline YQ Mok
Formerly Emeritus Regius Professo r of Forensic Medicin e Consultant Paed iatrician
Un iversity of Edinburgh; and Department of Community Child Health
Med ical Director Roya l Hospi tal for Sick Children
forens ic Medical Services Edinburgh, UK
0i HS Lo thian
Edin burgh, UK G Howard Moody
Consultant in Oral Pathology
Christian Giraud
Edinburgh Denta l Institute
Institut Universitaire de Medicin e Legale
Edinburgh, UK
La usann e, Switzerland

.-\lex M Graham John Pearn

Division of Pathology (Fore nsic Medicine) Professor of Paediatrics and Child Health
University of Edinb urgh The University of Queensland
Ed inbu rgh, UK Royal Children'S Hospita l
Herston, Queensland. Australia
Helen Hammond
Consultant Paedia tri can (Community) Waney Squier
Department of Community Child Hea lth Consul tant Neuropathologist
St John's Hospital Radcliffe Infirmary
Livingston, UK Oxford, UK

David J Harrison Angela Thom as

Professor of Pathology Consultant Paediatric HaematoJogist
Division of Pathol ogy (Forensic Medicine) Roya l Hospi ta l for Sick Children
The University of Edinburgh Edinburgh, UK
Edin burgh, UK
Dick Variend
Jean W Keeling
Consultant Paediatric Pathologist (retd)
fo rmerly Consultant Pa ediatric Pathologist
The Children'S Hospital
~oy al Hospital for Sick Children
Sheffield, UK
:::din burgh, UK
:y Milly Lo David Whittaker
Cl inical Research Fellow Emeritus Professor in Forensic Dentistry
)epartment of Ch ild Life and Health University of Wales
- e University of Edinburgh Cardiff, UK
~ inburgh, UK
Harry Willshaw
~at riceMa ngin Consultant Paediatri c Ophthalmologist
:nsritut Universitaire de Medici ne Legale Th e Birmingham Child ren's Hospital
_ausanne, Switzerl and Birmingham, UK
 I PREFACE
The possibility that a child may have been injured, abused,
neglected or otherwise ill treated rightly raises the indigna­
tion and anxiety of the caring professions and involves law
enforcement agencies. However, in the interests ofjustice a nd
fairness, a person accused of such injury or negl ect is entitled
to appropriate legal representatio n at any hearing and is
deemed to be innocent until proven guilty. Quite frequently,
dubiety and uncertainty linger about whether or not, in any
specific situation, observations made, clinically or pathologi­
cally, can be interpreted solely as a manifestation of inflicted
injury or neglect, or whether there is a possibility that the
observed findings could have come about in other, non­
criminal, circumstances. These matters necessitate advice
from those with expelience and expertise in this field.
The aim of this book is to furnish an authoritative, com­
prehensive tex t to assist practitioners of medicine and the
law dealing with such cases in the approp ri ate interpreta­
tion of these matters and to enable clinical and pathological
findings to be presented in an unbiased and dispassio nate
manner so that the co urts are able to better evaluate the
specialist evidence put before them.
The investigation and interpretation of findin gs of
alleged ill treatment of infants and children requires a
multidisciplinary app roach , centred on the child, his or her
well-being in both the short term and longer term, as well
as that of any siblings within the same environment. All
of the avai lable information about any in cident must be
care full y collected, collated and evalu ated. Laboratol),
data, both clinical and forensic, the results of radiological
investigatio ns and information from the examinat.ion of
the scene where any incident took place sho uld be carefully
sought and evaluated against the clinical findings. A team
approach is essential, with close collaboration of famil y
physicians, paediatricians involved in both community
and hospital practice, the clinical fo rensic medical examiner
and specialist pa tho logists, together with police and social
welfare serv ices. No incident should be looked at in isola­
tion but rather in the context of the child's development
and interaction with his or her family, environment and
peer group. The survivors of inflicted injul)' or neglect in
childhood must be carefully followed up, protected and
their family unit supported.
~-- ----~- -------­
. .
I
In this book, some of the top ics covered here are rele­
vant specifically to maltreatment in ea rly life, beginning
with the examination of an infant or child for whom ab use
is suspected, incorporating the family environment and set
against criteria for normal deve lopment. The difficult prob­
lem of suspected sexual abuse of children is considered
separately. The extensive clinical experience of the authors
of the opening chapters is readily apparent, highlighting,
as they do, the pitfalls of incomplete investigation and iIl­
considered interpretation. The ap propri ate level of investi­
gation of specific findings, interpretation of investigations
and consideration of differential diagnoses are addressed
in chapters contributed by a paediatric radiologist, a haema­
tolo gist and a clinical pathologist, respectively. Those areas
requiring specialist clinical expertise and experience - the
eyes, mouth a nd central nervous system - are considered
by specialists in those fields with ex tensive paedi atric
experience. The examination of the scene of death or injury
is discussed as a backg round to post-mortem examination
of the very yo ung. The interpretation of cerebral pathology
in the newborn, the investigation of sudden or suspicious
perinatal death and sudden death in both infa nts and older
children are addressed by experienced practitioners. Sepa­
rate consideration is given to sudden or suspicious deaths
that occur in hospital.
In subsequent chapters, more general areas of forensic
pathology, including asphyxia and thermal injury, drown­
ing, injury to road users and oth er accidents are add ressed
from a paediatric viewpoint. A similar approach is evident
in the chapters covering toxicological investigation, DNA
profiling and dental identification.
The book concludes with consideration of the role of th e
expert witness in criminal judicial cases and the provision
of reports in the civil medicolegal context.
Although the majority of contributors to this te xt are
UK based , the subject matter is presented, as far as possible,
wit hout national or geographic bias, so that the contents
have in te rnation al releva nce.
Anthony Busuttil
Jean W Keelin g
January 2008
 I ACKNOWLEDGEMENTS
We would like to thank our contributing authors for their
hard work and for their patience and ready responses in the
li g ht of requests for updates and answers to specific
quelies. Colleagues in Edinburgh and elsewhere in the UK
have read the Editors' contribu tions and made useful sug­
ges tions. JWK thanks Dr Roger Malcomson for his assistance
I
and expertise in the preparation of the illustrations for her
chapters. We would like to thank the many staff at Hodder
Arnold with whom we have been involved fo r their expert­
ise and encouragement ; in particular, Philip Shaw and our
Project Editor Amy Mulick for their most helpful sugges­
tions in the late stages of manuscript completion.
 I LIST OF ABBREVIATIONS USED I

AA amino acid CPP cerebral perfusion pressure

ARE accident and emergency CPR cardiopulmonary resuscitation
AC alternating current CPT cumulative pressure-time index
ACD a cid-citra te-de xtrose CPT 11 carnitine palmitoyltransferase type "
ACTH adrenocorticotrop hic hormone CPTlD carnitine palmitoyltransferase type 1
ADH antidiuretic hormon e deficiency
ADP adenosine diphosph ate CPT2D carnitine plamitoyltransferase type 2
ADPKD autosoma l dominant polycystic kidney deficiency
disease CSF cerebrospinal fluid
AIDS acquired immune deficiency syn drome CSM crime scene manager
ALTE appa rent life-threatening events CT computerized tomography
APOE apolipoprotein E CVP central venous pressure
aPTT activated partial thromboplastin time CVR cerebrovascular resistance
ARDS adu lt respiratory distress synd rome CZE cap illary zone electrophoresis
ARVD arrthythmogenic right ventricular dyspla sia
ATP adenosine triphosphate DAB DNA Advisory Board
AV atrioventricular DAI diffuse axonal injury
AvD0 2 arteriovenous ditTerence of oxygen DAVlD disaster and victim identification
DC direct current
0APP beta amyloid precursor protein DIC disseminated in travascular coagulation
BPA British Paediatric Association DMF decayed , missing and filled teeth
BPNA British Paediatric Neurology Association DNA deoxyribonucleic acid
BUN blood urea nitrogen DRVvr dilute Russell's viper venom time
DVD digital versatile disc
CACTO carn itine acyJcarnitine translocase deficiency
CAP common approach pathway ECF extracelluar fluid
CAPMI computer-assisted post-mortem identification ECG electrocardiogram
CASK ca rer-associated serial kill ing EDH extradural haemorrhage
CBF cerebral blood flow EDTA ethylenediamine tetra acetic acid
CD compac t disc EEG electroencephalography
CESDI Confidential Enquiry into Stillbirth and Deaths EFE endocardial fibroelastosis
SUDI in Infancy Sudden Unexpected Death in ELISA enzyme- linked immunosorbent assay
In fancy EM electron microscopy
CFAM ce rebra l function analys in g monitor EMIT enzyme-multiplied immunoassay technique
CFIR cystic fibro sis transmembrane conductance EPP polypropylene
regulator EPS expanded polystyrene
CHD congenital heart disease EPU expanded polyurethane
CHIRPP Canadian Hospital Injury Reporting and ERG electro retinograp hy
Prevention Program ESR erythrocyte sedimentation rate
CI confidence interval EVG elastic van Geison stain
CKI cytokeratin I
CMR0 2 cerebral metabolic rate for oxygen FAa fatty acid oxidation
CMV cytomega lovirus FBI Federal Bureau of Investigation
CNS central nelVOUS syste m FOP fibrinogen degradation product
CNV copy number variation FHM familial hemipl.egic migraine
COHB ca rboxyhaemoglobin FIl fabricated or induced illness
CaDIS Combined DNA Index System FLAIR fluid-attenuated inversion recovery
CPI combined paternity index FPIA fluoresce nt polarization immunoassay

~ ~----------

.
.
.
 List of abbreviations used I xvii

FPP fitness to practice panel MDA 3,4-methylenedioxyamph etamine

MDMA 3,4-methylenedioxymethamphetamine
GAS group A streptococcal infection MECC micellar electrokinetic capillary
GC-MS gas chromatography-mass spectrometry chromatograp hy
GCS Glasgow Com a Scale MELAS mitochondrial encephalomyopathy
GDP general dental practitioner MERRF myoclonic epilepsy with ragged red fibres
GI gastroin tes tin a I MfV mean flow volume
GMC General Medical Council MPS mucopolysaccharide
GMD genetic metabolic disorder MRI magnetic resonance imaging
GOS Glasgow Outcome Score mtDNA mitochondrial DNA
GSD glycogen storage disease MS/MS tandem mass spectrometry
MSUD maple syrup urine disease
HbF fetal haemoglobin MVC motor vehicle collision
HCM hypertrophic cardiomyopathy MVF mean flow volume
HDN haemorrhagic disease of the newbo rn
HE haematoxylin and eosin (e.g. HE stain) NAAT nucleic acid amp lification test
HELLP haemolysis, elevated liver enzymes, NAHI non-accidental head injury
low platelets NAI non-accidental injury
HIE hypoxic-ischaemic encephalopathy NAIT neonatal alloimmune thrombocytopenia
Hll hypoxic-ischaemic injury NEC necrotizing enterocolitis
HIV human immunod eficiency virus NEQAS National External Quality Assessment
HL H haemophagocytic Iymphohistiocytosis Scheme
HMWKS hi gh-molecular-weight kininogens NICHD National Institute of Child Health and
HPLC high-performance liquid chroma tography Development
HPV human papillomavirus NKH non-ketotic hyperglycinaemia
HSV herp es simplex virus NSPCC National Society for the Preven tion of Cruelty
to Children
ICD International C1a ssificatiol1 oj Diseases
JCH intracrania l haemorrhage OECD Organisation for Economic Co-operation and
ICP intracra nial pressure Development
ICU intensive care unit OJ osteogenesis imperfecta
lDDM insulin-dependent diabetes mellitus OR odds ratio
IgA immunoglobulin A OTS ornithine transcarba milase
IgE immunoglobulin E OXPHOS oxidat ive phosphorylation
IPH idiop athic pulmonary haemosiderosis
ISS inUlY severity score Pac0 2 partial pressure of arterial carbon dioxide
ITP idiopathic thrombocytopenic purpura PAI-I plasminogen activator inhibitor-I
lTU intensive therapy unit PBR Perls ' Prussian blue reactio n
IUGR intrau terine growth restriction PCR polym erase chain reaction
IVF in vitro Ferti Iization PDH pyruvate dehydrogenase
rVB intraventricu lar haemorrhage PET posit ron emission tomography
PICA posterior inferior cerebellar alielY
LCAD long-ch ain acyl-CoA deficiency PICU paediatric intensive care unit
LCN low copy num ber PK prekallikrein
LCHAD long-chain 3-hydroxyacyl-CoA PL phospho lipid
dehydrogenase PM post-m ortem examination
LQTS long QT syndrome PMCTD plasma membrane carnitine transpolier
LM laser microdissec tion deficiency
LoC loss of consciousness PSA phosph ate-specific antigen
LSD lysergic acid diethylamide PT prothrombin time
PTA post-traumat ic amnesia
\-1A DD multipl e acyl-CoA dehydrogenase deficiency
\,IlAP mean arterial pressure RBC red blood cell
.\1 CA middle cerebral artery RIA radioimmunoassay
\ilCAD medium chain acyl CoA dehydrogenase RCPCH Royal College of Paediatrics and Child Health
deficiency RFLP restriction fragment length polymorphism
 I CHAPTER 2
INVESTIGATION OF SUSPECTED
SEXUAL ABUSE
Jacqueline YQ Mok
Introduction 24
The colposcope in the medical examination 27
Forensic evidence 27
Skills and experience required 28
Consistent vocabulary 29
Normal female genital anatomy 29
Perianal findings 36
Acute, healing and healed anogenital trauma 37
INTRODUCTION
The widespread occurrence of child sexual abuse has been
Sexual abuse has been defined as the involvement of depend­
ent, developmentally immature children and adolescents in
sexual activity that they do not fully comprehend and to
which they are unable to give informed consent or that vio­
late the social taboos of family roles. l In this situation, there
is an imbalance of power between abuser and abused, and an
element of control of the child by a trusted adult. Sexual
abuse ranges from violent assault to gentle seduction. The
activities include all forms of digital-genital , oral-genital,
genital-genital contact between abuser and child, as well as
non-contact abuse such as exhibitionism and use of the child
in the production of pornographic material.
A diagnosis of sexual abuse has both civil and criminal
implications. The medical profession 's early involvement
with child sexual abuse was limited to psychiatrists, who
were interested in the behavioural manifestations following
sexual abuse. The subject received scant paediatric recogni­
tion until publications by Jaffe 2 and Kempe3 changed the
willingness and ability of paediatricians to recognize and
deal with the problem. Paediatricians must have a high index
of suspicion and be willing to consider sexual abuse as a dif­
ferential diagnosis when a child presents with behavioural
problems or somatic symptoms that suggest that sexual
abuse might have occurred. All physician s should act accord­
ing to their local child protection procedures, with the welfare
and protection of the child as paramount considerations.
- ~ ----- ­ - -- - ----- --
I
Female genital findings in sexual abuse 38
Signs of anal abuse 39
Conditions that mimic abuse 40
Screening for sexually transmitted infections 41
Interpretation of clinical and laboratory findings 42
Summary 43
References 44
known for many years, although the exact prevalence is diffi­
cult to define. Prevalence studies have been canied out by
interviewing adults about childhood events. In a summary of
19 studies conducted in the USA or Canada between J 980 and
1994, Finkelhor4 found that the rate of sexual abuse reported
by women varied from 2 per cent to 62 per cent, with an
approximate prevalence of 20 per cent. The ra tes reported by
men varied from 3 per cent to J 6 per cent, with a reasonable
estimate of approximately 10 per cent. However, studies based
on interviews of adults provide limited information owing to
differences in study design, the response rate, methods of data
collection, definition of sexual abuse, definition of a child, and
the accuracy of recall of events that might be traumatizing.
Data from studies of reported incidences of abuse pro­
vide information about the number of children recognized
annually, usually at individual centres. It does appear
that increasing numbers of children are referred because of
suspicions of sexual abuse, owing to either a t ru e increase
in the occurrence or better recognition of the problem.
In the USA it is estimated that between 120 and 150 per
10000 children have been subjected to sex ual abuse.
More recently, Jones et a1 5 identified a significant reduction
in the incidence of substan tiated child sexual abuse in the
USA, and a population-based study in Australia has also
provided evidence of a decline in the underlying rate of
child sexual abuse.6 Such findings may indicate the effect­
iveness of personal safety progra mmes for young children.
---,------­

An earlier paper from one city in the UK 7 described 51
children in 1985 who presented after they disclosed abuse,
a number that rose to 79 in the subsequent year. Thirty­
eight per cent of the children were aged less than 5 years of
age at diagnosis and the mean age was 8 years. Of the
alleged abusers, 60 per cent were related to the child, and
one-half of these were natural fathers. Because the vast
majority of children are abused by someone they know and
trust, force and restraint are not commonly used when
abusers engage children in sexually inappropriate activities.
There is usually little intent to harm the child, and bribes or
threats are often used to prevent the child from reporting
the abuse. Children are ideal victims for sexual exploitation
and abuse, as they are vulnerable and see adults as trusted
and powerful. The 'child sexual abuse accommodation syn­
drome's describes the stages of traumatic sexualization,
feelings of betrayal, powerlessness and entrapment that
lead to the child's accommodation of the abuse. As a result,
very few children disclose abuse immediately following the
event, making the retrospective interpretation of healed
trauma a great diagnostic challenge.
The Medical Evaluation
Child sexual abuse should rarely, if ever, be diagnosed on
the basis of physical signs alone. A clear statement from
the child is the single most important factor towards mak­
ing a diagnosis. The medical evaluation of children sus­
pected to have been sexually abused should be part of a
multidisciplinary process that spans the investigative,
diagnostic and therapeutic needs of the child and family.
The medical examination should be a comprehensive
health assessment that should aim to:
• establish any need for immediate treatment;
• provide background information that mayor may not
support the diagnosis;
• provide information or evidence to sustain criminal
proceedings and/or care plans;
• plan or coordinate ongoing care;
• reassure the child and family.
A minimum of physical examinations should be con­
ducted as is necessary, and if several medical specialists are
to be involved it is desirable that they should examine the
child together. Joint examinations can be performed by a
paediatrician and a forensic medical examiner (police sur­
geon) to encompass, in a single examination, the child's
medical needs with the legal requirements for evidence. 9
Although a carefully structured approach to the examin­
ation is required for legal purposes, the medical assessment
should be compassionate and thorough, resulting in a thera­
peutic experience for the child and family.lO.11 The examin­
ers must be both familiar and comfortable with normal
childhood behaviour and development, genital and anal
anatomy, as well as physical findings of abuse. Core skills
Introduction I 25
and case-dependent skills for any paediatrician or forensic
medical examiner who undertakes such examinations have
been defined by the Royal College of Paediatrics and Child
Health and the Faculty of Forensic and Legal Medicine
(October 2007).12 Core skills include: the ability to commu­
nicate with children and their carers about sensitive issues;
an understanding of the child's developmental, social and
emotional needs; a knowledge of consent and confidential­
ity issues as they relate to children; an understanding of
the range of normal genital and anal anatomy for the age
and gender of the child to be examined; an ability to docu­
ment and interpret the clinical findings; competence in the
production of a report; a willingness to communicate and
co-operate with other agencies; and the aptitude to present
the evidence and be cross-examined in civil and criminal
proceedings.
History
As with all medical consultations, the starting point is in
the history, taking care to avoid asking leading questions.
Investigative interviews are usually conducted by desig­
nated agencies (police and social services) to avoid repeti­
tive questioning of the child. This should not preclude
physicians asking relevant questions that are essential to
the medical examination. In eliciting the history, the clin­
ician should approach the child in the same manner as he or
she would in any other paediatric condition. Details should
be obtained about the child's birth and development, past
medical history, and family and social background. The
systematic enquiry should cover symptoms and signs relat­
ing to the genitourinary and gastrointestinal systems.
Specific questions should be directed to the presence of
pain, itch, rash, discharge or inflammation 'down below',
episodes of bleeding (on pants or on toilet paper), and fre­
quency and consistency of bowel movement. When appro­
priate, girls should be asked the menstrual history, type of
sanitary protection used, sexual history and previous
gynaecological examinations. The child's terminology for
the various body parts should be documented, and any
statement made by the child recorded verbatim. 13
The importance of the psychological aspects of sexual
abuse, both in the short and long term, has been high­
lighted. 14 ,15 Emotional difficulties include anxiety, sadness,
anger, behaviour problems, school refusal, sleeplessness,
withdrawal and sexualized behaviour. Somatic complaints
that have been reported include: eating disorders, abdom­
inal pain and headaches, as well as loss of bladder and bowel
control. The medical evaluation should include an enquiry
into the child's emotional status and general well-being, as
the carers may not associate behavioural difficulties with a
past history of abuse. The carers may even be unaware that
sexual abuse has occurred. Child sexual abuse is a diagno­
sis that, like other paediatric diagnoses, requires a consider­
ation of the history, physical examination and supportive
 26 I Investigation of suspected sexual abuse
laboratory tests when appropriate. 16 Although the medical
history qualifies as 'hearsay evidence', many states in the
USA permit an exception to the hearsay rule, for medical
histories obtained by physici ans,u
Technique of the Examination
Th e exa mination should be carried out in the presence of a
trusted adult, usually the child's mother. The whole child
should be examined, and this includes measuremen t of
hei ght and weight, assessment of the general appearance,
developmental milestones, demeanour and behaviour. As
part of the physical examination, general signs associ ated
with trauma should be sought, such as distribution a nd
pattern of bruises, grip marks, ' love bites', teeth marks and
scratches, as well as injuries within the mouth. In most situ­
ations, the disclosure involves past abuse and th e exam ­
in at ion can be planned to suit the child and family. Delay
should be minimized in the following circumstances, when:
• the a buse has occurred within the previous 72 hours;
• there is a history of acute trauma;
• there is a possibility of pregnancy resulting from the
abuse, so that post-coital contraception can be
prescribed.
Considerable reassurance will be required and to avoid
further distress no force or restraint should be used, no mat­
ter how well intentioned. An explanation ofwhat the exam­
ination entails should be given, with clear emphasis on 'a
health check' and 'taking a look'. Wi th calm reassurance and
an unhurried approach, most children can be examined
without the use of any sedation. The recommended position
for the female genital examination is the supine 'frog­
legged' position, with the hips flex ed and abducted, the
soles of the feet touching. Very young children can be
examined on their mothers' laps. Sometimes the adult can
assist by sitting astride the examining couch, cradling the
child as she leans back against the adult's body.
In the supine 'frog-legged' position , the external geni­
talia should be inspected for signs of injury. Gentle lateral
parting of the labia majora with the examiner's middle and
index fingers (labial separation) allows visualization of the
posterior fourchette, vestibule, perihymen al regions and
urethra; the hymen may be visible at this stage. In order to
visualize th e ma rgins, configuration and size of the
hymenal openin g, labial traction is applied . This involves
grasping the posterior ends of the labia majora between
thumb and index finger and pulling gently up an d out­
wards. The vaginal walls may be demonstrated by this pro­
cedure; however, digital examination of the vagina is
rarely indicated in the prepubertal child. Any sign observed
should be described and documented, using the clock face
to denote the location of the finding; in the anatomical
position, 12 o'clock is anterior while 6 o'clock denotes the
posterior position .
~ ----~ - -- -~ - -- - - - ------ - - ---­
Sometimes the hymen can be difficult to visualize
because of sticky or redundant folds. If labial traction does
not cause the hymen to open, douching with warm water
will result in 'floating of the hymen', Further definition of
the anatomy can be obtained with the use of a cotton bud,
applied behind the hymen to 'tease out' the folds, This
technique should not be used in prepubertal children, in
whom the hymen is extremely sensitive, unless preceded
by the gentle application of a local anaesthetic gel. In ado­
lescent girls, the hymenal edge can also be examined using
the Foley cath eter stretch technique,18 A 14-gauge Foley
catheter is inserted through the hymenal orifice into the
vaginal vault, and infl ated with 40 mL of air. Gentle pulling
toward the hymen results in stretching and displaying of
hymenal tissue over the surface of the balloon,19,20 The
toluidine blue dye test was deve loped by Lauber and
Souma 21 to help detect lacerations in adult rape victims.
Toluidine blue is a nuclear stain that will bind to nuclei in
the deeper dermis when exposed by lacerations in the skin.
In a study of girls "vho alleged sexual abuse, the applica­
tion of a I per cent aqueous solution of toluidine blue dye
was found to increase the detection rate of posterior
fourchette lacerations from 4 per cent to 28 per cent in
adolescents, and from 16.5 per cent to 33 per cent in the
paediatric population .22
The prone knee-chest position is recommended if there
is difficulty in visualizing the posterior hymenal rim. With
t he child resting on her knees and elbows, the buttocks are
parted upwards and outwards, using the examiner's palms.
This technique usually exposes the vestibule, perihymenal
structures and hymen. Any suspicious findin gs in the pos­
terior hymenal rim can be clarified , as the effect of gravity
causes the posterior rim to stretch out.
The perianal area is traditionally examined in the left lat­
eral position. With use of the colposcope, some examiners
prefer to continue with the examination in the supine pos­
ition, by asking the child to 'curl up into a little ball'. This is
achieved by flexion of the child's hips and knees against the
abdomen. The buttocks are gently separated using the palms
of both hands, and the perianal area inspected for signs of
abuse. The external anal sphincter usually relaxes during
this procedure, making it easier to visualize fissures in the
anal margin. Sometimes the ana l canal opens to reveal a
clear view of the rectum. A digi ta l examination of the anus
is rarely necessary and is unhelpful in assessing anal tone.
There is also no evidence to suggest that a digital examin­
ation provides useful clinical information. If there are suspi­
cions of injuries or pathology higher up the anal canal or
rectum, referral should be made to a paedi atlic surgeon for
an examination under anaesthesia.
Most units have a dedicated suite of rooms where chil­
dren can be examined in a child- fr iendly environment. The
minimum requirement is an area that is integrated into a
children's outpatient department to allow access to labora­
tories a nd other investigative facilities but which offers
some privacy. Other important considera tio ns include the

specific needs of adolescents, the gender of the examiner
and the availability of follow-up for sexual health coun­
selling and therapeutic support.
Examination of Boys
The same sensitive and age-appropriate techniques should
be used when examining boys. As with girls, any evidence of
trauma such as bruises, swellings, scratches and bites should
be sought for and documented . The genital examination
should focus on the groin, penis, urethra, scrotum and testes
while the ch ild is supine. The presence of discharge or warts
should also be noted. "'lhen examining the pelianal region,
various positions can be used, depending on the age of the
boy. The young child may be examined in the supine
knee-chest position, immediately after examination of the
penis and scrotum. Older boys may be more comfortable in
the left lateral position, with either the right knee or both
knees pulled up to the chest. The anal examination usually
involves inspection of the soft tissues only. An assessment of
the anal to ne can be made, with the buttocks parted. Find­
ings should be documented either by line drawings or photo­
graphs. The colposcope can be used as a source of light and
magnification, as well as photographic documentation.
THE COLPOSCOPE IN THE MEDICAL
EXAMINATION
The medical examination of children reporting sexual
abuse evolved from a search for ways to improve lighting
and visualization of the young child's genitalia. Methods
for recording observations were also important, to assist in
the interpretation of findings. In the early 1980s when
physicians began to examine the genitalia of children,
instruments such as a hand-held magnifying glass and an
auriscope were used to enhance visualization. In 1925, the
colposcope was invented by Hinselman , to assist in exam­
ination of the female cervix. Teixeira 23 used the colposcope
to examine victims of alleged sexua l assault. Five years
later, Woodling and Heger24 promoted the use of the colpo­
scope in the assessment of chi ldren who alleged sexual
ab use. The colposcope is a no n- invasive magnifying
instrument with a built-in light source, which allows mag­
nification of the external genitalia from 2 to 25 times. It
contains a binocular system of lenses of varying strengths,
co upl ed to an integral light source. The instrument can
either be mounted on the wall, the examination couch, or
on a caster system to enable mobility. Light intensity can
be varied to cope with changes in magnification, and most
instruments incorporate a green filter to improve the visu­
alization of abnormal vasculature or scars.
Tei xei ra reported that an additional 11.8 per cent of sus­
pected cases were corroborated than would have been
when conventional examination techniques were used. In
Forensic evidence I 27
the prospective study by Muram and Elias,25 130 prepubertal
girls (mean age 5.5 years) were evaluated bo th by an
unaided examination and by colposcopy. Overall, 92 girls
were found to have abnormal findings, the majority of
which (96 per cent) were observed during the unaided
examination. Of the four patients in whom findings were
detected initially by colposcopy, these were observed dur­
ing the repeat unaided examination. In only one patient
were the findings observed only by colposcopic examin­
ation alone. The authors concluded therefore that una id ed
examination by an experienced cl ini cian is adequate for
the evaluation of most victims of sexual abuse.
The most obvious advantage of the colposcope is the
integral photographic facility. Documentation of all visible
findings in abuse is increasingly expected as the standard
of good practice. Images produced by the colposcope can
be converted into slides or photographs using a single-lens
reflex camera or a Polaroid camera. Modern video technol­
ogy allows the image to be recorded onto a videotape,
compact disc (CD) or digital versatile disc (DVD) for imme­
diate viewing, and has the advantage over still photog­
raphy as it shows the dynamic variability of anogenital
anatomy as the examination proceeds.
The colposcope is now standard and acceptab le equip­
ment in the examination of sexua l abuse. With photo doc­
umentation, the most obvious benefit to the child is that
there is no need for repeated examinations. Many examin­
ers do not position their eyes in the binocular lenses of the
colposcope; rather they view the image produced on the
monitor at a more comf0l1able position and distance from
the ch ild . The child is also able to observe the examination
on the monitor, and this helps to achi eve a sense of control
and participation in the examination. The instrument is
usually in trod uced to the child to gain cooperation, and
many enjoy playing with it prior to the examination. The
use of high technology is both accepted and expected in
the modern health care system, and is preferable to
attempting to peer between a ch ild's legs wit h an auriscope.
The ability to capture images, either as slides, photographs,
videotapes, CDs or DVDs enhances undergraduate and
postgraduate physician training. 26 - 3o
FORENSIC EVIDENCE
Verba l consent is usually adequate for both the examin­
ation and photo-documentation. This should be obtained
from the person with parental responsibility and from the
child of sufficient maturity to understand the nature and
consequences of the examination. If the child is the subject
of legal proceedings, the consent of the court is required.
The person obtainin g consent must record the procedure in
the case notes. Written consent for each component of the
examination can be documented on a standard form and
on appropriate forms.
 28 I Investigation of suspected sexual abuse
Correct procedures must be followed for the collection
of forensic sampJes and evidential material, recognizing
the principle of the unbroken chain of evidence. This legal
concept requires that the origin and history of any exhibit
to be presented in a court of law must be clearly demon ­
strated to have followed an unbroken chain from its source,
through its examination and to the court. A note of the
persons handling the sample, time, date and place where
the sample was obtained, along with the places and condi­
tions of storage must be documented. In an acute assault
(within 72 hours), the clothing worn during the assault
should be coI1ected individually in paper bags and handed
to the investigating police officer. The skin should be
inspected for a ny stains, and swabs taken to examine for
saliva, lubricant or semen. If judged appropriate, a sample
of blood should be taken for DNA analysis. Stains on cloth­
ing are best preserved by drying and storing at cool room
temperature. Blood or semen collected on swabs is best
preserved by frozen storage. The presence of lubricant or
spermatozoa can be sought from swabs in and around the
mouth (taking care to swab in the labial-gingival sulci) and
from saliva. Young children will tolerate external and
internal anal a nd vaginal swabs, if the a ppropriate size of
swabs is used, moistened with water. Spermatozoa can be
detected for up to 14 hours in saliva J I 3 days in anal sam­
ples J 2 and 6 days in vaginal samples. JJ The time limits for
detection of seminal fluid are 3 hours in the anus and
J 2- J 8 hours in the vagina. Improved DNA diagnostic tech­
niques allow detection of useful evidence from unwashed
or partially washed clothing, bedding and other items used
during the assault. General guidelines for the collection of
forensic evidence in cases of acute sexual assault a re not
well suited for prepubertal children. In a review of 273
child victims of sexual assault aged less than 10 years, the
majority (64 per cent) of forensic evidence was found on
linen or clothing. Over 90 per cent of children with positive
forensic evidence were seen within 24 hours of the assault.
After 24 hours, all evidence, with the exception of one
pubic hair on a child, was recovered from clothing or
linens. The authors suggested that any collection of foren­
sic evidence from a prepubertal child 's body may not be
necessary more than 2 days after the assault. J4 In cases of
chronic abuse, when the last episode of contact was more
than a week previously, collection of evidence consists of a
careful histolY, thorough examination and screening for
sexually transmitted infections (STls) .
Although a highJy trained and experienced examiner is
not likely to miss any abnormalities during an unaided
examin ation, the colposcope allows a review of photographs
and video recordings when findings thought to be suspicious
during the examination might be interpreted on review as
normal or non-specific. Permanent documentation of visible
findings protects the examining practitioner by providing
objective evidence, thus only the interpretation of the find­
ings can be challenged. Still photographs, videotapes or CDs
of the injuries often provide key information for assessment
-- ~-------- - - ----­ -- ---
and evidence for prosecution in child abuse cases. Tradi­
tional systems for photographing abused children using a
single-lens reflex camera still provides the highest quality
image. The use of Hi-8 and Super VHS video cameras offers
significant improvements in image quality and resolution
over the traditional videotapes. More recent developments in
computer imaging technology have resulted in greater ease
in transforming analogue images to a digital format for
computer storage and conversion into slides or prints.
[mages can also be converted into a computer file, which
assures preservation without degradation of the initial
image. A video capture card, a scanner and/or a digital cam­
era, allow a computerized database to be developed. With
appropriate software products, images can be transmitted
over the Internet, enabling a second opinion to be sought
and thereby brings telemedicine into remote areas. J 5,J6
Photographs, video, CD and DVD recordings must be
properly verified and relevant, they must bear the patient
identifier, date and time of recording, and must be signed
by both medical and forensic examiners. Images are neces­
sary to explain or illustrate adequately the complexity of
the injuries, and are therefore relevant for courtroom pro­
duction. However, the production of sensitive images and
photographs of children's genitalia in the courtroom is not
common practice and should be discouraged. An alterna­
tive line diagram can always be used to illustrate the
injuries. All recorded material should be available to any
medical expert instructed by defence solicitors.
SKILLS AND EXPERIENCE REQUIRED
Although medical findings are not necessary for the legal
conclusion of sexual abuse, health-care professionals who
examine children for signs of sexual abuse are often asked to
render an opinion as to whether their examination was nor­
mal, non-specific, suggestive or indicative of abuse. Despite
this important role, the literature contains little information
regarding examiner competence in assessing children's geni­
talia. Undergraduate and postgraduate training on the subject
of child sexual abuse is woefully inadequate. The medical
investigation of children in whom abuse and neglect is
suspected has become a complex and technical specialty.
A new group of physicians has emerged, who are specialists
in the wide ramifications of child abuse and neglect. A
sound knowledge is expected in general, behavioural and
developmental paediatrics, as well as in . gynaecology and
infectious diseases. The child abuse specialist also needs to be
familiar with forensic examinations, civil and criminal laws,
child advocacy and public policy, and must be comfortable
when testifying in court. For these reasons, structured train­
ing programmes must be developed to improve the quality of
care delivered to abuse children and their families.37
Brayden et al J8 evaluated the interobserver reliability of
clinicians rating colposcopic photographs, and examined
 Normal female genital anatomy I 29

correlates of reliable interpretations. Seventy-one physici ans studies and a consistent application of established terminol­
and two nurse practitioners were asked to assess blindly six ogy that can prevent the misinterpretation of non-specific
photographs of prepubertal female genitalia, five of which or congenital findings as post-traumatic changes.
were taken from children who had given histories of sexual
abuse. Experts in the field of child sexual abuse assessment
made significantly more accurate assessments than paedia­
NORMAL FEMALE GENITAL ANATOMY
tricians, paediatric and family practice residents, and trainee
physicians. In another study, Adams and Wells 39 sought to
Figure 2.1 is a diagrammatic representation of the prepu­
determine how well medical examiners agreed on the
bertal female genitalia, with the parts labelled using recom­
significance of certain anogenital findings in children, by
mended terminology. The most consistent landmark is the
showing colposcopic photographs of 16 patients to 170 med­
clitoris, which is usually prominent in young girls because
ical examiners who were blinded to the history of each
of the lack of subcutaneous fat in the surrounding tissues.
patient. The agreement between the participants and the
The vestibule is the area enclosed within the labia minora,
experts on the abnormal cases (mean 81 per cent) was signif­
and includes the urethral opening which can be pinpoint or
icantly higher than on the normal cases (mean 71 per cent,
patulous. Bilateral lines (Hart's lines) drawn medially from
p = < 0.001). There was also higher agreement on genital
the posterior ends of the labia minora converge at the pos­
findings (78 per cent) than on anal findings (63 per cent,
terior fourchette. The introitus refers to the hymen and
p = 0.000). In the participants, higher experience level and
hymenal opening or orifice. Sometimes the vaginal walls
use of a colposcope were associated with higher overall
can be seen through the hymenal opening. The area
agreement with the experts (74 per cent versus 44 per cent,
between the posterior fourchette and inferior edge of the
p = < 0.0001). The history was also found to influence
hymen is referred to as the fossa navicularis. Between the
physicians' interpretation of girls' genital findings. Paradise
posterior fourchette and the anus, deep within subcuta­
et al 40 conducted a questionnaire survey of 1387 randomly
neous tissues, is the perineal body.
selected fellows of the American Academy of Pediatrics and
In order to identify the signs of sexual abuse, the clinician
all 802 members of child abuse professional groups, using
must have a sound knowledge of normal anogenital anatomy
seven simulated cases, in 6 of which the histories were
and its variants. Contemporary textbooks provide little
changed at a second mailing 4 months later. The proportion
insight into the characteristics of hymenal anatomy or the
of changed or reversed opinions varied from 0 per cent to 5.6
changes of the hymen in response to oestrogen levels during
per cent amongst experienced examiners; 1.6-19.8 per cent
the childhood years. The appearances vary considerably
amongst moderately experienced examiners and 3.6-27.2
according to the age of the child. In the newborn, the effects
per cent where the examiners were inexperienced. The like­
lihood of an interpretation being changed was influenced by
a diagnostic expectation (change in history) and ambiguity
of the photographs. Labium-­

majorum

CONSISTENT VOCABULARY Clitoris - - - - H - t - ­

Medical examiners are often asked to determine whether a

child has been sexually abused and whether penetration has
occurred, as well as how often a child has been abused.
More often than not, the physical appearances will be nor­
Labium - - i - t ' ­
mal or non-specific. In preparing the medical report, care minorum

(}L-1--\;\---\---'.-- Ure th ra I
must be taken to avoid terms that can be misinterpreted and opening

lead to confusion. Subjective descriptions such as a 'lax'

sphincter or a 'gaping' hymen are unhelpful unless sup­
Hy menal}2
ported by measurements. It is important that clinicians are orifice 'e
aware of the recommended terminology used to describe the -.....;:::7'--+-7--.1---- Hymen .E

--tt
Fossa ----''<---'..--~-
genitalia, as well as interpretation of anogenital findings, so navicularis ...... - _ _~,-'--_ Posterior

P"~,~:
that those who examine children for alleged sex ual abuse
can understand each other's descriptions. 41 - 43 Practice
guidelines are now available and should be widely dissemi­ "",,h""
nated. 44 - 47 In an excellent review of hymenal morphology
and non-specific findings in girls selected for non-abuse,

Heger et al 48 called for a thorough understanding of normal Figure 2.1 Prepubertal female genitalia.

Table 2.1 Cross-sectional studies of normal female genital anatomy
Authors
McCann et al 1990 52
Berenson et al 1991 57
Beren son et al 1992 53
Gardner 1992 55
Yo~an a ndYo~an 1992~
Emans et al 1994 119
Beren so n et al 2000 56
Heger et al 2002 81
Number of subjects Age Summary of common findings
93 10 month s-lO years Erythema of vestibule 56%
Periurethral bands 51%
Labial adhesions 39%
Hymenal mounds or projections 34%
Longitudinal intravag inal ridges 90%
Intravaginal rugae 89%
468 Ne wbo rn s Annular hymen 80 0/0
Fimbriated hymen 19%
Hymenal clefts (anterior segment only) 34%
External hymenal ridges 87 0/0
Longitudinal intravag in al ridges 53%
Hymenal tags 13 0/0
211 1 month-7 years, Vest ibular bands 98 0/0
mean 21 ::!: 20.6 months < 12 months, fimbriated hymen 46%
> 24 months, cresce ntic hymen 51%
Transverse hymenal diameter increa sed with age
79 3 month s-11.5 years, Increased vascularity 44%
mean 5.3 years Midline avascular area s 27%
Ragged posterior fourchette/notch in posterior fourchette 28%
Peri hymenal teth ers 14%
Hymena I bu mps 11 %
168 7-17 years Effects of puberty - labia minora increased in size and pigmentation; hymenal tissue increased in
thickness; increa sed vaginal se cretion s
300 Median 18 yea rs Complete hymenal clefts seen significantly more frequently in girls who were sexually active; not
related to participation in sports, prior pel vic examination or tampon use
Median hymenal diameter increased significantly with sexual activity
200 3-8 years, case - co ntrols Vulvar features: erythema 37%; partial agglutination 7010; linea vestibularis 7%; vaginal disch arge
for abu sed childr en 4%; friability 4%; perineal depression 2%
Hymen al features: periurethral band s 95%; longitudinal intravaginal ridg es 87%; vestibular bands
60%; bumps 46%; external ridges 80/0; prominent vessels 7%; superficial notches 5%; tags 5%
147 Mean 63 (::!:38) months Hymenal configurations: annular 53%; crescentic 29.2%; sleeve like 14.9%; sep tate 2%; other < 1%
Non-spec ific findings: perihymenal bands 91.8%; longitudinal intravaginal ridges 93.8%; hymenal edge
irregular 51.7 0/0 , thickened 45.5%, narrowed 22.4%; erythema 48.9%; vascular changes 37.4%; hymenal
bumps/mounds 34% ; hymenal opening size > 4 mm 30.6%; posterior hymenal concavity 29.5%; partial
posterior hymenal notch/cleft 18.3%; labial ad hesion s 15.6%; failure of midline fu sion 0.6%

of maternal oestrogens result in promin ence of the labia and
clitoris, a redundant and fleshy hymen and copious vaginal
secretions. In early infancy, the labia majora gra dually fold
across the introitus to form a protective pad. As the oestrogen
effects wear off, the labia majora appear flattened and the
labia minora are seen as thin folds. The inner margins of the
labia minora may be set deeply and easily mistaken for
the hymenal orifice.
Although there are several cross-sectional studies on
genital appea rances throughout childhood, a methodo­
logical flaw of most of these studies is the inclusion of girls
who may have been sexually abused, as the authors have
relied on parental history alone to exclude abuse. These
cross-sectional studies have included girls over multipl e
age groups, making it difficult to eva luate findings by age.
Table 2.1 summarizes some studies that have attempted
to define normal genital anatomy in non-abused girls.
Although cross-sectional studies cannot assess the forma­
tion of new features such as bumps or notches, they have
been useful in clarifying the range of normal findings, as
well as helped to define hymenal configuration with age.
Longitudinal studies have been useful in documenting
anatomical changes over time. Reports of longitudinal
studies have mainly come from one centre in the USA 49,50
and are summarized in Table 2.2.
Labial, Posterior Fourchette and
Vestibular Findings
Yordan and Yordan 51 have documented changes in the labia
minora with age and onset of puberty [see Table 2.2). In the
absence of oestrogens, the labia minora appear small, thin and
without pigment. The epidermis of the vestibule is thin and
smooth. The fossa navicularis is characterized by a network of
Table 2.2 Longitudinal studies on normal female genital anatomy
Authors Number of subjects Ages examined
Berenson 1993 49 62 Birth and 1 year
Berenson 1995 50 134 <2 months and about
3 years; 42 also examined
at about 1 year
Normal female genital anatomy I 31
fine submucosal blood vessels that extend onto the edge of the
hymen . The hymen itself is thin and almost translucent. With
the onset of early puberty, there is a reduction in the superfi­
cial vascular prominence in the vestibule, fossa navicularis
and hymen. As puberty progresses, the labia majora and
minora become pigmented. Secretions are abundant in the
vestibule, which becomes darker and more textured in its
appearance. Subcutaneous deposits of adipose tissue may be
seen. The hymen becomes fleshy and redundant, making
detailed examination of the rim difficult.
Labial adhesions, either partial or extensive [Fig. 2.2),
have been reported in 5-39 per cent of girls studied. 52 ,53
The presence of labial adhesions does not imply that sexual
abuse has occurred, as they could result from poor hygiene,
inflammation or recurrent infection, especially in a young
child still in nappies.47
Increased friability (Fig. 2.3) is defined as a smal l dehis­
cence of the tissues of the posterior fourchette, with or with­
out bleeding. This is usually seen towards the end of the
examination and has been reported in 4.7 per cent of girls
examined in the cohort reported by McCann. 52 In a minolity
of girls abnormalities that had previously been documented
in sexually abused girls were noted. These included
increased vascularity, mid-line avascular areas, as well as
urethral dilatation (Fig. 2.4). Although it is likely that sexu­
ally abused girls may have been inadvertently included in
these cohorts, these studies highlight the dangers of over­
interpreting minor anatom ic findings that may lie within the
range of normal variation. 54 .55 In a case-control study of
anatomic changes resulting from sexual abuse, no difference
was observed in the percentage of abused versus non-abused
children with regard to labial adhesions, increased vascular­
ity, linea vestibularis, friability, a perineal depression, or a
hymenal bump, tag, longitudinal intravaginal ridge, external
ridge, band or a superficial notch. Furthermore, the mean
Summary of findings
Decrease in hymenal tissue seen in 33 out of 57
Hymenal notches seen for the first time in nine girls;
resolved in five
Hyme nal tags seen for first time in four girls; resolved in two
Posterior hymenal notches not observed
Change from annular/fimbriated to crescentic hymen in
87 out of 134
External hymenal ridges resolved by 3 years
Tags decreased by 1 year; resolved by 3 years
Longitudinal intravaginal ridges and periurethral bands more
easily seen at 3 years
Mean horizontal and vertical transverse hymenal diameters
increa sed with age
Posterior hymenal rim did not va ry with age; > 2 mm
Racial differences in hymenal anatomy
 32 I Investigation of suspected sexual abuse
Figure 2.2 Extensive adhesions of the labi a majora in a l -year­
old girl, obscuring introitus. The labia minora are seen at the
anterior end of the adh esions.
Figure 2.3 Friable posterior fourchette in an 8-year-old girl,
with sp litting and bleeding at the end of the examination. Notch
seen at 5 o'clock position of hymen.
Figure 2.4 Imperforate hymen in a 10-year-old girl, with a
dilated, patulous urethra that could be mistaken for hymenal
orifice.
- ----- ~--------- - - - --
number of each of these features per child did not differ
between the two groups of children. Vaginal discharge was
seen more frequently in abused children (11 per cent versus
4 per cent, p = 0.01) especially if the abuse took place within
the last 7 days, involved penile penetration and was reported
on three or more occasions. 56 This finding is not pathogno­
monic for abuse, however, as infection, allergy or the pres­
ence of a foreign body can cause vaginal discharge.
Hymenal Configuration
In the prepubertal chi ld, the hymen can be fimbriated (with
a ru ffled edge). crescentic (absence of tissue anteriorly),
annular (tissue present all round, 360°), imp erforate, sep­
tate or cribriform. At birth, the annular sleeve-like hymen
is th e most commonY A decrease in hym enal t issue is
observed by the age of 1 year in approximately two-thirds
of the children49 and by th e age of 3 years, a crescentic
hymen is th e most predomin ant finding.5o Variations of
hymenal morphology have been reported consistently in
all studies of norm al anatomy and can be explained by the
lack of consistency in terminology, as some authors c1as­
si fy annular hymens with ventral clefts as crescentic.
Exam ples of an imperforate, an annular and a crescentic
hymen are shown in Figs 2.4-2.6.
Hymenal Ridges, Bumps, Tags and Bands
External hymenal rid ges are usually seen between the
external surface of the hymen and the vestibular wall, either
in the 6 o'clock pos ition, or in the sub-urethral region. This
congenital feature has been observed in up to 80 per cent of
newborn in fan tsY There are no published data on the fre­
quency of external ridges in children older than four years
Figure 2.5 Annular hymen in a 3-year-old girl, with
peri hymenal bands at 10 and 2 o'clock.
- - ----- -
 Normal female genital anatomy I 33

Figure 2.6 Crescentic hymen in a 6-year-old girl. Note the Figure 2,7 Bump or mound of tissue at 6 o'clock on a crescentic
absence of hymenal tissue between 11 and 1 o'clock and the hymen.
translucent smooth-edged posterior rim.

old, but it is likely that the majority of these ridges resolve

by three years 01d. 50 External ridges may persist in some
individuals and should not be interpreted as a sign of abuse.
Longitudinal intravaginal ridges that extend from the
vagin a onto the internal surface of the hymen have been
described in all four quadrants of the hymen. Although
first described in a bused children,58 they have also been
found in cohorts of non-abused girls and are likely to be
congenital findings, occurring in 56 per cent of newborn
females. 57 Beyond the neonatal period, longitudinal intrav­
aginal ridges have been reported in 25-94 per cent of girls
(see Table 2.1, p. 30), and these differences may reflect vari­
ations in definitions or terminology. Interestingly, there
may be racial differences as intravaginal ridges were noted Figure 2.8 Fleshy tag, arising from between 10 and 11 o'clock
to occur more frequently in black girls than in Caucasian or on the hymen, and resting on right labium minus.
Hispanic girls.50
A hymenal bump represents a localized, rounded, thick­
ened mound of tissue on the edge of the hymen, is wider
than it is long and can be seen anywhere on the edge (Fig.
2.7). This feature has been reported in 7 - 34 per cent of girls reported in up to 98 per cent of newborn in fants57 and
without a history of abuse, and in older studies, hymenal 92 per cent of prepubertal girls. 4B Perihymenal or vestibular
bumps or mounds were thought to represent an area adja­ bands are often symmetrical, and create a semilunar­
cent to a healed tear in the hymen. 58 More recent studies shaped space between the bands, which ca n be mistaken as
on children selected for non-abuse suggest that this is a the hymenal orifice (see Fig. 2.9).
common and non-specific fe ature, and may represent an
area where a longitudinal intravaginal ridge ends at the
hymen. 48 ,52,53,56 Hymenal Notches, Clefts and Transections
A tag is an elongated projection of tissue above the
hymen, and is usua lly longer than it is wide. It can some­ A U-shaped, concave indentation on the edge of the
times be seen to extrude from the vaginal canal onto the hymen, cau sing a dip beneath the baseline or a break in the
hymen. Tags are commonly found in the midline (Fig, 2.8), continuity of the membrane, is usua lly referred to as a
and have been observed in 13 per cent of newborns,57 notch, while a cleft is V-shaped and angular. The concav­
Although most resolve in the first few months of life, many ity in the hymenal tissue is persistent, independent of the
persist for years. 48 - 50,56 examination technique and posi tion of the child. The
Bands can extend from the hymen to the vestibule, and extent of a notch/cleft can be classified as superficial or
can be found in all four quadrants. They have been partial, when its depth is less than or equal to one-half of
 34 I Investigation of suspected sexual abuse
Figure 2.9 Symmetrical perihymenal bands in a 4-year-old girl.
The semilunar shaped spaces created by the bands could be
mistaken for the hymenal opening.
the width of the hymenal membrane, or deep, when it is
more than one-half of the width. Notches or clefts are seen
commonly at the 12 o'clock position of annular or redun ­
dant hymens. They should not be reco rded as pathological
findings in the fimbriated hy men (because of its fringed
nature) or between 11 a nd 1 o'clock in a crescentic hymen ,
whe re the re is a natural absence of hymenal tissue. 56 The
sign ificance of a hymenal notch or cleft depends on the
location and ex tent of the defect. Anterior and latera l
notches, between the 9 and 3 o'clock positions, have been
observed in 35 per cent of newborn girls 57 are usually
caused by asymmetrical insertions of a crescentic hymen,
and should be considered normal. Anterior notches tend to
widen to form crescentic hymens, whereas some notches
disappear as hymena l tissue becomes less redundant. 49
Although the observation of a latera l notch in the first 3
years of life has been documented as a new finding, none
of the children followed up longitudinally developed a new
posterior notch. 50 Care must be taken to distinguish a true
concavity from an appa rent hollo wing in the hymenal rim
caused by an adjacent mound or tag. Until recently it was
thought that notches in the posterior portion of the hymen
were rarely observed in non-abused children. 53 ,5 7 With a
more consistent application of terminology, two groups of
examiners have established that superficial notches /partial
clefts can be found in the posterior hymen in 5-18 per cent
of girls who had been screened for non-abuse, and have
proposed that this is a non-specific finding. 48 ,56 A recent
study 59 suggested that some partial tears of the hymen heal
to leave shallow notches. Figure 2.3 demonstrates a notch
in the 5 o'clock posi tio n of the hymen, with friability of the
posterior fourchette.
A hymenal transection describes a complete deficiency
in hymenal tissue that ex tends to the junction between the
hymen and vestibule and, when seen in any location, is
associated with trauma. 56 ,60 Other terms used to describe
this a ngular concavity include complete notch/cleft, tear,
-- ~ --- - ~- - ­ - - ----- - ---
Figure 2.10 Transection of the hymen in the 5 o'clock position.
The hymen also shows ea rly oestrogen effects, appearing
thickened and redundant.
healed laceration, rupture and gap. It is widely agreed that
a deep/complete cleft or transection located between the
4 o'clock and 8 o 'clock positions in the hy men is consistent
with abuse (Fig. 2.10), as these have never been reported in
non-abused children .47 ,48,56
Gaping/Narrowing/Attenuation of the Hymen
The hymenal orifice has been described as gaping w hen,
with the child in the supine frog-legged position, a view is
obtained into the vagina, with thigh abduction and without
the exam iner manipulating the labia majora (Hobbs, Wynne
and Thomas, 1995).61 This subjective term is best avoided. A
narrow ing of hymenal tissue is referred to as attenuation of
the hymen. The narrowed rim is usually found at the poster­
ior edge and should be confirmed with the child in the
knee-chest position. Although usually associated wit h
chronic abuse, this term is comparative and implies that the
examiner has prior knowledge of the hy menal anatomy.
Unless the child has been examined previously it is not pos­
sible to determine w hether the hy menal rim is narrowed.
Heger et al 48 defined a narrow hymenal membrane as the
w id th of th e membra ne viewed in the coronal plane (i.e.
from the edge of the hymen to the muscular portion of the
vaginal introitus) of less th an 2 mm . A 'thickened edge' was
a term used to describe the relative amount of tissue
between the internal and external surfaces of the hymenal
membrane. Using these terms, the hymenal rim was found
to be thickened in 45.5 per cent; irregular in 51.7 per cent
and narrowed in 22.4 per cent of 147 prepubertal girls
screened for non-abuse. Over three-quarters (78 .7 per cent)
of girls with a narrowed hym enal rim were found to be
overweight (> 75th centile for weight).
The posterior hymenal width cannot be measured accu­
rately, especially when the rim is narrow. However, a com­
plete absence of the posterior hymen (confirmed in the
----- ---------
 Normal fem ale genital anatomy I 35

knee-chest position) has not been reported in new borns or

in prepubertal girls selected for non-abuseY

Hymenal Measurements

Owing to the absence of a single diagnosti c test for ch ild

sexual abuse, physicians have attempted to establish sta nd­
ards of dia gnosis by develo ping objective criteri a, such as
the measurement of the hyme nal open ing. Much of the early
literature concentrated on small differences in the diameter
of th e hymenal opening. One criterion freq uently cited as
dia gnosti c of child sexual abuse was the presence of a
hymenal openin g in excess of 4 mm. 62,63 Later studies have
show n that the hymenal ap pea ran ce and diameter vary with
the age of the child, 52,53 as well as wi t h the position of the Figure 2.12 Same child as shown in Fi g. 2.11. Labial traction
child during examin ation and the examination technique. 54 reveals an enlarged hymenal opening, allowing a cle ar view of the
In a cohort of non-abused pre-pubertal children, about one­ vag inal wa lls.
third had a hymenal openi ng size of g reate r than 4mm.48
The kn ee-chest position and th e supin e labial trac ti on
method were sup erio r to the supine lab ial separation tech­
nique in opening the vaginal introi tus. The largest vertica l
transhymenal diameter was produced in the kn ee-chest
position, whereas the greatest transverse horizonta l diameter
was produced by the supine lab ial traction procedure. Other
fa cto rs that influence the hymenal diameter include th e
hymenal configuration, state of relaxatio n of the child an d
the skill and ex perience of the examiner. Examiners should
expect to vis ualize the intravaginal contents during labial
traction, especially when the child is relaxed. Measurements
obtained und er general an aesthesia will be greater tha n
those taken when the child is awake. The hymen is an elas­
tic and dy namic piece of tissue, wit h its ap pearance and size
changing du ring normal physiological activities su ch as Figure 2.13 Labial tra ction showing an annular hymen with
breathin g and coughing. The effects of examination tech ­ perih ymenal bands in a 2-year-old girl.
niqu e, as well as genera l anaesthesia, on the configuration
and size of th e hymenal orifice are shown in Figs 2.11-2.14.

~
Figure 2.14 Same ch ild as in Fig. 2.13, examined under general
Figure 2.11 Labial separation shows a flesh y annular hymen in anaesthesia, using the same technique. Note the dilated
a 5-year-old girl. appearance of th e hymen .
 F'

36 I Investigation of suspected sexual abuse

With t he onset of puberty, hymenal elasti city increases

to such a n extent that penile penetration can occur without
any sig ns of trauma. Meas urement of the hymenal dia meter
then becomes an unreliable a nd useless exercise. Hymenal
measurements in non-abused children have provided data
to refin e a normal database against which to comp are the
anatomy of suspected child abuse victims. However, a nar­
row focus on the hymenal diameter as a criterion for sexua l
abuse has the potential to do more harm t han good. 54 - 55
There is subst antial overlap of mean transverse hymenal
diameters between prepubertal sex ually abused girls and
t hose selected for non-abuse. The hymenal diameter is
non-discriminatory for sex ual abuseY
Figure 2.15 A visi bly relaxed anus, with venous pooling seen at
the end of an examination in the supine knee-chest position .
PERIANAL FINDINGS

The skin of the anal verge and epithelium of the anal canal
exhibit regular folds when the anus is closed. Midline anterior
or posterior smooth areas in the anal verge are common find ­
ings, where the defe ct is lined by unbroken skin and repre­
sents a gap in the fibres of the superficial part of the external
anal sphincter. The significance of lateral defects in the anal
verge is unclear. The perianal skin may look red and moist,
due to poor hygiene, irritation, inflammation, infections or
infestations. Caution must be exercised when interpreting
perianal hyperpigmentation in non-white children. The
plexus of veins underlying the peri anal skin can be seen to
distend if examination is prolonged, producing a purple-blue
discolouration around the anus (venous poolin g). This is a
normal finding, caused when buttock traction and increased
intra-abdominal pressure occludes venous drainage (Fi g.
2. 15). Haemorrhoids in children are extremely rare.
Intermittent cont ractions of the external a nal sphincter Figure 2.16 Gross 'reflex anal dilatation', allowing a clear view
can sometimes be seen (twitching or winking) , a sign which of the anal canal. The child was examined in the supine
has been observed in normal children. This must be differen­ knee-chest position.
tiated from reflex anal dilatation, when the anus opens
widely after a brief period of buttock separation. Reflex a nal
dilatation is an incompletely understood phenomenon that
must be interpreted with caution a nd in contex t. It is best
demonstrated by the buttock separation test in the left lateral
position, when the buttocks are parted gently and observa­
tion takes place for 30 seconds. A positive result is obtained
when the ex ternal anal sphincter contracts briefly, follow ed
by dramatic rel axation of both the external and intern al
sphincters to reveal a cylindrical opening into the rectum
(Figs 2.16-2.18). Although first described as indicative of
anal abuse,57 the significance of this sign rema ins controver­
si al. Dilatation of up to 10 mm has been reported in up to
49 per cent of 267 normal children. 58 Other authors have
reported a prevalence of the sign in 4-14 per cent of non­
abused children. 59 ,7o Reflex anal dilatation is observed more
freq uently when the exam ination is prolonged, in the
knee-chest position, and in children who have a history of
constipation. 7I ,72 Chronic straining at stool may lead to Figure 2.17 Same child as in Fig. 2.16. Appearances following
weakness of the anal sphincter, and a 'visibly relaxed ' anus anal dilatation, showing perianal erythema and loss of anal folds.

..-:- ----- - - -~ - - ------- ~- - - - -----­


Figure 2.18 Same child as in Figs 2.16 and 2.17. Follow-up
examination 3 weeks later. Scar tissue is seen in the 12 o'clock
position.
(Fig. 2.15) has been described when stool may be present in
the rectum. Other situations when anal dilatation is observed
include the period following defecation, neurogenic bowel
disorders, use of musc]e relaxants during assisted ventila­
tion, general anaesthesia and during post-mortem examina­
tions. Under these circumstances, the sphincters are relaxed
and the anal dilatation is passive. They should not be misdi­
agnosed as reflex anal dilatation, which is a dynamic
response. Reflex anal dilatation remains a controversial,
highly variable and non-specific physical sign.
Early reports on children who had been anally abused
described erythema, swelling, venous engorgement, skin
tags, wedge-shaped smooth areas and anal dilatation. 67 .?3
Many of the physical signs, however, are minor and non­
specific but can be used to corroborate the child's history.
They should not be used in isolation, as they have also been
reported in cohorts of children who have not reported abuse
(Table 2.3). In a study of 267 prepubertal children selected
for non-abuse, McCann et al 68 encountered a relatively high
incidence of perianal soft tissue changes, and urged exam­
iners to exercise caution when rendering an opinion on the
Table 2.3 Perianal findings in non-abused children
Author Number of subjects Age
McCann etal1989 68 161 girls, 106 boys 2 months-11 yea rs
Berenson et al 1993 74 89 girls Mean 10.7 months
-
- - -- =-
Acute, healing and healed anogenital trauma I 37
significance of medical findings; venous engorgement was
observed in 52 per cent of the children, after 2 minutes in the
knee-chest position; some dilatation of the anus was
observed in 49 per cent of the children. The frequency of
these findings were confirmed in a cohort of 89 children of
less than 18 months 01d. 74 Findings infrequently noted in the
absence of abuse included fissures (in the absence of a recent
history of constipation), skin tags outside the mid-line, and
irregularity of the anal orifice without dilatation. Anal/
perianal bruising, anal lacerations, tears and scars have not
been observed in the absence of abuse. 47
ACUTE, HEALING AND HEALED
ANOGENITAL TRAUMA
Controversy remains about the frequency of abnormal find­
ings in children examined for sexual abuse. Different study
populations and the accessibility of clinics to the children
who disclose abuse when the injuries can be documented
can explain some of the differences. Public and professional
awareness of child sexual abuse will lead to earlier disclos­
ures in children and therefore result in more abnormal
findings. The conduct and timing of the medical examin­
ation, interest, experience and expertise of personnel
involved also differ between centres, and depend on the
presenting complaint. Children and adolescents may pre­
sent with a variety of symptoms and signs to health profes­
sionals at busy emergency departments or outpatient
clinics, where the diagnosis of sexual abuse is not sus­
pected. Disclosures may be made to staff at educational
establishments, social services or law enforcement agencies.
When appropriate child protection procedures are followed,
the medical examination is planned and forms part of the
investigative process. However, the medical examination on
its own must not be relied upon to diagnose sexual abuse.
With the introduction of standard terminology and
photo-documentation, practice and research have become
more reliable. The rates of 'abnormal findings' have
dropped as previously reported abnormal findings have
Summary of findings
Erythema 41 0/0
Hyperpigmentation 30%
Venous engorgement after 2 minutes of examination 52%
Midline wedge-shaped smooth areas 26%
Anal skin tags/folds 11%
Anal dilatation 49%
Smooth areas 26%
Hyper/hypopigmentation 10%
Erythema 7%
Skin tags 3%
Venous pooling 1%
--------~~
-- ~
 38 I Investigation of suspected sexual abuse
been recognized as non-specific findings, present also in
control subjects who have been carefully selected for non­
abuse. It is also important to remember that some forms of
child sexual abuse will leave no physical signs. Other fac­
tors that influence physical findings include the position of
the child in relation to the abuser, the size of the object
inserted relative to the size of the orifice, the degree of
force used, the use of lubricant, the frequency and duration
of abuse, as well as the period of time that has elapsed
between the last episode of abuse and the examination.
Erythema and superficial abrasions may disappear within
2 days, and bruises fade after a few days. A single episode
of penetrative anal abuse may result in perianal svvelling
and a gaping anus, signs that will resolve within 1 week.
There is now good documentation of norm a l variants in
anoge nital anatomy, but literature on the healing process
of the traumatized anogenital region is scant. In general,
hea ling is rapid and occurs through rege neration and
repa ir. Wound healing by regeneration is complete within
48-72 hours, and normal appearances are restored by 6
weeks. Deeper injuries heal by repair, a process involving
granulation and scarring that takes approximately 2
months. Because of the frequent delay between the alleged
abuse, disclosure and medical examination, there is a need
to understand the heal ing process, how an injury might
affect the genital anatomy of the prepubertal girl and
whether signs of trauma persist after the onset of pubeliy.
Finkel 75 reported on the findings of seven children who
experienced acute genital and anal trauma, and were fol­
lowed up until their injuries were healed . All of the chil­
dren had acute signs of injury documented after the
incident, most of which were superficial lacerations. In
only one child, who had a deep lacerated penetrating
wound, was there any sca r formation that distorted the
"hymenal and perihymenal tissues. Obvious acute superfi­
cial trauma healed without residua, by a process of regen­
eration. Healing was rapid, and by the second examination
(3-13 days later) t here was little apparent sign of injury.
McCann et al76 followed the resolu tion of acute genital
injuries in three children who had been sexually assaulted.
Initial signs of injury included erythema, oedema, submu­
cosal haemorrhages, abrasions, lacerations and hy menal
transections. Acute tissue changes resolved quickly, with ery­
thema and oedema disappearing by 11 days, whereas submu­
cosal haemorrhages resolved within 27 days. The shaJ1l
jagged edges of the damaged hymen gradually smoothed out
as healing took place, and there was relatively little scarring
in the hym en or posterior fourchette. With the onset of
puberty, the hymenal changes in one gi rl were obscured by
hypertrophy of the hymen. This highlights the importance of
using techniques to enable a thorough examination of the
fimbriated, redundant hymen seen in pubertal girls.
The natural history of perianal injuries has also been
studied in four children. Acute signs seen included ery­
thema and perianal oedema, which caused the folds to
thicken or flatten . Venous congestion was present in three
- - ~ -- - - ----- - -- - - - - - - - - -----
children, and two children had slight anal dilatation. Va ri­
ous degrees of laceration were noted in all the children. By
the eighth day, acute signs of trauma were present in only
one child. Second-degree lacerations healed over 5 weeks,
leaving na rrow bands of scar tissue; however, deeper lacer­
ations took 12-14 months to hea l. Even deep injuries faded
into surrounding tissues with time, making detection of
periana l injury increasingly difficult. 77
More recently, the healing patterns of anogenital injuries
in 94 prepubertal children were described in a longitudinal
study over 10 years. 59 The authors confirmed that anogenital
trauma healed quickly, often without residua. Of the 171 acute
irUuries see n, diagnostic changes remained in only 25 (14.6
per cent). All trauma to the labia majora or minora healed
without residua. Injuries to the posterior fourchette and fossa
navicularis (24 out of 47) healed with non-specific residua
(labial fusion, vascular changes and scarring) . Hymena l tran­
sections did not heal unless repaired surgically, whereas 5 out
of 37 partial tears of the hymen healed with residual sha llow
notches. Only J out of 31 anal injuries healed with anatomic
changes (tag, scarring and hypeJ1ligmentation).
FEMALE GENITAL FINDINGS IN
SEXUAL ABUSE
Over the past decade a great deal of information has been
gathered by medical exam iners who see children who have
been sexually abused, and a wide range of findings have
been reported. It is well documented tha t ab normalities of
the genital tract associated with abuse may be minor and
non-specific, occurrin g a lso as a result of inflammation,
infection or poor hygiene. Although Emans et a l58 found
that a number of genital findings were significantly more
common in sexually ab used children than in asy mptomatic
girls, many of the findings were also present in girls who
were examined bec ause of symptoms relating to vaginitis,
vulvitis, bleeding or dysuria. Debate continues regarding
the frequency of ab normal findings in sexually abused
children because of difficulties in reaching an agreement
on what constitutes an abnormality and the lack of a true
go ld standard diagnostic test for sexual abuse. There may
be no physical findings in more than one-half of the chil­
dren examined for suspected sexual abuse. The absence of
physical signs neither confirms nor negates a diagnosis of
abuse. An earlier study by Muram 78 on 31 girls whose per­
petrators confessed abuse reported that the medical exam­
ination failed to detect any abnormalitY in 29 per cent of
the girls. In a case review of 236 children whose peJ1letra­
tor had received conviction for sexual abuse, Adams et al 79
found that the genital examination was normal in 28 per
cent, non-sp ecific in 49 per cent, suspicious in 9 per cent
and abnormal in 14 per cent of cases. Abnormal anal find­
ings were found in only 1 per cent of patients. Using dis­
criminating ana lYSis, the two factors that significantly
correlated with the presence of abnormal genital findin gs
--- -- ­

in girls were (1) the time since the last incident and (2) a
history of bleeding being reported at the time of ab use.
Among the clinical findings that require diagnostic
interpretations are concave va ri ation s of the hymenal rim.
These have been referred to as clefts, notches, tears, fis­
sures, transections, healed laceratio ns or hymenal attenua­
tions. The cha racteristics and clinical correlates of concave
hymena l variations were described in a large cohort of
13 83 prepubertal and adolescent girls who alleged sexual
abuse. Concavities in the hymen were found in 174
patients (12 .6 per cent), of whi ch 3 5 per cent were anterior.
Of these, two -thirds were smooth and curved, wh ereas one­
third were angular and /or irregular. Posterior or lateral
concavi ties were found in another 57 per cent of the girls,
of whom one-third were curved and smooth, and two­
thirds were angular and/or irregu lar. Multiple hymenal
concavities were found in 13 girls. Each hymenal concav­
ity was categorized by location, contour (curved or angu­
lar) and smoothn ess (smooth or irregular). The concavities
were fUlther sorted into one of five categories, reflecting
the presence or absence of the clinical impression of
anogenital tra uma: benign, benign with reservations,
uncertain, suggestive of trauma , and trauma. Poste ri or
and/or lateral location of the concavity, angular contour
and rim irregu larity of the concavities we re found to be
significantly associated with trauma .60 Abnormal genital
findings are more likely to be documen ted when the girls
report bleeding at the time of the assault, or when the
examina tion occurred within 72 hours of the last episod e
of abuse. so The vulvovaginal signs of ab use have been
summarized by a working party of the Royal College of
Physicians of London,12 with emphasis on the type of
abuse and t he timing of the examination (Table 2.4). The
Table 2.4 Vulvovaginal signs of abuse'
Non-specific signs when seen within 72 hours of abuse
Erythema
Bruising
Superficial laceration/a brasion
Oedema
Signs supportive of abuse
Notch/cleft in posterio r hymena l rim, which may be associated
with scarring
Scar in posterior fourchette
Hym enal opening> 15 mm due to attenuat ion of hymen
Signs diagnostic of penetrating injury
Fresh laceration of hymen
Old tear in hymen that may have healed with scarring and
interruption of hymenal rim
Attenuation of hymen/disappearance of hymenal rim, usually
posteriorly
'Adapted from ref. 46.
Signs of anal abuse I 39
1997 publication has been replaced by an evidence-based
review of physica l signs of child sex ual ab use.47
These signs are currently subject to further scrutiny in
the light of more recent studies that suggest that the geni­
tal examination of the abused child rarely differs from that
of the non-abused child. Using a case-control study
design, 192 prepubertal children who alleged penetratio n
were examined using a colposcope, and the photographic
records we re compa red with 200 control subj ects who were
matched for age and ethnicity, and who were found not to
have been abused after careful screenin g. Abnormal find­
ings (hymen al transection, perforation or deep notch) were
seen in only 2.5 per cent of subjects or four children, aLi of
whom gave a history of penile or digita l penetration. How­
ever, only 17 ab used children were examined within 1
week of the abuse.56 A prospective study of 2384 children
referred to a tertiary centre from 1985 to 1990 for possible
sexual abuse found that 96.3 per cent of all children had a
normal medical exa min ation. Even with a history of vaginal
or anal penetration, the rate of abnormal medical signs was
only 5.5 per cent, seen in 61 out of 444 (6 per cent) girls and
2 out of 251 (1 per cent) boys. Most children in this study
were examined within 1 week of abuse. s1
In a study of female sexua l assault victims , 94 per cent
of 200 women were found to have trauma at one or more
locations - posterior fourchette, labia minora, hymen and
fo ssa navicularis. Injuries were found to vary by site, with
tears appearing most often on the posterior fourc hette and
fossa navicularis, abrasions on the labia, and ecchymosis
on the hymen. s2 However, the mean age of the cohort was
24 years (range 11-85 years) and the majority were exam­
ined within 24 hours of the assa ult, factors that a re atyp­
ical of children who disclose sexual abuse. Eighty-four
(88 per cent) women returned for follow-up examinations
within an average of 25 days (range 4- 50 days). All
injuries had resolved , with no evidence of t rauma in 71 (87
per cent) women. Tears to the hymen and fossa navicularis
did not reunite, and no scarring was observed at follow-up.
Kellogg et al s3 conduc ted a retrospective case review of
36 pregnant ado lescent gi rls (age range 12.3- 17 .8 yea rs)
who presented for sexual abuse exa minations. Despite
definitive evidence of sexua l contact (pregnancy), evidence
of penetrative trauma (clefts in the posterior, one-half of
which extended to the base of the hymen) was found in
only 2 of the 36 girls, whereas four had suggestive findings
(deep notches in the poste rior half that did not extend to
the base of the hymen, or visible scars) .
SIGNS OF ANAL ABUSE
Non-specific signs that may be seen in the acute stages
include erythema, bruising and oedema, which vary in
exte nt depending on the degree of friction and force
involved. Bruising may result in a haematoma formation at
the anal verge, caused by sheari ng of the subcutaneous
 40 I Investigation of suspected sexual abuse
venous plexus. Fissures or tears may be single or multiple,
seen as superficial or deep clefts in the perianal skin or
mucosa, which, in the acute situation, are painful. The fis­
sure is usually triangular or fan-shaped, with its apex point­
ing towards the anal canal and the lower edges pulled apart
by the fibres of the external sphincter. There may be bl eed­
ing or anal spasm associated with the fissure, or on its own.
Deep fissures are likely to heal with scarring and there may
be an associated skin tag caused by skin overgrowth during
healing. Opinions differ as to the significance of anal fissures
in relation to abuse, and the presence of a fissure gives no
information on its aetiology. Fissures resulting from consti­
pation cannot be distinguished from those caused by anal
abuse. They have been reported in normal cohorts of infants
and children,68,74 and also in children with constipation .lI ,72
Chronic trauma may lead to thickening of the anal canal
folds. There may be squamous metaplasia of the columnar
epithelium. The perianal tissues look and feel calloused,
and hypertrophy of the external sp hincter results in a
prominent ring of tissue surrounding the anal orifice. This
appeara nce has been referred to as the 'tyre sig n',67 a nd is
illustrated in Fig. 2.19.
An isolated finding of reflex anal dilatation is of no
diagnostic significance unless it is > 15 mm in size and
reproducible. When seen in association with a history of
anal abuse, especially in the presence of other signs such as
fissures, it supports the diagnosis. In the abse nce of a dis­
closure from the child, follow-up is justifi ed. A summary of
the perianal signs of abuse is shown in Tab le 2. 5. These
signs, from the 1997 publication by the Royal College of
Physicians, have been replaced by an evidence-based
review of physical signs of child sexual abuse .47
CONDITIONS THAT IVIIIVIIC ABUSE
General paediatlicians are often asked to see children because
a physical finding may raise concerns of sexual abuse. A lack
Figure 2.1 9 'Tyre' sign, showing thickened anal folds with
venous congestion.
-.:,,1 --~- --~- --- - ---- --~ -~ ----------­
Table 2. 5 Perianal signs af abuse'
Non-spe cific acute changes
Erythema
Oedema
Fi ssures
Venou s co ngestion
Bru isi ng
Signs supportive of abuse
Anal laxity without other explanation
Reproducib le reflex anal dilatation> 15 mm
Chronic changes, i.e., thickening of anal skin verge, increased
elasticity and reduced anal sphincter tone
Bite marks
Signs diagnostic of blunt force penetrating trauma
Fresh lace ration
Healed sca r extending beyond anal margin on to perianal sk in in
the absence of a reasonable alternative explanation
-Ada pted from ref. 46.
of fa miliarity with the normal female genital anatomy, as
well as the manifestation of infection, common childhood
conditions and accidental trauma, will cause physicians to
misdiagnose sexual abuse. In the atopic individual the vulval
skin can be moist and inflamed, as a result of contact derma­
titis from use of cosmetics or washing material (bubble
bath, biological soap powders). Vulvovaginitis is the most
common gynaecological complaint in childhood, and major
causes include poor hygiene, threadworms and infec tion with
Candida or group A beta-haemolytic streptococcus.84­86
Soreness and itch lead to scratching or rubbing, and bleeding
may be a presenting complaint.
A prolonged history of 'vulvovaginitis' may be a pre­
sentation of lichen sclerosus, often misd iagnosed as recur­
rent thrush. This is a disease of unknown aetio logy and
with an unpredictable course. Affected areas include the
external genitalia and perianal skin. In boys, th e prepuce
and glans penis may be more variably involved. The child
may present with bleeding associated with pain and itch,
and on exa mination the perineal skin is thin and friable,
with white shiny macules. There may be vascula r or pur­
puric areas, sup erfici al abrasions, haemorrhagic bullae,
erosion and ulceration . Usua lly the affected skin is sharply
demarcated from the surrounding normal skin, rarely
extendin g beyo nd a figure-of-eight distribution around the
labia and anus. Lichen scl erosis has been mistaken for sex­
ual abuse 87-89 Submucosa l haemorrhages in lichen sclero­
sus are seen in Fig. 2.20.
Physicians in the children's emerge ncy department are
sometimes asked to see children who present with perineal
injuries, and the question of sexual abuse may arise. Unin­
tentional injuries to the perineum in children are usually
accompanied by a witnessed account of the event and are
usually superficial. Hymenal injuries are rarely the result of

Figure 2.20 Lichen sclerosis. Cutaneous and submucosal
haemorrhages seen in the posterior fourchette and vestibule. The
patient was a 6-year-old girl who presented with pain and
bleeding 'down below:
accidental injury, when the labium minomm is the most
frequent structure involved. Deeper injuries tend to be
anterior or lateral to the hymen. In a multicentre study of
56 prepubertal girls who sustained unintentional perineal
injuries, the hymen was involved in only one child who fell
in a park with her legs abducted. A pinpoint abraded area
was found on the hymen at the 3 o'clock position. 9o
Congenital abnormalities such as haemangiomas and
failure of fusion of the median raphe raise concerns of
abuse when they first present. The raphe is a flesh-coloured.
slightly raised, linear structure in the perineum, and when
failure of fusion occurs the opposing edges may split and
appear like a traumatic lesion, with bleeding. Midline avas­
cular areas in the posterior part of the vestibule or posterior
fourchette may be seen as a white line (linea vestibularis) or
a white spot (partial linea vestibularis), and is easily mis­
taken for scar tissue. They have been observed in 25 per
cent of newborn girls,91 and can vary in size as well as con­
figuration in the first year of life. 92 In contrast with scar tis­
sue the linea vestibularis is generally in the midline, without
accompanying disturbance in vascularization. Normal vari­
ants such as bumps and notches between 3 and 9 o'clock on
the hymen can be mistaken as residua of sexual abuse.
Urethral prolapse is a circular eversion of the mucosa that
usually occurs at the urethral meatus without accompanying
symptoms. Some bleeding may occur following straining,
and assumed by carers to have come from the vagina. 9 ] Care­
ful examination reveals an oedematous area anterior to the
hymen, through which the urethral opening may be identi­
fied (Fig. 2.21). Crohn's disease can involve any part of the
alimentary system, and one presentation consists of extra­
intestinal manifestations. There may be perianal and vulval
ulceration with oedema, suggesting trauma. 94 ,9 5 Clinicians
may also have to sort through histories given by adults
who have misinterpreted nomlal childhood masturbation or
Screening for sexually transmitted infections I 41
~
Figure 2.21 A 10-year-old girl with a 4-year history of 'vaginal
bleeding'. An oedematous and erythematous 'polyp' is seen
anterior to the hymenal opening. The appearances suggest a
prolapse of the urethra.
sexual play between children, and by parents engaged in
custody disputes. The normal, age-appropriate sexual
exploitative behaviour must be differentiated from disturbed
behaviour arising from child abuse.96.97
SCREENING FOR SEXUALLY TRANSMITIED
INFECTIONS
Debate continues as to whether the screening for STls
should be routine in the evaluation of sexual abuse. The
presence of a sexually acquired organism can indicate prior
sexual contact in a child, and when accompanied by other
indicators of sexual abuse supports the diagnosis. Non-sex­
ual transmission of STls is rarely an issue in adults,
but when the same diseases are found in children, there is a
tendency to attribute them to an asexual mode of transmis­
sion. 98 The immature anogenital tract is more vulnerable to
infection, especially if there is a breach of the mucosal lin­
ing following traumatic abuse. The low numbers of children
reported to have acquired STIs from sexual abuse may rep­
resent the lack of systematic screening, or the non-recogni­
tion by clinicians that children can be infected through
sexual abuse. However, other routes of transmission that
must be considered, although difficult to exclude are:
• perinatal acquisition from an infected mother who may
be asymptomatic (Chlamydia trachoma tis, Neisseria
gonorrhoeae, Trichomonas vagina/is, herpes simplex
vims, human papillomavirus [HPV], human
immunodeficiency virus [HNJ);
• non-sexual adult to child contact (HPV, herpes simplex
vims);
• auto-inoculation (HPV, herpes simplex virus).
The risk of a child or young person acquiring an STI
depends on the prevalence of STis within the local population:
maternal STI during pregnancy that might lead to vertical

 42 I Investigation of suspected sexual abuse
transmission; the type of sexual contact during abuse ; injuries
to the genital tract; the sexual maturity of the victim; and
whether a condom was used during abuse. Transmission from
mother to child is welJ documented for most STIs and the
organism can lie dormant for up to 2 years (and possibly
3 years). In dealing with a child who is less than 3 years old, it
is therefore important to establish if the parents are infected
and to screen them if possible. Sexual abuse should be sus­
pected when an infection is diagnosed after infancy and
before sexual activity occurs in the older child. Under these
circumstances, gonorrhoea and syphilis would be diagnostic
of sexual abuse, whereas infection with Chlamydia, Tri­
chomonas, herpes simplex virus and HPV would be suggestive
of sexual abuse. 12,44,99,100
A prospective study of 1538 children who were examined
for possible sexual abuse found the overall prevalence of
STIs to be 6 per cent. The diagnosis of a STI was highly cor­
related with a history of sex ual contact (alleged by 49 per
cent of the children) and the presence of a discharge. l01 The
authors had attempted to exclude perinatal acquisition as a
source of infection by selecting verbal children who could
understand questions regarding sexual contact. Although
the prevalence of STIs was only 3.7 per cent in a cohort of
159 girls who had been sexuaJly abused, Robinson et al 102
also found a significant association of sexually transmitted
organisms with the presence of a vaginal discharge. The
authors recommended that the presence of vaginal discharge
in sexuaJly abused girls was an indication for STI screening.
The use of an algorithm to assess the risk of gonococcal
and chlamydial infections was found to reduce the cost and
trauma of unnecessary sampling. Major factors in the risk
assessment were genital-genital or genital-anal contact or
penetration, suspicious anogenital findings, genital dis­
charge and suspicion of an STI in the perpetrator, although
vaginitis and referral for suspected abuse in a sibling were
designated minor factors. Restricting testing to children
with at least one major or two minor factors in their study
population of 3040, 45 per cent would have been tested and
all known cases of gonorrhoea or chlamydia infection
would have been identified. Testing would have been
avoided in 51 per cent of girls and 72 per cent of boys. 103
More recently, the Association of Genitourinary Medicine
and the Medical Society for the Study of Venereal Disease in
the UK produced guidelines for the management of sus­
pected STls in children and young people. The recommen­
dations included the need for sta Ff worki ng in genitourinary
medicine to be: alert to the possibility of child abuse and
neglect ; aware of local guidelines; and trained in child pro­
tection procedures and protocols. Screening should be con­
sidered in all young people who may have been sexually
abused or who have been found to have an STI. 104
Human papillomavirus infection is one of the most
common STIs, and is estimated to affect 10 per cent of the
adult population. Subclinical infection is common, and the
increasing incidence of condyloma in children probably
reflects the increased prevalence of HPV disease in the
~ --- - '- - - - - ---~- -----­
adult population. Results from studies of children with
anogenital warts suggest that the majority of children do
not acquire these sexually. Vertical transmission is an
important source of infection in young children, and
household members of children with anogenital warts
should be routinely screened for the presence of non-geni­
tal as well as anogenital warts. 105.106 At present, there is no
reliable method to determine the mode of acquisition in a
child with HPV infection.
Children who disclose sexual abuse should be screened
for STIs under the following circumstances:
• presence of symptoms and signs at the time of assault
or subsequently (bleeding, discha rge, dysuria);
• suspicious anogenital findings;
• history of genital-genital or genital-anal contact or
penetration;
• alleged abuser known to have, or at high risk of
acquiring, STI;
• STI prevalent in the community;
• request from child or young person.
The forensic science laboratory will not examine samples
obtained for screening of STls. For each organism, sampling
techniques and transport media should be individually tai­
lored and must meet the criteria for maintaining the chain of
evidence. Swabs must be taken by a doctor with appropriate
training and experience. lt is therefore important to establish
good liaison with the diagnostic laboratories, as well as
physicians in genitourinary medicine. Fine, wire-mounted
urethral swabs can be passed through the hymen, even in
young children, to enable cultures for STls. 107 A vaginal wash
procedure, using stelile saline introduced through tubing
from a butterfly needle, has also been found to be acceptable
as a diagnostic test for STIs in prepubertal girls.!08 Testing for
infection with HlV involves a blood test for HlV antibody,
obtained at the time of the initial examination and at follow­
up. Newer, more sophisticated tests for HfV-RNA or proviral
DNA need only be done in a young child to exclude vertical
transmission. 109 Serological testing for infection with Tre­
ponema pallidum, hepatitis B or hepatitis C virus should be
considered on an individual basis. The child or young person
may require active and passive immunization against hepati­
tis B infection as well as antiretroviral therapy. Expert opin­
ion should be sought from an infectious diseases specialist.
In general, there have been relatively few studies where
children with a particular STI have been evaluated for the
possibility of child sexual abuse. This has resulted in a lim­
ited evidence base to determine whether a particular STI is
a marker for sexual abuse 47
INTERPRETATION OF CLINICAL AND
LABORATORY FINDINGS
Many changes have occurred in the medica l evaluation of
children suspected of having been sexually abused, and

how physical findings are interpreted. A comprehensive list
of the clinical and laboratory findings seen in abused and
non-abused children was first drawn up by Adams et al. 43
Sometimes known as the Adams Classification System, it
has been developed over the years and the latest revision 45
is the result of examining published data and consensus
amongst 18 experienced physicians. The classification sys­
tem can be summarized as:
• findings documented in newborns, or commonly seen
in non-abused children:
- normal variants;
- findings commonly caused by other medical
conditions.
• indeterminate findings (owing to insufficient or
conflicting data):
- physical examination fmdings, for example deep
notches or clefts in the hymen; smooth,
uninterrupted rim of hymen of < 1 mm wide;
immediate dilatation of the anus> 2 cm;
- Lesions with indeterminate specificity for sexual
transmission: for example, anogenital condyloma
accuminata, anogenital herpes.
• fi ndings diagnostic of trauma and/or sexual contact:
- acute trauma to external genital/anal tissues, for
example fresh laceration or extensive bruising;
- residual (healing) injuries, for example perianal
scar;
- injuries indicative of blunt force penetrating
trauma: for example acute laceration, extensive
bruising, healed hymenal transection;
- presence of infection that confirms sexual
deliberate omission of pertinent facts or knowledge and
contact, for exam pIe gonorrhoea;
misquoting of medical literature H5
- diagnostic of sexual contact: for example,
pregnancy or sperm identified in specimens taken
directly from the child's body.
Until a better evidence base is available, this system
provides a useful tool that both assists physicians in inter­
preting clinical findings and helps to achieve some consist­
ency in terminology.
SUMMARY
Whenever abuse of a child is suspected the clinician must
refer to local multi-agency child protection procedures for
appropriate action. Siblings and close friends of the victim
must also be interviewed and examined if appropriate, as it
is likely that they may also have been exposed to the
alleged perpetrator. 110 There may be an association between
different types of abuse, and following a diagnosis of any
form of abuse a medical assessment for sexual abuse
should also be considered.
In the investigation of sexual abuse of children, clinicians
should remember that the medical examination is only a part
of the jigsaw. Physical evidence is neither essential for, nor
Summary I 43
predictive of, conviction. The child's history is the single
most important factor in the accurate diagnosis of most
cases of sexual abuse. III ,11 2 In a review of child sexual abuse
criminal court cases, those involving the youngest victims
were found to have a significantly lower conviction rate.
Successful prosecution, particularly in the youngest chil­
dren, depended on the quality of the verbal evidence and the
effectiveness of the child's testimony. II ) The literature
reports a marked variability of medical findings in children
examined for sexual abuse and the diagnosis of sexual abuse
cannot be made on the basis of physical findings alone. It
has been suggested that a model could be developed for the
prediction of anatomic findings, based on an inventory of
variables that would include the age of the child, type of
contact, whether penetration occurred, and a history of pain
or b1eeding.1l4
Expert medical testimony may help to interpret the
presence or absence of physical signs, but the overall effect
on the legal outcome is unknown. Increased communica­
tion between prosecutors and paediatricians may improve
the outcome, especially if the paediatrician is told whether
she is a witness to fact, or an expert witness. The distinc­
tion is important to allow adequate preparation and to
avoid an injustice to the child. A medical expelt should be
able to demonstrate relevant training or experience in child
abuse cases that are similar to ones in which he or she has
been called on to provide expert testimony. Irresponsible
medical testimony must be avoided; this includes use of
unique theories of causation, unusual interpretations of
medical findings, alleging non-existent physical signs,
Medical , social and legal professionals have relied too
heavily on the medical examination in diagnosing sexual
abuse. Normal findings are consistent with abuse, and the
examiner must document all signs, positive or negative,
whenever any child is examined. There is now a consensus
on terminology and interpretation of findings in child sex­
ual abuse, 12,44,45,47,48,56 but these guidelines are only helpful
if examiners are meticulous in documenting their findings.
There needs to be an agreement on the minimum standards
for the training of medical professionals who perform sex­
ual abuse examinations. 9 ,37,116,117 Medical care for children
and young people who disclose abuse has evolved into a
complex and technical specialty. Expert medical testimony
can be crucial to the legal outcome of a case of alleged sex­
ual abuse. The evaluation of child sexual abuse is not rou­
tinely taught in undergraduate or postgraduate training
programmes. Specialized training programmes must be
developed to ensure better research as well as the dissemin­
ation of knowledge and expertise.37, 11 8 In September 2006,
the American Academy of Pediatrics recognized child abuse
as a new pediatric subspecialty. It is hoped that the accred­
itation of trained specialists will benefit child victims of
sexual abuse, families, health-care professionals and those
in social services and law enforcement.
 44 I Investigation of s us pected sexual abu s e
REFERENCES
Sch echter IVID, Roberge L. Sexual ex ploitatio n. In Hel fe r RE,
Kempe CH (ed s) Child A buse and Neglect: Th e Family and th e
Community. Ca mbridge, MA: America n Academy o f
Pedia tri cs, 1978.
2 Ja FFe AC, Dynn eson L, ten Bensel RW. Sex ual a bu se of
children: a n epi demiological study. Am J Dis Child 197 5;
129 :689 - 92.
3 Kemp e CH. Sexu al a buse, a nother hidd en pedi a tri c probl em.
Pedia trics 1978; 62 :382-9 .
4 Finkelhor D. Current in formation on the scope and na tu re of
child sexual abuse. The Future ofCIlildren J994; 4:3 1-5 3.
5 Jones LM, Finkelhor D, Ko pie c K. Why is sexual a bu se
declin ing? A survey o f state chil d pro tecti on a dministra tors.
Child A buse Et Neg lect 2001 ; 25 : 11 39- 58.
6 Dunne MP, Purdi e DM, Cook MD et al. Is child sex ua l abuse
declining? Evid ence Fro m a popul a tio n-based survey o f men
and wo me n in Australi a. Child A buse & Neg lect 2003 ;
27:141- 52.
7 Hobbs CJ , Wy nne JM. Child sexual abu se: an in creasing ra te
of di ag nos is. Lancet 1987; ii: 837- 4 5.
8 Summit RC. Th e child sexual abuse acco mmodati on
sy ndrome. Child A buse & Neglect 1983; 7 : 177-93.
9 Mo k JYQ, Busuttil A, Hamm ond HF. The j o in t paedia tric­
fo rens ic exa min ati on in child a buse. Child A buse ReIJ 1998 ;
7 :194-203.
10 Levitt CJ. Sex ua l abu se in children : A compassiona te ye t
thorough approach to eva lua tion. Postg rad IVIed 198 6 ;
80 :201-15.
11 De San Lazaro C. Ma king paedi atric assessment in susp ected
sex u al abus e a therap euti c ex perience. Arch Dis Childh 199 5;
73 : 174- 6.
12 Th e Royal CoJl ege o f Pa edia trics and Child Hea lth and Fac ul ty
of Foren sic a nd Legal Medi ci ne. Guida nce all pa ediatri c
f orensic examina tions in relation to poss ib le child sexu.al
abuse. September 2004. www.rcpch .ac. uk/doc.aspx? id_
Resource=3 409
13 Roya l College of Paediatri cs and Child Hea lth . Child protection
companion. Lond on, RCPCH.
J4 Kendall-Tacke tt KA, Meye r Williams L, Finkelhor D. Impa ct of
sexu al abuse on children: A review and synthesis of rec ent
empirical st udi es. Psychol Bull 1993; )] 3 : 164 -8 0.
15 Be rliner L, Elli ott DM. Sexual a bu se of children. In Briere J,
Be rliner L, Bulkley C et al (eds) The APSAC handbook on child
maltreatment. Thousa nd Oaks, CA: Sage, pp. 51-71.
16 Lev itt CJ. The medica l ex amin a tio n in child sexua l abuse:
A ba lance betw ee n history a nd exam. J Child Sex Abuse
199 2; 1:11 3-21.
17 Myers JEB. The ro le of the physici an in presen tin g verbal
evid ence of child abuse. J Pediatr 19 86; 109 :409 -11.
18 Ferrell J. Foley cath eter balloon technique fo r visu ali sing th e
hym en in fe male ad olescent sexu al a buse vic tims. J f merg
Nurs 1995 ; 21 :585-6.
19 Starling SP, Jenny C. Fo ren s ic exa mina tion of adolescen t
fema le genitalia: th e Foley cathete r technique. A rch Pedia tr
Adolesc Med 1997 ; 151 : 10 2- 3.
20 Persa ud Dl, Squires JE, Rubin-R emer D. Use of Foley ca theter
to examin e estrogenised hy mens for evidence of sexual abu se.
J Pediatr Adolesc GYl1ecol 1997; 10: 83-5.
21 Lauber A, Souma G. Use o f toluidine blue fo r docum enta ti on
of trauma tic intercourse. Obs tet Gyneeol 1982 ; 60:64 4-8.
22 McC au ley J, Gorma n RL, Guzins ki G. Toluid in e blue in the
detecti on o f pe rineal la cera ti o ns in pedi a tri c a nd adolescen t
sex ual a buse victims. Pediatri cs 1986; 8 6: 1039-43.
23 Tei xei ra WR . Hy menal col poscopi c ex amina tion in sex ual
oFfences. A m J Forensic Mcd 19 81; 2 :209-14.
, -- - - ---- - - -- - - - --- ------
24 Woo dling BA, Heger A. The use of the colpos cope in the
diag nos is o f sex ua l a bu se in th e pedi a tri c age group. Child
A buse Et Neglect 1986 ; 10: J11 - 14.
25 Mura m D, Eli as S. Ch ild sex ual abus e: geni ta l tra ct findin gs in
prep ubertal girls . n. Comp arison of colp oscopi c and unaided
examinati ons. Am J Obste tr Gy J1 ecol 1989; 160 :3 33 - 5.
26 Ada ms JA, Phillips P, Ahm ad M. The usefulness o f
coJ pos copic pho to graph s in the evaluat io n o f suspected child
sexual abuse. Adolesc Pedia tr Gyneco /1 990; 3:75-8 2.
27 Mura m D, J ones CE oTh e use o f vid eo- colposcopy in the
gynecologic exa minati o n of infa nts, children, and yo ung
adolescents. Ado lesc Pediatr Gym'col 1993 ; 6: 154-6.
28 McCann Jl. The colposcop ic ge nita l exa min atio n o f the
sex ua lly abu sed prep ub erta l fe male. Adolesc Pedi atr GYl1 ecol
1993 ; 6:123 - 8.
29 Hobb s C, Wy nn e J. Use of the colpo scope in examin ation for
sex ual abuse. Arch iues of Diseas e in Childhood 1996 ;
75 :53 9- 42.
30 Heger AH. Twe nty years in the ev aluati on of the sex ually
a bused child : h as medi cin e helped or hUlt th e child a nd the
fa mily? Child Abuse Et Neglect 1996; 20:893 -7.
31 Will ott GM, Cross MA. Th e de tec tion o f sperm a tozo a in the
mo uth. J Fo rel/sic Sci Soc 1986 ; 26 :12 5-8.
32 Willot t GM, All ard JE. Sp erma tozo a - their persi sten ce a Fter
sex ual in terco urse. ForeJ1 sic Sci Int 1982; 19:13 3-5 4.
33 Davi es A, Wilson EM . Th e persi sten ce of se minal constitu ents
in the hum a n vagina. Forensic Sc ience 1974; 3 :45-55.
34 Christ ian CW, Lavelle JM, De Jong AR et al. Fore nsi c ev id ence
fi ndi ngs in prepub erta l v ictims of sex ual assault. Ped iatrics
2000 ; 106 : 100-4.
35 Fin kel MA , Ricci LR. Docum enta ti on and prese rva ti o n o f
visua l evid ence in ch ild a buse. Ch ild Ma ltreatment 1997;
2 :322-30.
36 Finkel !VIA. Techni cal condu ct of the CSA medic al
exa min at ion. Child Abuse & Neglect 1998 ; 22 :555-66.
37 Bl ock RW, Palusci V1. Ch ild Abuse Pedia trics: A new pedi atri c
subs peci alty. J Pediatr 2006 ; 148:711- 12.
38 Brayden RM, Altemeier WA IJ], Yeager T, Muram D.
Interp reta tio ns o f coJposcopi c photogra ph s: Evidence fo r
com peten ce in assess ing sex ua l a buse. Child Abuse Et Neglect
1991 ;15:69-7 6.
39 Adam s J A, Well s R. Normal versus a bno rma l gen ital findings
in ch ildren: how well do exa min ers agree? Child Abuse &
Neglect 1993 ; 17:663-7 5.
40 Paradi se l E, Win ter MP, Finkel IVIA et al. Influ ence of hi story
on physici ans ' interp ret ation of girls ' genit al finding s.
Pediatrics J999 ; 10 3 :980- 6.
41 Bu tle r-Sloss E. Report of the inqui ry il1 to child abuse ill
Clellelan d 19 87. London: HMSO.
42 Adams JA. Gra nd round s : medica l iss ues and case s tudi es.
Sign ifican ce of medi cal findin gs in s uspec ted sex ual a buse:
mo vin g towards con sensus. J Child Sex Abuse 1992 ; 1 :9 1-9.
43 Ada ms J A, Ha rper K, Kn ud son S. A proposed syste m for the
cla ss ifi cati on of anogenital findings in children with
susp ected sex ual abus e. Adolesc Pediatr Gy neco l199 2;
5:73-5.
44 Ameri can Aca demy of Pediatrics Committee on Child Abuse
a nd Neg lect. Gonorrhea in prepu berta l children. Pedia trics
1998; 101: 13 4- 5.
45 Ada ms J A. App ro ach to the in terpreta ti on of medica l and
labora tory findings in sus pected child sex ual abu se: a 2005
revisio n. Th e APSAC Advisor 200 5; Summer:7-1 3.
46 Royal College of Physi Cia ns of Lond on. Physical Sig ns of
Se.rual A bus e in Child re n. Lond on: Lav enham Press , 1997.
47 Ro yal College o f Paediatrics a nd Child He alth. Th e physical
signs of child sexu al abuse. A n elJ ide nce-based relJ iew and
gu idance fo r best pra ctice. Lo nd on, RCPCH.
----- ----- -

48 Heger AH, Ticson L, Guerra L et al. Appearance of the
genitalia in girls selected for non-abuse: review of hymenal
morphology and non-specific findings. ] Pediatr Adolesc
Gynecol2002; 15:27-35.
49 Berenson AB. Appearance of the hymen at birth and one year
in infants 18 months of age and younger. Pediatrics 1993;
of age: A longitudinal study. Pediatrics 1993; 91:820-5.
91 :838-40.
50 Berenson AB. A longitudinal study of hymenal morphology in
75 Finkel MA. Anogenital trauma in sexually abused children.
the first 3 years of life. Pediatrics 1995; 95:490-6.
51 Yordan EE, Yordan RA. The hymen and Tanner staging of the
breast. Adolesc Pediair Gynecol1992; 5:76-9.
52 McCann J, Wells R, Simon M, Voris J. Genital findings in
prepubel1al girls selected for nonabuse: a descriptive study.
Pediatrics 1990; 86:428-39.
53 Berenson AB, Heger, AH, Hayes, JM et al. Appearance of the
hymen in prepubertal girls. Pediatrics 1992; 89:387-94.
54 McCann J, Voris J, Simon M, Wells R. Comparison of genital
examination techniques in prepubertal girls. Pediatrics 1990;
85:182-7.
55 Gardner JJ. Descriptive study of genital variation in healthy,
nonabused premenarchal girls.] Pediatr 1992; 12:251-7.
56 Berenson AB, Chacko MR, Wiemann CM et al. A case-control
study of anatomic changes resulting from sexual abuse. Am ]
Obstet Gynecol 2000; 182:820-34.
57 Berenson A, Heger, A, Andrews, S. Appearance of the hymen
in newborns. Pediatrics 1991; 87:458-65.
58 Emans SJ, Woods ER, Flagg NT, Freeman A. Genital findings
in sexually abused, symptomatic and asymptomatic girls.
Pediatrics 1987: 79 :778-85.
59 Heppenstall-Heger A, McConnell G, Ticson L et al. Healing
patterns in anogenital injuries: a longitudinal study of
injuries associated with sexual abuse, accidental injuries, or
genital surgery in the preadolescent child. Pediatrics 2003;
112:829-37.
60 Kerns DL, Ritter ML, Thomas RG. Concave hymenal variations
in suspected child sexual abuse victims. Pediatrics 1992;
90:265-72.
61 Hobbs CJ, Wynne JM, Thomas AJ. Colposcopic genital
findings in prepubertal girls assessed for sexual abuse. Arch
Dis Child 1995; 73:465-9.
62 Cantwell H. Vaginal inspection as it relates to child sexual
abuse in girls uncler thirteen. Ch ild Abuse Et Neglect 1983;
7:171-6.
63 White ST, Ingram DL, Lyna PRo Vaginal introital diameter in
the evaluation of sexual abuse. Child Abuse Et Neglect 1989;
13:217-24.
64 Kerns DL. Cool science for a hot topic. Child Abuse Et Neglect
1989; 13:177-8.
65 Paradise JE. Predictive accuracy and the diagnosis of sexual
abuse: a big issue about a little tissue. Child Abuse Et Neglect
1989; 13:169-76.
66 Heger A, Emans SJ. Introital diameter as the criterion for
sexual abuse. Pediatrics 1990; 85:222-3.
67 Hobbs CJ, Wynne JM. Buggery in childhood: a common
syndrome of child abuse. Lancet 1986; ii:792-6.
68 McCann J, Voris J, Simon M, Wells R. Perianal findings in
prepubeltal children se.iected for nonabuse: a descriptive
study. Child Abuse Et Neglect 1989; 13:179-93.
69 Priestley B. Reflex anal dilatation and abuse. Lancet 1987;
ii:1396.
70 Stanton A, Sunderland R. Prevalence of reflex anal dilatation
in 200 children. EM] 1989; 298:802-3.
71 Clayden G. Reflex anal dilatation associated with severe
chronic constipation in children. Arch Dis Childh 1988;
63 :832-6.
72 Agnarsson U, Warde C, McCarthy G, Evans N. Perianal
appearances in childhood constipation. Arch Dis Childh 1990;
65:1231-4.
References I 45
73 Hobbs CJ, Wynne JM. Sexual abuse of English boys and girls:
the importance of anal examination. Child Abuse Et Neglect
1989; 13:195-210.
74 Berenson AB, Somma-Garcia, A, Barnett, S. Perianal findings
Pediatrics 1989; 84:317-22.
76 McCann J, Voris J, Simon M. Genital injuries resulting from
sexual abuse: a longitudinal study. Pediatrics 1992;
89:307-17.
77 McCann J, Voris J. Perianal injuries resulting from sexual
abuse: a longitudinal study. Pediatrics 1993; 91:390-3.
78 Muram D. Child sexual abuse: Relationship between sexual
acts and genital findings. Child Abuse Et Neglect 1989;
13:211-16.
79 Adams J, Harper K, Knudson S, Revilla J. Examination
findings in legally confirmed child sexual abuse: it's normal
to be normal. Pediatrics 1994; 94:310-17.
80 Adams JA, Knudson, S. Genital findings in adolescent girls
referred for suspected sexual abuse. Arch Pediatr Adolesc Med
1996; 150:850-7.
81 Heger A, Ticson L, Valesquez 0, Bernier R. Children referred
for possible sexual abuse: medical findings in 2384 children.
Child Abuse Et Neglect 2002; 26:645-59.
82 Slaughter L, Brown CRV, Crowley S, Peck R. Patterns of
genital injuIY in female sexual assault victims. Am ] Obstetr
Gynecol 1997; 176: 609-16.
83 Kellogg ND, Menard SW, Santos. A. Genital anatomy in
pregnant adolescents: 'Normal' does not mean 'Nothing
happened'. Pediatrics 2004; 113 :e67-9.
84 Paradise JE, Campos JM, Friedman HM, Frishmuth G.
Vulvovaginitis in premenarchal girls: clinical features and
diagnostic evaluation. Pediatrics 1999; 70:193-8.
85 Straumanis JP, Bocchini JA. Group A beta-hemolytic
streptococcal vulvovaginitis in prepubertal girls: a case report
and review of the past twenty years. Pediatric Inject Dis]
1990; 9:845-8.
86 Pierce AM, Hart CA. Vulvovaginitis: causes and management.
Arch Dis Childh 1992; 67:509-12.
87 Jenny C, Kirby P, Fuquay D. Genital lichen sclerosus mistaken
for child sexual abuse. Pediatrics 1981; 83:597.
88 Handfield-Jones SE, Hinde FR, Kennedy CT. Lichen sclerosus
et atrophicus in children misdiagnosed as sexual abuse. EM]
1987; 294: 1404-5.
89 Muhlendahl KE. Suspected sexual abuse in a 1O-year-old girl.
Lancet 1996; 348:30.
90 Bond GR, Dowd MD, Landsman I, Rimsza M. Unintentional
perineal injury in prepubescent girls: A multicenter,
prospective report of 56 girls. Pediatrics 1995; 95:628-31.
91 Kellogg ND, Parra JM. Linea vestibularis: a previously
undescribed normal genital structure in female neonates.
Pediatrics 1991; 87 :926-9.
92 Kellogg ND, Parra JM. Linea vestibularis: follow-up of a
normal genital structure. Pediatrics 1993; 92:453-6.
93 Johnson CF. Prolapse of the urethra: Confusion of clinical and
anatomic characteristics with sexual abuse. Pediatrics 1991;
87 :722-5.
94 Wallis SM, Walker-Smith J. Case report: an unusual case of
Crohn's disease in a West Indian child. Acta Paediatr Scand
1976; 65:749-51.
95 Sellman SPB, Hupertz VF, Reece RM. Crohn's disease
presenting as suspected abuse. Pediairics 1996; 97:272-4.
96 Friedrich WN, Grambsch P, Broughton D et al. Normative
sexual behavior in children. Pediatrics 1991; 88:456-64.
97 Friedrich WN, Fisher J, Broughton D et al. Nonnative sexual
behavior in children: a contemporary sample. Pediatrics 101 :e9.

 46 I Investigation of suspected sexual abuse
98 Neinstein LS, Goldenring J, Carpenter S. Nonsexual
transmission of sexually transmitted infections: an infrequent
occurrence. Pediatrics 1984; 74 :2 17-25.
99 American Academy of Pediatrics. Sexually tra nsmitted
diseases. In Pickering LK (ed .) Rep ort oj th e Co mmittee on
InJectious Diseases, 26th ed n. Elk Grove Village, lL: Ballinger,
2003, pp. 157-67.
100 Am erica n Aca demy of Pediatrics Co mmi ttee on Child Abuse
and Neglect. The eva luation of sexua l abuse of children.
Pediatrics 2005; 11 6:506-12.
101 In gra m DL, Everet t D, Lyna PR et al. Epidemi ology of adult
sexually transmitted disease agents in children being
evaluated for sexual abuse. Pediatr Inject Dis J 1992 ;
11 :9 45-50.
102 Robinson AJ , Watkeys JEM, Ridgway GI. Sexually tra nsmitted
organisms in sexually abused ch ildren. Arch Dis Childh 199 8 ;
79: 356-8.
103 In gram DM, Miller WC, Schoenbach VJ et al. Risk assessment
for gono co cca l an d ch lamy dial infec tions in young children
underg oing evalua tion for sexua l ab use. Pediatrics 2001:
107 :e73-80.
104 Th o mas A, Forster G, Robinson A, Rogsta d K for the Clinical
Effectiveness Group (Association of Genitourinary Medicine
and the Medical Society for the Study o f Venereal Diseases).
Nation al guideline for th e management of suspected sexu ally
transmitted infections in children a nd yo ung people. Sex
TrailS Inject 2002; 78:324-31 .
105 Handl ey J, Dinsmore W, Maw R et aJ. Anogenital warts in
children; sexual abuse or not? Int J STl ft AIDS 1993 ; 4:271-9.
106 Si eg fri ed E, Rasnick-Conley J, Co ok S et al. Human
pap ill o mavirus screening in pediatric victims of sexual abuse.
Ped iatrics 1998 ; 101 :43- 7.
107 Steele AM , de Sa n Lazaro C. Transhymenal cultures for
sexually transmissible organisms. Arch Dis Childh 1994;
71 :423-7.
J. - - - - - -
108 Embree JE, Lindsay D, William s T et al. M. Acceptability a nd
use fulness of vaginal washes in premenarch eal girls as a
diagnostic procedure for sexua lly transmitted infections.
Pediatr Inject Dis J 1996; 15 :651-66.
109 Mok JYQ. Routine Hrv testing after child sexual abuse? Child
Abuse ReI! 1998; 7:6 3-9.
110 Muram D, Spe ck PM, Gold SS. Genital abnorma lities in
female siblings and friends of child victims of sexual abuse.
Child Abuse ft Neglect 1991 ; 15: 105-10.
III Bays J, Ch adwic k, D. Med ical diagnosis of the sexually
ab used child. Ch ild Abuse ft Neglect 199 3; 17:9 1-11 0.
112 Muram D. Child sex ual ab use. Curl' Opin Obstetr GYllecol
1993 ; 5:784-90.
11 3 De Jong AR, Rose M. Legal proof of child sex ual abuse in the
absence of physical evidence. Pediatrics 199 1; 88:506-11.
114 Kerns DL. Triage and referral s for child sexual a buse medical
examinations: Which children are likely to have positive
medical findings? Ch ild Abuse ft Neglect 1998 ; 22:515-18.
115 Chadwick DL, Krous HF. Irresp ons ibl e testimony by medical
ex perts in cases involv ing th e physica l abuse and neglect of
ch ildren. Child Maltreatment 1997; 2: 313 -21.
116 Adams JA. Th e role of photo do cu mentati on of genital
findings in medical evaluations of suspected child sexual
abuse. Child Maltreatment 1997; 2:341-7.
117 Mok JYQ, Busu ttil A. Medical exa minati ons for Child Sex ual
Abuse in Scotla nd: good enough practice? Child Abuse
Revi ew 2004; 2004; 13:324-37.
118 Jen ny C. Pediatric fellowships in child abuse and neglect: the
development o f a new subspecialty. Child Maltreatment 1997;
2:356-61.
119 Emans SJ, Woods ER, Allred EN, Grace E. Hymena l findings
in adolescent wo men: impact of tampon use an d consensual
sexual activity. J Pedlatr 199 4; 125: 153-60.
---- -- -­

RADIOLOGY OF CHILD ABUSE
Maeve McPhillips
Role of the radiologist 47
Radiological investigations 48
Skeletal injuries 51
Head injury 60
Visceral injuries 68
Child abuse has been a concem of paediatric radiologists since
Caffey's description of multiple fractures in children with a
subdural haematoma in 1946. He describes fractures of the
long bones as a 'common complication of infantile subdural
haemato ma' and, although suggesting that there is a trau­
matic origin, states that 'the causal mechanism remains
obscure'.! In 1971 , Guthkelch, a British neurosurgeon, first
proposed shaking as a cause of infantile subdural haematoma,
and the following year Caffey published his paper 'On the
theory and practice of shaking infants'. 2,J Since then, non­
accidental injury has been a subject of great importance to
those who have responsibility for the care, diagnosis and
investigation of children who may have been victims of abuse.
Teams of clinicians work closely together, aware that only
when they are in possession of all the information available
about the child can the correct diagnosis be reached.
ROLE OF THE RADIOLOGIST
Paediatric radiologists have expertise in the interpretation
of imaging of in fants and children and a familiarity with
the appearances of the normal brain and ske leton. The
appeara nces of the skeleton and brain followin g accidental
trauma will be well known to them, and they will have an
understanding of possible mechanisms of injury, and so be
able to judge whether the given history is appropriate for
the injury.
The first role of the radiolo gist is to be alert for injuries
that, in the light of the history supplied, may not have
I CHAPTER 3 I
Soft-tissue injury 69
Differential diagnosis 69
Conclusion 73
References 73
occurred accidentally, or have been identified as incidental
findings on a radiograph performed for another purpose.
Such cases w ill need to be discussed with the referring
clinician .
When referred a case of possible non-accidental injury,
the radiologist should review the presenting injury, and
discuss the patient with the clinician. A knowledge of nor­
mal skeletal development and variations in ossification can
prevent unn ecessary investigation. 4 - 7 There shou ld be a
recognized departme ntal protocol for a skeletal survey for
suspected non - accidental injury and all im ages should be
reviewed by the radiologist before the child leaves the
department. Th is is to ensure high-quality images and to
allow for any necessary supplementary views. An assess­
ment should be made of the age of any injuries. The possi­
bility of an underlying bone disease or other condition,
together w it h the need for appropriate im aging, should be
considered at this stage.
The report should be commun icated verb ally to the
responsible clin ician as soon as is practicable and a formal
written report issued promptly. If there is doubt as to the
presence or significance of a lesion, this should be clearly
stated together w ith a plan of investigation to clarify the
findings. The need for further imaging, in particular neuro­
imaging, should be discussed with the clinician at th is
stage. The radiologist should be available for any discus­
sions with clinicia ns and members of the chi ld protection
team, including the police. Because of the possibility of
child protection or other legal proceedings, all images and
reports should be retain ed in a secure place.
 48 I Radiology of child abuse
RADIOLOGICAL INVESTIGATIONS
A separate exposure should be obtained of each
Skeletal Survey
The radiographic skeletal survey is the mainstay of investi­
gation of non-acc idental injury in yo ung children and
infants. There has been much variability in the quality of
exa minations 8,9 a nd national stand ards have been intro­
duced by the American Academy of Paediatrics, JO the
American College of Radiology,l1 the Royal College of
Radiology and the Royal College of Paediatrics and Child
Health. 12 Table 3.1 shows a suggested protocol.
A ' babygram' of the whole body with a si ngle exposure
or coverage of the whole body using th ree or four films is
inadequate. High-quality radiographs must be obtained
with optimum exposure factors and good coning. High­
resolution cassettes should be used with no grid. The patient
name, side marker, and date and time of the examina tio n
must be clearly visib le on the radiograph . The names of the
radiographers must be recorded: they should wo rk in pairs.
Not only does this allow for increased efficiency in what
can be a prolonged investigation, but it provides legal safe­
guards for both the patient and the ra diographers.
The radiograp hers should obtain positive identifi cation
of the patient from the accompanying staff or carer and the
identity should be checked on the na me band. It is impor­
tant that the responsible clinician has discussed the need
for the skeletal survey with the parents or caregivers, as
they should be allowed to accompany the child during the
investigation. It may also be appropriate that a member of
the nursing staff is in attendance.
Table 3.1 Suggested protocol for initial skeletal survey
Skull AP and lateral views
Towne's view if occipital injury
suggested clinically
Spine Lateral views of cervical, thoracic and
lumbosacra l sp ine
Chest AP to include clavicles
Lateral
Oblique views of both ribs - each to
include whole chest
Abdomen AP whole abdomen, to include pelvis
and hips
Limbs AP both upper arms
AP both forearms
PA both hands
AP both femora
AP both lower legs
AP both feet
Supplementary Additional views of any questionable areas
AP and la tera l coned vie ws of metaphyses
AP, anteroposterior; PA, posteroanterior.
- - - ~ -
anatomical area to optimize image quality, reduce geomet­
ric distortion and allow for the detection of subtle abnor­
malities. At least two views of the skull should be obtained.
Both oblique views of the ribs should be obtained routinely.
Two perpendicular views should be obtained of any focal
injury. Coned views of the metaphyses, in anteroposterior
(AP) and lateral projections, may be helpful to confirm or
exclude classic metaphyseal lesions.
The skeletal survey is not an emergency investigation and
should be performed during the standard working day. It
requires two radiographers and a lo t of time. It should be
performed in the radiology department unless the child is
critically ill , in which case it may be performed using mobile
radiography equipment in the paediatric intensive care unit.
Image quali ty may be compromised in this situation.
Most radiol ogy departments no longer use hard-copy
radiographs for reporting. Soft-copy reporting has been
shown to be superior to hard copy owing to its abil ity to
vary the grey -scale settings and to magnify sections of the
image selectively. 13
Follow-up Skeletal Survey
If there is ongoing clinical concern, the skeletal survey should
be repeated in 10-14 days, apart from the skull radiographs.
This is particularly relevant if the presentation has been with
head injury and the initial skeletal survey has been normal. If
there are areas of radiological concern, repeat radiographs
should be obtained of these areas, also at lO-14 days (Fig.
3.1). For patients in whom there is no particular diagnostic
unceltainty, it is still important to obtain repeat chest radio­
graphs, including obliques, and views of the id entified bony
abnormalities. It has been shown that repeat skeletal surveys
can identify a greater number of fractures, particularly of
ribs, and can confirm suspected fra ctures. 14-16 Lack of change
over time can confirm a normal valiant. Radio grap hs taken
on two separate occasions can also help more accurate dating
of injury and show evidence of different ages of injuries.
Post-mortem Skeletal Survey
The same high standards should be applied to the examina­
tion of the deceased child as to the live child. A 'ba bygram'
should not be performed . Individual exposures of each
anatomical area should be obtained, with particular atten­
tion paid to the presence of COITect identifiers on each image.
The skeletal survey must be obtained prior to the autopsy
(Fig. 3.2). If necessary, the autopsy should be delayed to
allow this. The reporting radiologist should co mmunicate
his or her findings to the pathologist as soon as possible.
A verbal report will suffice, as long as it is documented in
the notes and in the final radiological report. Detailed
radiography of removed specimens of bone can be very
- - ---- --- -- - - -­
 Radiological investigations I 49

helpful (Figs 3.3 and 8.1, p. 147). Perpendicular views shou ld

be obtained of all bones. For the ribs this mean s that the
additional view is an axial , or supero-inferior image. 17, 18

Ultrasound

Ul trasound is not used routinely in the assessment of non­

accidental musculoskeletal injury, mainly because of its

Figure 3.2 Skeletal survey following a post- mortem

showing bowel in the chest. Known rib fra ctures are obscured
by bowe l.

Figure 3.1 (a) Admission film. Healing fracture s of the necks of

the seventh , eighth, ninth and tenth ribs can be seen on the left
as we ll as a mid-shaft fra cture of the clavicle. (b) Follo w-up fi lm
12 days later. Now visible are healing fra ctures of the left clavicle
and of the necks of the fifth to eighth ri bs on the left and of the
sixth to ninth ribs on ri ght. The left tenth rib has compl etely
remodelled in the 12-day interval and the ninth rib is identifiable
only by minimal irregul arity.
Figure 3.3 Subtle metaphysea l fra cture identified on post­
mortem skeletal survey prior to autopsy. High-definition specimen
radiograph following stripping of soft tissues. The irregu larity in
the distal femoral metaphysis was co nfirmed by histo logy to be a
fracture. It was not palpable.

- ~ "'- -

-----~--
 50 I Radiology of child abuse
Figure 3.4 Transfontanellar high-resolution ultrasound image.
Cystic cavities (arrows) at the junction of the grey and white
matter are tears caused by shearing.
dynamic, operator-dependent nature. Positive findin gs will
need to be documented using a nother imaging modality.
Ultrasound can be used, particularly in the very young child,
to assess cartilaginous epiphyseal fracture-separation, which
may be underdiagnosed radiographically. 19,20 Its use has also
been described in the detection of rib fractures. 21
The use of transfontanelJar ultrasound is more esta blished
although it is still a subjective examination and not useful i~
a legal setting, as experts find it difficult to review another
operator's images. It is easily performed in the clitically ill
infant at the bedside and has a definite role in the manage­
ment of an infant with head injUly, to monitor extra-axial
collections, ventricular dilatation and intracranial blood flow.
High-resolution ultrasound using high-frequency probes
(10 MHz) can differentiate subarachnoid from subdural
fluid, particularly when colour-flow imaging is used, Con­
tusional tears at the grey-white matter junction can be
demonstrated elegantly and may not be visualized on com­
puted tomography (CT) (Fig. 3.4). Fresh tears may contain a
clot and older tears appear cystic; these cysts may later col­
lapse. 22 Collapsed tears may be overlooked at autopsy
unless specifically targeted.
Fluid is well visualized by ultrasound as free intraperi­
toneal or pleural fluid. The liver, spleen and kidneys are
easily assessed by ultrasound, with an accuracy of 94 per
cent,23 but CT is more widely used in th e assessment of
blunt abdominal trauma. 24
Computed Tomography
Prior to the development of CT for neuroimaging, subdural
haemorrhage was identified indirectly by cranial angiography
- - -- - - - - ---
or air encep halography, or at post-mortem examination. In
accidental and inflicted head injury, CT shows different
patterns of injury.25.26 It is the imaging modality of choice
for suspected brain injury. lt is recommended as part of the
investigation of all suspected non-accidental injuries in
children less than 1 year of age. 12 There is some controversy
about this, particularly in a child with no neuroloo'ical
abnormality, regarding the radiation burden. If neuroimag­'"
ing is not performed, the rea son should be documented in
the patient's notes.
After presentation of a non-accidental injury, CT scan­
ning should be performed as soon as possible, without
intravenous contrast. This method of imaging is widely
available, easily performed, and fast. With newe r scanners
the need for sedation or general anaesthesia in the restless
child is less. Standard brain and bone windows should be
provided. Acute haemorrhage is readily visualized, even in
the subarachnoid space. Lesions requiring neurosurgical
intervention are reliably seen. Changes of brain oedema
can be subtle and may be missed. Known fractures, visible
on skull radiography, may not be visible on CT images, as
they may lie in the plane of the imageY Three-dimensional
(3D) volume rendering may be helpful in identifying frac­
tures or, indeed, in differentiating normal variants, such as
parietal fissures, accessory sutures and synchondroses,
from suspected fractures. 28 This imaging modality can be
used to assess the bony integrity of th e cervical spine and
fractures of the facial bones. 28 Suspected visceral trauma
may also be investigated using contrast-enhanced CT.
Although not used routinely, post-mortem CT scanning
may be useful, particularly as the normal limitations to dose
no longer apply and a high-resolution volumetric scan can
be obtained in a very short tim e. Owing to the amount of
data produced, reporting such an investigation will be time­
consuming for th e radiologist but may give considerable
further information about metaphyseal and rib fractures.
Magnetic Resonance Imaging
Magnetic resonance imaging (l'v1RI) is complementary to CT
scanning. Although l'v1RI is less sensitive for acute haemor­
rhage, especially in the subarachnoid space, it elegantly
demonstrates subacute subdural haemorrhage, contusions and
oedema (Fig. 3.5). Images should be obtained in transverse,
coronal and sagittal planes. The recommended sequences are
T1-weighted, T2-weighted, FLAIR (fluid-attenuated inversion
recovery), gradient echo, which is sensitive to blood products
and detects small areas of haemorrhage, and diffusion­
weighted imaging/apparent diffusion coefficient (OWl/AD C),
which demonstra tes areas of evolving brain injury, in parti­
cular areas of hypoxia-ischaemia. After presentation, an l'v1RI
should be ca rried out as soon as possible. Because the signal
from the damaged tissue normalizes after 7 days, OWl is most
useful in the first week. 29 It may be appropriate to assess vas­
cul ar structures with mag netic resonance (MR) arteriography
---- - -- - - - --

Figure 3.5 Magnetic resonance scan, Tl-weighted image,
showing a fresh small linear subdural in the right posterior
parietal region and high signal as a result of fresh haemorrhage
(arrows). Overlying this is a scalp haematoma. A skull fracture was
not visible on magnetic resonance imaging.
and venography, and to use transverse Tl-weighted fat­
saturated sequences to search specifically for inj ury to the
major vessels in the neck. 28 Spinal cord injury is best assessed
with MR, using short tau inversion recovery (STIR) sequences.
Post-mortem MR scanning has been shown to be more
sensitive than autopsy alone in the detection of a shearing
injury, cortical haemorrhage and mastoid fluid, and to be
equal to autopsy in the demonstration of cerebral oedema,
focal contusions and subfalcine herniation. 3D
Scintigraphy
Isotope bone scanning, using technetium-99-labelled
methylenediphosphonate (99Tc-MDPl, can show bone
pathology. In one series,3l 10 per cent of fractures were
seen only on scintigraphy, and Mandelstam has shown that
50 per cent of rib fractures may be visible only on isotope
bone scanning 32 (Fig. 3.6). The yield from radiography may
have improved since the introduction of oblique views of
the ribs at presentation. Many fractures also become appar­
ent on follow-up radiographs. Scintigraphy has a Jow sen­
sitivity for skull fractures. Because of the high activity in
normal growth plates, metaphyseal fractures may not be
apparent on isotope bone scanning, particularly if bilateral
(Fig. 3.7). Fractures may show increased activity on
scintigraphy for up to a year after injury. Any areas of
increased activity detected using scintigraphy should
Skeletal injuries I 51
Figure 3.6 Isotope bone scan of the upper body show ing
multiple areas of increased uptake in the posterior ribs,
representing fractures of the posterior shafts at their necks. A
spi ral fracture of the left humeral shaft is identifiable by the
asymmetrical uptake in the humeri.
therefore be further assessed by radiography. Bone scintig­
raphy is complementary to radiographic skeletal survey,
and may be useful in individual patients.
SKELETAL INJURIES
Fractures in child abuse are most common in infants and
children under 2 years of age. It is convenient to separate
fractures into two groups: (a) those that are seen fre­
quently, but are not very specific for non-accidental injury;
and (b) those that are highly specific for inflicted injury but
are less commonly seen (Tables 3.2 and 3.3).33
Periosteal New Bone
Fractures repair by the laying down of subperiosteal new
bone. With no obvious fracture, the presence of sub­
periosteal bone may be due to injury to the periosteum by
rough handl ing, or by acceleration-deceleration forces
during shaking. Physiological periosteal reaction is seen in
normal infants as young as 8 weeks, up to about 8 months
of age. It is smooth, rarely more than 2 mm in thickness
and is seen along the diaphyses of the long bones. It is usu­
ally symmetrical, though may be more obvious on one side.
Long-bone Fractures
Diaphyseal fractures of long bones are common in non­
accidental injury; some authors have found them to be four
 52 I Radiology of child abuse

(a) Table 3.2 Fractures that are frequent but with low specificity for
PO :;. T non-accidental injury

LT Midclavicular fractures
Simple linear skull fractures
Single diaphyseal fractures
Reproduced with pe rmission from Cart y33

Table 3.3 Fractures considered to have a high specificity for

child abuse

Meta physeal fractures

Rib fractu res
Scapular fractures
Fractures of the outer end of the clavicle
Vertebral fractures or subluxations
Finger injuries in non-ambulant children
Fractures of different ages
Bilateral fractures
Complex skull fractures
(b)
Reproduced wi th permission from Carty33

Figure 3.7 Isotope bone scans of lower body. (a) Posteroanterior

view shows metaphyseal fracture of the left proximal tibia (arrow).
(b) Lateral view of lower limbs. The fracture cannot be seen.
A posterior rib fracture is visible in the lower left chest (arrow).

times more common than the more specific metaphyseal

fracture. J4 It is important to correlate the appearance of the
fracture with the clinical history (Fig. 3.8).
Suspicion of abuse increases when there is evidence of
healing, indicating a delay in seeking medical care for a
child in pain. Excessive callus formation is seen owing to
movement at a fracture site, causing repetitive subperiosteal
bleeding (Fig. 3.9). Multiple fractures, especially in different
stages of healing, without an obvious history of significant
trauma, suggest at least poor parenting. In the presence of a
Figure 3.8 A 6-week-old infant presented to the accident and
emergency department. The mother gave a story of the child
waking up and not using her arm, saying she must have caught
it in the cot bars. The story is inappropriate for this fracture, which
was almost certainly caused by the child being roughly lifted by
this arm.

. .- -- - -- -- - - - -- ­

Figure 3.9 Toddler presenting with severe head injury. The
radiograph shows a new buckle fracture of the distal radius and
ulna and an older, untreated supracondylar humeral fracture, with
resulting extensive subperiosteal reaction.
fracture of high specificity, a diaphyseal fracture takes on
increased significance. An isolated shaft fracture becomes
more suspicious when there is other evidence of physical
abuse, such as bruising. Bilateral forearm fractures in
infants and toddlers are worrying beca use young children
do not ex hibit the normal protective instinct to break their
fall using their outstretched arms .34
There are welJ-recognized mechanisms for typical spiral,
oblique and transverse fractures. The described mechanism
given in the history should be assessed for each individual
fracture. It should be remembered that non-accidental injury
may occur when the limb is used to lift or drag the child, or
if the child is thrown, or be the result of direct injury. In par­
ticular, transverse fractures of the forearm bones can be sus­
tained as defensive injuries to ward off a blow (Fig. 3.10). It
should be remembered that infants who cannot yet roll over
are unlikely to fall off an elevated surface and that reported
falls in children who are not yet cnlising should raise suspi­
cion. 35 It is, therefore, important to know the degree of
mobility of the individual child. The spiral fracture of the
~------
Skeletal injuries I 53
Figure 3.10 A 14-year-old boy who died as a result of abuse.
Fractures of shafts of rad ius and ulna. These are typical defensive
fractures sustained when the arm is held up to ward off a blow
from an assailant.
tibia, the 'toddler's fracture', is very commo n as children
start to walk. Impaction fractures in the lower limbs appear
as torus or 'buckle' fractures, and are caused by transmission
of force vertically up the shaft of the bone when the child is
forcibly thumped down onto a hard surface in the standing
position (Fig. 3.11).
Metaphyseal Fractures
Although these are highly specific for non-accidental injury
in infants less than 1 year of age, they are seen in only 11-39
per cent of children surveyed. 36 .37 Their detection depends on
the quality of the radiography and they may be seen as cor­
ner or 'chip' fractures, suggesting local avulsion, or as
'bucket-handl e' fractures depending on the projection. They
are more commonly seen in the lower limb, but are also seen
in the upper limb. They can be caused by shaking, but, when
seen in a single limb, a twisting, pushing or pulling mecha­
nism may be responsible. This description is supported by the
report of classic metaphyseal fractures in a group of children
being treated for clubfoot, in whom the foot was pushed and
twisted into forced eversion and dorsiflexion before being put
into a cast. 38 Similar fractures are recognized rarely following
Caesarean section,39 a breech delivery or an armling delivery.
Metaphyseal corner fractures do not require any specific
treatment. They are not palpable.
 54 I Radiology of child abuse

Figure 3.11 Healing impaction fracture of the distal femoral

shaft with disruption of the normal contour anteriorly and
subperiosteal new bone (arrows). The apparent defect in the
anterior tibia is not a further fracture but is the site of the
unossified tibial tubercle.

These classic metaphyseal lesions are planar fractures

through the primary spongiosa of the metaphysis, with tra­
becular disruption giving rise to a disc of bone and cartilage.
At the periphery, adjacent to the periosteal collar, there are
more trabeculae so the fragment is denser there. The increased
visibility of the peripheral bone gives rise to the radiographic Figure 3.12 (a) Anteroposterior and (b) lateral views of typical
appearance of a comer fracture, sometimes with a subtle bucket handle metaphyseal fracture of the distal tibia (arrows).
metaphyseal lucency. An oblique view of the metaphysis may
demonstrate the disc-like fracture fragment. Some bones, such
as the tibia, are more likely to show a 'bucket-handle' appear­ incomplete and not pass through the whole metaphysis. Volu­
ance, as the metaphysis is relatively straight (Fig. 3.12). The metric CT imaging of metaphyseal fractures may demonstrate
comer fracture appearance is more likely to be seen in meta­ the disc-like fragment, but would appear to be unwalTanted
physes with a more complex appearance, such as the proximal except in specific cases, or for research purposes. Appearances
humerus or distal femur 40 (Fig. 3.13). The fracture may be during healing are variable. The fracture may become more

~ - --- ---- - -- - - ---

 Skeletal injuries I 55

Figure 3.15 Metaphyseal fractures of the distal femur and

proximal tibia with periosteal new bone along the lateral aspect of
the proximal tibial metaphysis. Periostea l new bqne along the medial
tibia, lateral fibula and femur is confined to the diaphysis and is
Figure 3.13 Typical metaphyseal corner fracture seen at the
probably physiological. Reproduced with permission from Carty33
posterior aspect of the distal femur (arrow).

Figure 3.14 Coned view of the ankle showing a healing corner

fracture of the posterior tibia and a fresh corner fracture anteriorly.
The healing posterior fracture shows periosteal new bone.

apparent, maximal at 10-14 days. Frequently, there is no other

evidence of healing as the periosteum may not be disrupted. If
it has been, there may be faint subperiosteal new bone forma­
tion, although more extensive new bone formation can result
in thickening of the adjacent cortex and a squaring off of the
metaphyseal contour40 (Figs 3.14-3.16). Healing is usually
complete within 6 weeks (Fig. 3.17). Extensive periosteal
Figure 3 .16 Healed metaphyseal fracture of the right proximal
humerus with subtle irregularity and sclerosis. No evidence of
subper iostea l new bone formation. Healing fractures of two
posterior ribs are visible (asterisks).
 56 I Radiology of child abuse
Figure 3.17 Close-up view of the knee. The lucent lines on the
distal femur and medial half of the proximal tibia are healed
metaphyseal fractures.
reaction extending up the shaft of the bone can suggest local
trauma, perhaps a result of forceful gripping of the bone
(Figs 3.18-3.21).
Epiphyseal Plate Injury
An epiphyseal plate injury may be identified using a radi­
ographic skeletal survey but, without a metaphyseal com­
ponent, when the epiphysis is un ossified they may be
missed. Ultrasound or MR scanning can show displacement
of the epiphysis and confirm the presence of a fracture/sep­
aration injury, and should be considered in a child who
does not use a limb and with no obvious explanation on
the skeletal survey.19
Rib Fractures
In infants and young children, rib fractures are highly spe­
cific for non-accidental injury. They can be seen in signifi­
cant accidental trauma, such as motor vehicle accidents, or in
infants with bone disease, such as rickets or bone disease of
prematurity.41,42 They are recognized as occurring rarely as a
result of birth trauma, presumably because of pressure from
-- -----
Figure 3.18 Healed metaphyseal fracture of the distal tibia,
lateral view, showing squared contour and slight irregularity.
There is irregular periosteal new bone extending proximally up
two-thirds of the shaft of the tibia, a disproportionate reaction
for a relatively minor fracture.
the maternal symphysis pubis, in which case there will prob­
ably be a history of a large baby and a difficult deJivery.43.44
Rib fractures are splinted by adjacent soft tissue and
neighbouring ribs. They may result in mild respiratory dis­
tress. If several ribs are fractured in more than one loca­
tion, this will give rise to a flail segment of chest wall, with
more severe respiratory consequences.
In child abuse, although ribs may be fractured by direct
trauma and impact, or by compression by kneeling on the
chest, the usual mechanism of injury is squeezing of the
chest by the perpetrator's hands. The infant is held facing
the adult, with the fingers on the back of the chest, the
palms at the side and the thumbs in front. Pressure causes
anteroposterior compression of the rib cage.
Although fractures can be seen anywhere along the rib
arc, the most characteristic fracture site involves the posterior
rib, at the rib head and the costovertebral junction45 (Figs 3.1,
3.6 and 3.7, pp. 51-52). This is due to leverage of the poste­
rior rib on the fulcrum of the transverse process of the verte­
bra, with an adjacent fracture. Similar fractures have also
been demonstrated in a cadaver study simulating median
sternotomy with rib retraction. 46 In the same paper, a study
was done comparing sternal compression and AP manual
compression in rabbits. No rib fractures were seen following
---- - --
 Skeletal injuries I 57

Figure 3.20 Healing metaphyseal fracture of the distal humerus

with gross periosteal new bone extending almost to the proximal
meta physis.

Figure 3.19 (al Admission radiograph. Tiny metaphyseal

fractures of the distal tibia and fibula (arrows). There is soft-tissue
swelling of the calf. (bl Follow-up film shows massive healing
subperiosteal new bone, mainly due to healing of a subperiosteal
haematoma, the extent of which was totally inappropriate for the
tiny fracture. This represents very extensive trauma to the leg.

sternal compression on a firm surface and the posterior ribs Figure 3.21 Healing bucket handle metaphyseal fractures of the
showed no change in their relationship to the spine; however, distal tibia bilaterally with disproportionate periosteal new bone
fractures did occur with manual compression. Lateral rib formation. Irregular periosteal new bone on the fibular diaphyses
fractures are due to the compressive force on the ribs and are bilaterally. The appearances probably represent a direct gripping injury.

-... . , - ~
 58 I Radiology of child abuse
shown at autopsy to have distraction of their outer surface
and impaction of the inner surface 47 (Figs 3.22 and 3.23).
Fractures of the anterior ends of the ribs involve the costo­
chondral junction. These show disruption of the posterior
surface, which may be a result of direct pressure from the
thumbs.
Rib fractures may be difficult to see unless there is some
displacement. Only 36 per cent of fractures identified at
autopsy were visible on the skeletal survey in Kleinman 's
study.47 Lateral rib fractures may show adjacent extrapleural
opacity, owing to focal haemorrhage and pleural reaction.
Posterior lib fractures may also be obscured by the trans­
verse processes. Oblique views of the ribs are useful in detec­
tion of both posterior rib fractures and fractures at the
costochondral junction. However, many rib fractures do not
become visible until the follow-up radiograph shows callus
formation. Volumetric CT scanning of the thorax may reveal
rib fractures, but the radiation dose associated would pre­
clude its routine use for this purpose.
Specimen radiography of resected ribs should be per­
formed in both the frontal and the axial projections.
There have been studies of children which have shown no
radiographic evidence of lib fractures resulting from car­
diopulmonary resuscitation (CPR)48.49 A recent studyS° found
autopsy evidence of rib fractures in 11 per cent of 70 infants
who received CPR immediately before death. The fractures
were all anterolateral, linear and often bilateral. There was
little, if any, associated blood and no reactive change.
Figure 3.22 Child admitted moribund. Healed fractures right
sixth and seventh ribs postero'atera"y (arrows) plus fresh corner
fracture proximal right humerus (white arrow). The healed left
clavicular fracture could be a birth injury.
. --
Spinal Trauma
-----
Severe spinal trauma is much less common than other frac­
tures in non-accidental injury. It may be underdiagnosed, but
routine lateral radiographs of the whole spine (Fig. 3.24) and
follow-up skeletal surveys may increase its identification.
Vertebral body fractures are thought to be due to hyper­
flexion. This may be during a shaking episode, holding the
child by the chest or the shoulders, or by slamming the
child down onto a hard surface on the buttocks, transmit­
ting the force vertically up the spinal column. These may
be anterior compression fractures, fractures extending to
the superior end plate, or a combination of the two. The
superior end plate fracture is thought to be similar to the
classic metaphyseal fracture.5l Fractures may also occur of
the spinous process, with a similar mechanism. The carti­
lage tip of the spinous process may be avulsed with only a
small fragment of bone, which may be recognized on the
lateral spine radiographs. If diagnosis is delayed, the mar­
gins of the avulsed fragment may be irregular, no longer
matching the defect in the spinous process. This is due to
further growth of the avulsed fragment. 52 Fracture disloca­
tion of the spine may occur, which can be associated with
cord compression and long-term neurological damage. It is
described at any level in the spine, including the sacrococ­
cygeal region. It may present as an unexplained kypho­
sis. 53-56 Traumatic spondylolysis of C2 , the hangman's
fracture, has been described with anterior subluxation of
C2 on C3. 57 Full visualization of the vertebral injury may
Figure 3.23 There are multiple healing rib fractures in both
axillae with periosteal new bone around the fractures.
- --
 Skeletal injuries I 59

require CT, but MRI will be necessalY to exclude a ny asso­ fracture healing. A radiologist whose practice involves paedi­
ciated cord injury. atric trauma will have co nsiderable experience in the report­
ing of acute and healing fractures, when the timing of the
Digital Injuries injUly is precisely known . This can then be extrapolated to
the appearance of fractures for which there is an inad equate
history. Despite the lack of published data, there is remark­
Injuries to the metacarpals, metatarsals and phal a nges are
able agreement between radiologists. Approximate dating of
uncommon, but have a high specificity for abuse. They are
usually torus fractures, and the mechanism is thou ght to be
twisting, bending or hyperextension 58 (Fig. 3.25). Crush
fractures may also be seen.

Other Bony Injuries

Fracture of the outer end of the clavicle usuall y resu lts

from a fall. A midclavicular frac ture may be as a res ult
of birth trauma and an appropriate histOlY should be
sought. Fractures of the acrom ion and the body of the
scapula 56 ,59 (Fig. 3.2 6) have been described, as have sternal
fractures. GO Pelvic fractures usually involve the superior
pubic rami. 61

Dating of Injuries

Precise dating is impossible. Accepted g uidelines ha ve been

published (Table 3.4),62,63 but there have been few studies of

Figure 3.25 Healing fractures of the bases of the proximal

phalanges, great and little toes, due to hyp erexte nsion injuries.

Figure 3.26 Ossifying haematoma projected over scapula in a

Figure 3.24 Lateral view of the lumbosac ral spine show in g child who had been severe ly beaten. Reproduced with permission
compression fracture of L2 (arrow) with a mild lum bar kyphosis. from Ca rty,s9

-
~ -
~-------

-
~
- -- , -
- --~--
 60 I Radiology of child abuse

Table 3.4 Dating of fractures (adapted from O'Con~or .IF, Cohen J

In: Diagnostic imaging of child abuse. Baltimore: Williams and
Wilkins, 1987112)

Time Peak

Soft-tissue resolution 2-10 days 4-10 days

Early periosteal new bone 4-21 days 10-14 days
Loss of fracture line 10-21 days 14-21 days
definition
Soft callus 10-21 days 14-21 days
Hard callus 14-90 days 21-42 days
Remodelling 3 months to
2 years
Reproduced with permission from Carty] 3
Figure 3.27 Posterior linear parietal skull fracture in an abused
child. The only indication that this was caused by abuse was the
fractures is always given as a range, with the limits becoming la ck of clinical history for the fracture.
wider as time elapses from the injury. The appearance of sub­
periosteal new bone formation is the earliest sign of healing.
It has been described as being visible as early as 4 days after
injury and seen in 50 per cent within 2 weeks.63

HEAD INJURY

In non-accidental injury, the infant and young child may

suffer injury to the scalp, skull and face, suffer an intracra­
nial or brain injury, or a combination of the two. Brain
injury is the leading cause of mortality and of significant
neurological impairment in survivors. 64 - 67
The infant with brain injury may present acutely as
encephalopathy or seizures, less acutely as irritability or
vomiting, or as a relatively well child with macrocrania or
failure to thrive. The initial diagnosis may be of meningitis,
apnoea or collapse of unknown cause, or a search may be
made for an infectious or gastrointestinal cause for vomiting.
Findings suggestive of non-accidental injury may be seen on Figure 3.28 Wide diastatic skull fracture, allegedly caused by
other investigations, such as incidental healing rib fractures falling off a sofa. In addition, there is a further extensive linear
on chest radiographs, or subdural collections on cranial ultra­ fracture across the lower part of the skull.
sound. These will then prompt further investigations.

bilateral fractures, complex or stellate fractures and grow­

Skull Fractures
It is important to remember that a child may sustain a skull
fracture with no associated brain injury, or may have severe
brain injury without a fracture.68 A sk ull fracture requires
direct impact or compression. 69 There have been many stud­
ies of falls, including population studies/ o- 73 showing that
skull fractures are unusual in low falls and that a fall from
about 1 m is usually required for a skull to fracture. This is
the approximate height of a fall from an adult's arms.
A typical accidental skull fracture is a unilateral linear
parieta l fracture, which is also the most common fracture
seen in non-accidental head injury (Fig. 3.27). However
ing fractures are more common in non-accidental injury74
(Figs 3.28, 3.29 and 14.6, p. 299). Bilateral fractures can be
seen in accidental trauma with a fall onto the occiput or
the vertex, with symmetrical transmission of force. No
sku ll fracture type is characteristic, and in any individual
patient the appearance of the fracture alone does not allow
ajudgement as to its cause.
Dating of Skull Fractures
Skull fractures cannot be dated radiographically. Some
indication of age may be given by the presence of adjacent

- - - - . - --- -- -- - - -­

Figure 3.29 Symmetrical horizontal fractures, which meet in
the midline, and an additional linear parietal fracture on the left.
There was no clinical history to explain how these fractures were
sustained.
soft-tissue swelling, which resolves gradually according to
the size of the haematoma. However, care should be taken, as
a swelling may appear after some delay as a result of seepage
of cerebrospinal fluid (CSF) through a fracture, or enlarge­
ment of a scalp haematoma during resolution. The clinician
should also realize that focal scalp swelling may occur in the
absence of an underlying fracture (Figs 3.30 75 and 3.31).
Extradural Haematoma
Extradural haematoma is unusual in children, and rare in
non-accidental head injury. Radiologically, it is a lentiform
collection overlying the brain, tending to compress the
underlying brain substance. There may be an associated
fracture, with typical tearing of the middle meningeal
artery, although venous bleeding can cause an extradural
haematoma.
Subdural Haematoma
Subdural haemorrhage is caused by bleeding from bridging
veins crossing the subdural space, which are stretched and
torn when the brain moves excessively relative to the over­
lying dura. A degree of trauma is required to tear these veins,
though the minimum degree is not known. Mechanisms of
injury in inflicted trauma are discussed in Chapter 14.
Subdural haemorrhage is seen in birth trauma 76 but has
been shown to be present in 8-17 per cent of asymptomatic
neonates, including 6-26 per cent of spontaneous vaginal
Head injury I 61
deliveries. 77 ,78 Subdural blood was seen most frequently in
the posterior fossa, and no interhemispheric blood was
seen in either series, of 199 babies in total. Delivery by
Caesarean section may still be associated with subdural
haemorrhage, as there may be some difficulty in disimpact­
ing the engaged head. In Whitby's cohort of 111 babies, fol­
lowed for 2 years, the subdural haemorrhages had all
cleared by 4 weeks?7 There were no recurrent bleeds.
Accidental trauma can result in subdural haemorrhage,
most commonly over the cerebral convexities and at a single
site, localized to or opposite the impact, although occasionally,
when there has been major trauma, the haemorrhage may
spread over the convexity. Interhemispheric haematoma is
rare, but is described, and is therefore not specific for inflicted
head injury?9 Homogeneous hyperdense haematoma is seen
more often in accidental injury, though this changes in the
days after presentation in 25 per cent. 79 Subdural haemor­
rhage is, however, more common in inflicted head injUly. In a
series of 100 children admitted as a result of head injury,
Duhaime et al 80 identified 76 as accidental and 24 as inflicted.
Only three children with accidental head injury had subdural
haemorrhage, all of whom had been involved in road traffic
accidents. By contrast, ) 3 of the 24 children with non­
accidental head injury showed subdural haemorrhage, with
either no history of injury or history of a low fall (~l m) in
22 patients, and admitted assault in two. Ewing-Cobbs et al 81
studied 40 children with a head injury, 20 accidental, 20 non­
accidental. Subdural haemorrhages were seen in 16 of the
non-accidental group, but in only nine of the accidental
group, most of whom had been involved in motor vehicle
accidents and none of whom had fallen. Hymel et al S2 com­
pared CT scans of 39 children with inflicted head injury with
scans from 39 control patients with accidental head injury.
Subdural haemorrhage was found in 17 in the non-accidental
group, but in only four in the accidental group; interhemi­
spheric falx haemorrhage was seen in 17 and 2 respectively.
In non-accidental head injury, subdural haemorrhage is
usually bilateral, and almost always interhemispheric (Figs
3.30 and 3.32); it spreads over the cerebral convexity and
may be seen in the middle cranial fossa (Fig. 3.33, p. 64).
Although it is often obvious, it may be a shallow layer of
blood, easily overlooked. It is uncommon for subdural
haematomas at presentation to cause significant mass
effect on the underlying brain (Fig. 3.30), though blood
may continue to accumulate with time. S) Mixed density
haematoma is more common at presentation in non­
accidental injury, but has been described within 2 days79
and 1 week of accidental injury.s4
Precise dating of subdural haemorrhage is difficult on
both CT and MR scans, unlike the dating of intracerebral
haemorrhage in adults. There seems to be considerable
variation in density and intensity, which may be related to
the initial volume of haemorrhage, the patient's haemoglo­
bin level, dilution by CSF, the oxygen tension within the
CSF, and by a layering out of blood within the haemor­
rhage, the 'haematocrit effect' (Fig. 3.34, p. 65). Intervention,
 62 I Radiology of child abuse

Figure 3.30 Magnetic resonance scans of an infant. (a) Fluid-attenuated inversion recovery [FLAIR) coronal, (b) Tl-weighted sagittal at
presentation and (c) T1 sagittal 10 weeks later. High-intensity interhemispheric subdural haemorrhage is indicated by arrowheads. The
low-intensity posterior fossa subdural collections [stars) are presumed to be older. Secondary herniation of the craniocervical junction
(white arrows) resolved along with the subdural. An ill-defined, subependymal, high-intensity signal suggests a shearing injury [arrow).
There is also a large subgaleal fluid collection [asterisk). Adapted with permission from McPh il lips.75

- 4.: - - -- --- ~ - ~

Figure 3.31 Computerized tomography scan shows a fresh,
high-density scalp haematoma. Subarachnoid haemorrhage is
visible in the sulci, mainly over the left parietal and occipita l
lobes.
Figure 3.32 (a) Acute computerized tomography scan showing high-density fresh interhemispheric subdural haemorrhage poster iorly.
There is some loss of the normal grey-white matter different iation due to mild cerebral oedema . (b) Follow-up scan 4 weeks later. There is
mild generalized cerebral atrophy.
Head injury I 63
such as the insertion of a pressure monitoring device or
a subdural tap, will also have an effect. Appearances on
sequential scans and comparison of initial CT and MR
scans may be helpful.
Subarachnoid Haemorrhage
Spontaneous subarachnoid bleeds can occur due to rupture
of an arteriovenous malformation or an aneurysm, which
is thought to be rare under 1 year of age. The distribution
of blood differs from that in non-accidental injury.76 Acute
subarachnoid blood is more clearly seen on CT than on MR
(Fig. 3.31). In inflicted brain injury subarachnoid blood is
not unusual and may result, acutely, in arterial vasospasm
and later in secondary hydrocephalus.
Parenchymal Brain Injury
Focal areas of parenchymal contusion and haematoma are
caused by impact of the brain against the adjacent skull or
dura, such as the falx or tentorium, and are most common
in the cortical grey matter of the frontal and anterior tem­
poral lobes. It is not clear if contusional injUly can be
caused by shaking alone, or whether impact against a firm
surface is required. While these injuries may be seen on CT
64 I Radiology of child abuse

Figure 3.33 (a) Fluid-attenuated inve rsion recovery (FLAIR) coronal MR image at presentation shows a thin layer of subdural blood over
the convexity on the right and in the subtemporal region bilaterally (arrows). T2-weighted transverse images. (b) At presentation there is
loss of grey-white matter differentiation on the right. with sl ight swelling and subtle midline shift. (e) Ten days later there is increased
intensity in the right hemisphere and left frontal lobe. with prominent sulci due to early atrophy. (d) Nine months later there is extensive
atrophy on the right. sparing only the basal ganglia. The left hemisphere is normal at this level. Adapted with permission from McPhillips.75

- - --
 Head injury I 65

scanning, MR scanning, particularly in the coronal plane,
demonstrates them well, and gradient echo (GRE) images
are very sensitive for blood products, which are seen as low
intensity. They may be subtle at autopsy, and knowledge of
the MR findings will direct attention to the appropriate
areas 30
Shearing injury is unusual in head injury but, when seen
in the absence of a history of a high-velocity injury, such as
a road traffic accident, it is highly suggestive of non-accidental
injury. The most common site for focal shearing injury is at
the grey-white matter junction in the frontal or frontopari­
etal lobes or in the corpus callosum (Figs 3.30 and 8.17,
p. 157). At the grey-white matter junction it may actually
be seen as a tear, which is readily visualized using high­
frequency ultrasound scanning, and is seen as a low­
echogenicity focus. 22 Tears may be seen on CT, but as they lie
close to the vertex they are not easily characterized. On MR
they may be seen as haemorrhagic (Fig. 3.34) tears initially,
later becoming cystic. Follow-up imaging shows that some
persist as cysts 85 (Fig. 3.35), while some show focal gliosis

Figure 3.34 Sagittal Tl-weighted magnetic resonance images. The presentation scan (al shows increased intensity posteriorly, in
keeping with layering in a subdural haemorrhage (arrowheads), and a focal area of high intensity (arrow) over the frontoparietal
convexity, suggestive of focal clot formation. Six days later (bl, an ill-defined, high-intensity signal can be seen in the frontoparietal
subcortical white matter (arrow), representing a tear. Two months later (cl, there is residual focal atrophy (white arrows). Adapted with
permission from McPhilli ps l5

..
-------
 66 I Radiology of child abuse

Figure 3.35 Coronal fluid-attenuated inversion recovery (FLAIR) magnetic resonance im ages at (al 3 days, (bl 3 weeks and (cl13
months after presentation. (a) III-defined, high-inten sity signal over the left convexity (arrow) and subtle irregularity inthe underlying
parenchyma. (b) Cortical tear with low-intensity cyst formation (arrow) . (c) The tear now appears as a slit-like lesion (arrowl with
adjacent gliosis (arrowheads) and subtle atrophy. Adapted from with permission from McPhilli ps 85

and others collapse with associated white matter loss 86 Oedema or swelling of the brain may be as a direct result
(Fig. 3.3 4) . Sometimes shearing injury may be seen as tiny of trauma, or may be a result of hypoxic-ischaemic damage
focal haemorrhagic areas on MR scanning, and knowledge of or hyperaemia. Traumatic oedema is usually associated with
these findings may direct the pathologist to a subtle lesion. 3D contusion or parenchymal haemorrh age. Hypoxic-ischaemic

0- _ - - - ­ - - ­

Figure 3.36 CT scan showing an acu te reversal sign with low­
density brain, loss of normal grey-white matter differentiation
and relative sparing of the thal am i an d basal ganglia, which are
dense by comparison. Also visible are multiple areas of
intra cerebra I petech ia I haemorrhage; bi lateraI ch ron ic subd ura I
haemorrhages with fresh interhemi spheric subdural blood. There
is an old linear infarct in the left occipital lobe.
mJury is the most common radiological manifestation of
non-accidental head injury and is often quite extensive. In
severe cases, there is sparing only of the cerebellum and pos­
sibly the basal ganglia and th alami , giving rise to the 'rever­
sal sign' on CT scanning, in which the cerebral hemispheres
are hypodense compared with these structures 87 (Fig. 3.36).
Except for the occasional documented vascular dissection ,
the underlying causes are not well established. There are
many theories abou t the seconda ry mechanisms of brain
injury and cerebral oedema, wh ich are discussed further in
Chapter 14. In hypoxic-ischaemic damage there is loss of
clarity of the grey- white matter interface on all scanning
techniques (Fig. 3.33). On CT scanning, the affected area may
show abnormally low density (Fig. 3.32, p. 63), although on
MR sequences the in tensity will reflect the increased water
content, low on Tl-weighted sequences and high on T2­
weighted sequences. Th e combination of diffusion-weighted
imaging (OWl) and ap parent diffusion coefficient (ADC)
mapping will show restricted diffusion, and is more sensitive
than other MR sequences; however, this is not specific and
may be seen in metabolic disorders, following seizures, and
in encephalitis. It may be useful in predicting the prognosis.
~ ­ ---~-
Head injury I 67
Sequelae of Head Injury
Some children who have suffered non-acciden ta l hea d
injury will have no residu al radiological abnormality on
follow-up several years later. 64 - 67 Others show mild cere­
bral atrophy with prominent ve ntricles and CSF spaces
(Fig. 3.35). Focal injury may be seen as areas of focal
infarction, while some who have had extens ive
hypoxic-ischaemic injury may have widespread multicys­
tic encephalomalacia and severe atrop hy (Fig. 3.33).
Hydrocephalus may be seen secondary to haemo rrhagic
arachnoiditis. Some children with sub dural haemorrhage
who have also developed cerebral atrophy may have chronic
subdural collections (Figs 14.9 and 14.10, pp. 300- 30 1).
Spinal Cord Injury
Radiological appearances of spina l inju ry in non-accidental
injury are not well described. There are reports of
fracture-subluxation of the vertebrae with associated cord
compression.53 . 54 For eva luation of suspected injuries to the
spinal cord, MR is necessary as injlllies may be present in the
absence of other rad iographic abnorma lity (spinal cord
injury with out radiographic abnorma lity, SCIWORA). Under­
lying abnormalities, such as atlan to- ax ial instability or
block vertebra, which would predispose to cord injury,
should also be sought. 28
Cerebrospinal Fluid Spaces
The depth of CSF fluid spaces surrounding the brain
changes in the first 2 years of life. Paediatric radiologists
who scan young children are familiar with the healthy
child who has widened CSF spaces and minim al, if any,
ventricular dilatation. It is considered a developmental
variant. Kleinman et al 88 reviewed the CT scans of 34 nor­
mal children, finding that the extraventricular subarach­
noid space is increased in children under 2 years of age,
normalizing at aro und 2 years and becoming 'monoto ­
nously uni form ' by 3 yea rs. They postulated theories for
this transient increase in CSF spaces, suggesting that it was
a transient alteration in CSF dynamics, perhaps associa ted
with a response to the grow ing brain. 88 Kapila et al 89 stud­
ied CSF dynamics using nuclear cisternography a nd found
no evidence of communicating hydrocephalus. Libicher
and Trbger90 scanned 89 infants to determin e an upper limit
of normal, based on the 95th percentile. Wilkinso n et al 91
obtained similar results, with some variation of depth with
position. Fessell et al 92 reviewed cranial ultraso und scans
of 38 patients with macrocrania, an occipitofrontal head
circumference of 95 per cent or greater for age, who had
been followed up for a mean period of 55 weeks. They
showed that a CSF depth of < 10 mm had a 94-100 per cent
 68 I Radiology of child abuse

negative predictive value for the development of neurolog­

ical abnormality.
These [SF spaces should co ntain fluid with the same
characteristics as the intraventricular [SF. If the density on
[T images, intensity on MR images or echogenicity on
ultrasound scanning is different, then this raises the possi­
bility of inflammation or haemorrhage, and requires fur­
ther investigation. Ultrasound scanning, using colour flow
imaging to demonstrate subarachnoid vessels, can demon­
strate which comp artment is involved.

VISCERAL INJURIES

Almost any organ can be injured as a result of inflicted

trauma, bu t visceral injury is very unusual in child abuse,
when compared with the incidence of skeletal and head
injury. The typical abdominal and chest injuries result from
kicking or punching the abdomen and chest, or from kneel­
ing or standing on the child. The child may present acutely
owing to peritonitis and blood loss, or with late complica­
tions of injury, such as su bacute gastrointestinal obstruc­
tion resulting from stricture forma tion or pancreatic
pseudocyst format ion. The investigation of visceral injury
should be the same as for other blunt abdominal trauma. Figure 3.37 Barium meal in a child who presented wi th vomiting,
dehydration and weight loss, showing proximal duodenal dilatation
secondary to duodenal haematoma causing an apparent stricture.
Liver, Spleen, Kidneys and Adrenals The jejunal mucosa is thickened secondary to haemorrhage.

trauma. Upp er gastrointestinal contrast studies may be use­

The mechanism of abusive injury of these organs does not
differ fro m accidental blunt trauma . Direct compression
can cause laceration, rupture and haemorrhage. Associated
rib fractures may be seen and the presence of other non­
accidental injury and the lack of a specific history of injury
may suggest abuse. 93 ­95
Pancreas
Pancreatic trauma can result in acute pancreatitis, often
haemorrhagic. The presence of associated injuries, particu­
larly to nearby bowel, may suggest the traumatic aetiology.
Pseudocyst formation can occur with associated mass
effect. For the assessment of the degree of injury and asso­
ciated abnormality it may be necessary to use [T scann ing
with intravenous and oral contrast. 96 -9S
Bowel Injury
Duod ena l and jejunal injury can occur from direct trauma ,
compressing the duod en um against the spin e, but may also
result from deceleration forces with shearing injury to the
root of the mesentery.99 Duod enal hae matoma may be seen
or transection of the j ej unum in the region of the ligament
of Treitz. Ultrasound is not usually useful for ep igastric
ful if the child is not acutely ill (Fig. 3.37), bu t abdominal
[T (Fig. 3.38) with intravenous and oral contrast may be
most helpful. Mesenteric tears with associated vascular
injury may present late with multiple strictures. 94 Gastric
rupture is a rare occurrence, and is likel y to be due to com­
pression of a distended stom ach. 95 Pneumatosis of the gas­
tric wall has been described. 100
Rectal perforation is usually the result of penetrating
injury, associated with sexual ab use. 94 • 10 1
Visceral Chest Trauma
Injury to the lungs, heart and mediastinum is unusual, a nd
likely to be a result of direct compression injury. This can
result in diaphragmatic rupture, pneumothorax or pneu­
mopericardium. Airway obstruction may cause a pneumo­
mediastinum. Rib fractures may cause a haemothorax (Fig.
3.39). Secondary inflammatory change .may extend to the
mediastinum from adjacent structures, most commonly
being involved with pancreatitis.
Penetrating Trauma
Although pharyngeal perforation is a well-recognized iatro­
genic injury, it is a rare manifestation of non-accidental

-- - - - - - --

Figure 3.38 Abdominal computerized tomography with
intravenous contrast. (a) Fluid in dilated duodenal loop (asterisks)
with denser haematoma medially. Also visible is free fluid around
the liver and spleen. (b) Caudal image of the same child also
shows free intraperitoneal fluid and intense enhancement of
bowel wall due to 'shock bowel'.
injury, but has been described. In infants it is due to the
insertion of a finger or a sharp object. IO l-1 03 In older chil­
dren, penetration may be related to sexual abuse. Resulting
retropharyngeal abscess formation may compromise
the airway and inflammation may extend into the
mediastinum.
Ng et al 94 described two cases of penetration by multi­
ple needles. In one case, the needle marks were visible and
CT scanning showed one needle to lie in close proximity to
the carotid artery. In another case, the needles had been
inserted through the umbilicus and probably per rectum,
with abscess formation in the abdomen and pelvis. Foreign
bodies seen on radiographs must not be assumed to be
external to the patient. 94
Differential diagnosis I 69
Figure 3.39 Healing rib fractures in both axillae with a right
haemothorax. Recent angulated fracture of right humerus. The
raised right hemidiaphragm is due to liver contusion.
SOFT-TISSUE INJURY
It is not unusual for soft-tissue injury to be sufficient to
cause obliteration of fat planes on radiographs, and such
changes can also be seen on ultrasound and MR imaging.
But, in general, radiology cannot reliably detect bruiSing.
Older children may be beaten, with focal haemorrhage in
muscle and soft tissue. These may ossify and appear radi­
ographically as heterotop ic new bone formation. 104 The
mechanism is similar to post-traumatic myositis ossificans.
Carty has described soft-tissue calcification of a neck­
lace distribution in the neck. It is thought that this may
represent focal fat necrosis and ischaemia. 105
DIFFERENTIAL DIAGNOSIS
To miss a diagnosis of non-accidental injury and to fail to
safeguard the child may result in further injury or death. To
misdiagnose non-accidental injury, when there is another
cause for the child's condition and radiological appear­
ances, is to cause heartbreak and devastation to a family. It
is therefore important that all paediatric radiologists have a
knowledge of the differential diagnosis of both head
injuries and individual skeletal abnormalities.
Head Injury
The major differential for scalp swelling and skull fracture
is accidenta l injury. The history should be sufficient to
explain the findings and the degree of injury. Bony skull
defects with bruising or the appearance of periorbital
haematoma are recognized presentations of leukaemia and
metastatic neuroblastoma.
 70 I Radiology of child abuse
Subdural Haematoma
Even with a bleeding diathesis, some trauma is required to
cause a subdural bleed, but it may be relatively minor in
nature. Haematological investigation is appropriate in
unexplained or suspicious cases of subdural haemorrhage.
Rarely, an aneurysm or arteriovenous malformation may
bleed, causing subarachnoid and subdural haemorrhage.
Subdural effusions, which may be abnormal when com­
pared with intraventricular CSF on scanning, can be seen
in meningitis. This can be confirmed clinically and by lab­
oratory testing. Herpes simplex encephalitis can be haem­
orrhagic; again clinical assessment and virological
investigation should clarify this. Glutaric aciduria is
uncommon, but may present with macrocrania. It can be
recognized by the combination of cerebral atrophy, with
particular widening of the Sylvian fissure and the CSF
space a nterior to the temporal lobes, and abnormal signal
in the basal ganglia and periventricular white matter. 106
Physiological Periosteal Reaction
Physiological periosteal reaction is seen in normal infants
between the ages of 8 weeks and about 8 months (Fig. 3.40).
It is rarely more than 2 mm in thickness, smooth and seen
along the diaphyses of the long bones. It does not extend
to the epiphyseal plate. It is usually symmetrical, though
may be more obvious on one side and, if identified on the
tibia, should be visible on the femur (Fig. 3.15, p. 55).
Accidential Injury
In accidental injury, the fracture should be compatible
with the mechanism described in the history. It is more
Figure 3.40 Simple physiological subperiosteal new bone seen
along the diaphysis of both femora , both medially and laterally.
~--
important to be aware of the patient's stage of develop­
ment than their actual age, can they roll, do they stand,
are they cruising? This information, together with the
history, can often clarify a worrying scenario (Figs 3.41
and 3.42).
Normal Variants
The paediatric radiologist should be familiar with the
appearances of nutrient foramina, which may be mistaken
for fractures by the unwary.107 There are numerous variants
of ossification, particularly around metaphyses. These
may be spurs, beaks and the 'step-off appearance, which
resemble metaphyseal fracture. But close inspection, using
coned views and magnification of high-quality radio­
graphs if necessary, will show that these structures are in
continuity with normal bone and do not represent frac­
tures. 4 The acromion may also show irregular ossification. 6
Irregularity of the anterior cortex of the proximal tibia
is also a normal variant,4 as is irregularity of the radial
shaft.
There are numerous sutures and synchondroses in
the skull, most of which are easily recognized. The parietal
fissure, or 'strip suture', is frequently misdiagnosed as a
fracture. It is a short tapering lucency extending from the
skull vertex into the parietal bone. They are no wider than
Figure 3.41 Typical innocent toddler's spiral fracture.
­ - --

2 mm, and are not assoc iated with overlying soft-tissue
swelling.
Birth Trauma
The most common fracture sustained during delivery is of the
midshaft of the clavicle. It may not be recogni zed at birth but
it either presents at 2-3 weeks of age, with a palpable lump,
or is seen as an incidental finding on a chest radiograph. Most
h ave healed by 12 weeks and are undetectable. A few fail to
unite, leaving a pseudoal1hrosis of the clavicle, which may be
mistaken for a new fracture at a later date. Other diaphyseal
fractures are usually recognized owing to reduced movement
of the limb, humeral fractures in shoulder dystocia and
femoral fractures in extended breech delivery. Rib fractures,
resulting from impaction against the maternal symphysis
pubis, are usually unilateral. 43 ,44 They may present as a result
of crepitus or respiratOlY distress, or be seen as an incidental
finding on a chest radiograph . Metaphyseal fractures may
rarely be due to birth trauma, when a limb has been twisted or
pulled, in the course of Caesarean section,39 an armling deliv­
ery or a breech delivery. The finding of subdural haematoma
in a significant number of asymptomatic neonates has been
Figure 3.42 (a) Simple torus fracture of the distal tibia with
slight change in the normally smooth contour of the anterior
cortex. This fracture is technically metaphyseal, but this is a
simple accidental injury, not non-accidental injury. (b) Healing
with sclerosis at the fracture site.
Differential diagnosis I 71
described, and it should be remembered that it can be detected
for up to 4 weeks following delivery.77,78
Osteomyelitis
Periosteal reaction and minor metaphysea l or cortical
irregularity may be see n on ultrasound or, sev era l days
later, radiographically. In the absence of a histolY of
trauma or clinical evidence of infection this may be
ascri bed to non-accidental injury.
Metabolic Bone Disease
Several conditions can present with reluctance to use a
limb, in the absence of trauma. Radiographs may show a
metaphyseal lucent line. Careful inspection of the radi­
ograph will show that this is present at every metaphys is,
though more obvious at some. There is also no visible frac­
ture. This is a well-recognized presentation of leukae mia
and neuroblastoma, and the child is likely to be systemi­
cally unwell. Langerhans cell histiocytosis may h ave a sim­
ilar presentatio n but the bone lesions are unlikely to be
symmetrical and generalized.
Meta physeal fractures are well recognized in scurvy,
Menke syndrome and metabolic disease of the newborn.
Scurvy has typical radiological appeara nces, with osteopenia,
blurred metaphyses and poor trabeculation. Clinical appea r­
ances and biochemical investigations will clarify the diagno­
sis in most cases.
Copper deficiency is rare, but well described. 108 Prema­
turity, mal nutrition, malabsorption and dietary defic iency
are predisposing factors. Copper has been added to modern
milk formulae and parenteral feeds to prevent its defi­
ciency in preterm infants. Fractures are seen in copp er defi­
ciency only in the presence of widespread symmetrical
skeletal chan ges, showing periosteal reaction, metaphyseal
spurs and cupping and fraying of the metaphyses, visible
especially at the wrists, knees and costochondral junctions.
Biochemical investigations will confirm this disorder.
Rickets is characterized by metaphyseal fraying and cup­
ping, with osteopenia and poor trabeculation. Fractures may
occur, but only in the presence of obvious bony abnormal­
ity, confirmed by biochemical abnorm alities (Fig. 3.43).
Congenital syp hilis can be confirmed by clinical
features and serology. Metaph yseal in'egularity is seen, but
the appearances are those of undermining of the metaphysis.
Osteogenesis Imperfecta
Osteogenesis imperfecta is a disorder of synthesis of type I
collagen, with resulting bone and connective tissue abnor­
malities. It is an inherited condition, with four major types
described by Sillence. 109
 72 I Radiology of child abuse

Figure 3.43 Immigrant child with nutritional rickets and severe osteopenia. (a) Left lower limb shows healing fractures of the tibia and
fibula and a fresh corner fracture of the proximal tibia. (b) Upper limb shows a healing fracture of the radius and abnormal periosteal new
bone extending to the humeral metaphysis. Typical metaphyseal fraying and cupping best seen in the distal ulna. The faint metaphyseal
sclerosis is due to healing, following treatment. A diagnosis of non-accidental injury was not made because of the rickets, although there
were grave concerns about the family.

• Type I is the most prevalent form, with blue sclerae, a
family histolY of deafness, and sometimes abnormal
dental development. Ten per cent of people with
osteogenesis imperfecta show fractures at birth. This is
an autosomal dominant condition but new mutations
do occur.
• Type II can be recognized in utero and is usually lethal
in the fetal or perinatal period. It is thought to be
transmitted as an autosomal recessive condition.
• Type III is a rare autosomal recessive form of
osteogenesis imperfecta, resulting in deformity of the
limbs, spine and skull. Fractures are often present at
birth.
• Type IV is regarded as a rare form of osteogenesis
imperfecta. The sclerae are normal in appearance. The
bone disease is very variable in severity. This group
may be subdivided, according to the presence of
dentinogenesis imperfecta. In a mild case, particularly
because of the normal sclerae, it may be difficult to
distinguish osteogenesis type IV from non-accidental
injury.
It should be remembered that children with osteogenesis
imperfecta susta in fractures similar to fractures of other
children from accidental trauma, just with a lesser degree of
force. They w ill therefore have fractures that are similar in
distribution to those in accidental trauma, and would not be
expected to show metaphyseal corner fractures. Simila rly,
rib fractures and skull fractures are not commonly seen in
osteogenesis imperfecta. Because fractures are painful and
cause the child distress, there should not be any significant
delay in presentation to medical attention; therefore, previ­
ously unidentified fractures in various stages of healing
would not be seen. Subdural haemorrhage is a rare feature
of osteogenesis imperfecta, recognized in severe cases, in
which the diagnosis is not in doubt. 110 Brain injury is not a
feature of osteogenes is imperfecta.

.. - -- - ­

There are rep orts of suspected non-accidental injury
that have later been proven to be osteogenesis imperfecta .
Review of the described histories would sugges t that fea­
tures of osteogenesis imperfecta or clinical features
unusual in non- accidental injury were overlooked.IIL.112
Temporary Brittle Bone Disease
There have been two hypotheses proposed for conditions
that wo uld render a child more susceptible to fractures for
a short period of time, with spontan eo us resolution of the
abno rmali ty and with no biochemical or pathological
abnormalities at the time of diagnosis. Paterson et al sug­
ges t tha t this is caused by a temporary copper deficiency,
resulting in a temporary bony fragility w hich does not
resolve until the copper levels have returned to normal. liJ
His cohort included patients with rib, metaphyseal and
skuJl fractures. Although metaphyseal fractures are recog­
nized in established copper deficiency, bone changes are
acco mpanied by abnormal biochemistry and by anaemia:
skull fractures are not seen. 108 A review by Chapman and
Hall has co nsidered Paterson's proposed condition in
detail. I14 Another group suggest that reduced bone density
due to lack of fetal intrauterine movement could cause
abnormally low bone density and a higher propensity to
fracture. 115 The validity of this data is questioned 115 and
contradicted by other authors. 117
CONCLUSION
It is the du ty of the paediatric radiologist to be an advocate
for the child, who should be protected fro m abuse. Natural
conditions which could mimic aspects of non-accidenta l
injury, or underlying conditions which could predispose to
fra cture or subdural haemorrhage, must be identified where
present. It must be remembered that, although rarely, such
a child may also be abused. IIB.119
A mistaken diagnosis of child abuse, when the appear­
ances have been caused by another disorder, can be devastat­
ing for the child and family. However, if we fail to recognize
inflicted injury, the child may be the victim of further assault,
which may result in life-long disability, if not death.
REFERENCES
Caffey J. Multiple Fractures in the long bones of infants
suffering from subdural haematoma. A]R 1946; 56:163-73.
2 Guthkelch AN. Infanti le s ubdu ra l haematoma a nd its
relationship to whiplash injuries. EM] 1971; 2:430-1.
3 Caffey J. On the theory and practice of shaking infants. Its
potential residual effec ts of permanent brain damage and
mental retardation. Am] Dis Child 1972; 124:161-9.
4 Kleinman PK. Belan ger PL, KareJlas A, Spevak MR. Normal
metaphyseal radiologic variants not to be confused with
ftndings of inFant abuse. A]R 1991 ; 156:781-3.
References I 73
5 Kleinman PK. Schm id-like metaphyseal chondrodysplasia
Sim ul a tin g child abuse. A]R )991; 156: 576-8.
6 Kleinman PK, Spevak MR. Variations in acromial ossificatio n
Simulating infant abuse in victims of sudden infant death
syndro me. Radiologv 199 1; 180:185- 7.
7 Sbapiro R. Anoma lous parietal sutures and the bipartite
parietal bon e. A m] Ro el1tgenol Radium Ther Nucllvled 1972;
115:569-77.
8 Omah AC, Hall CM. Observational study of skeleta l surveys in
suspected non-accidental. injuly. Clinical Radiology 2003; 58:
70 2-5.
9 James SU, Halliday K, Somers J, Broderick N. A survey of
non-accidental injury im aging in England, Scotland and
Wales. Ciin Radio/2003; 58:696-701.
10 Sane SM et al for American Academy of Pediatrics. Dia gnost ic
imagi ng of child abuse. Pediatrics 2000; 105:1345-8.
11 American College of Radiology. ACR Practice Guidelinefor
Skeletal Surveys in Children (Res. 47, 17,35). Reston, VA: ACR
Standards American College of Radiology, 2001, pp. 145-9.
12 Royal Co llege of Radiolo gists and Royal College of Paediatrics
and Child He alth. Standards for Radiological hll!es tigatiolls of
Su spected Non-accidental h~j!II)!. London: Royal College of
Paediat rics and Child Health, 2008.
13 Omah AC, Moon L, Hall CM, Todd-Pokropek A. Diagnostic
accuracy of fracture detection in suspecte d non- acc idental
injury: the effect of edge enhancement and digital display on
observer performance. Ciill Radiol 2006; 61: 163-73.
14 Zimm erman S, Mako roFF K, Ca re M, et al. Utility of Follow-up
skeleta l surveys in suspected ch ild physical abuse evaluations.
Child Abuse Et Neglecr, 2005; 29: 1075-83.
15 Kleinman PK, Nimkin K, Spevak MR, et a!. Follow-up skeletal
surveys in susp ec ted child abuse. A]R 1996; 167: 893- 6.
16 Anilkumar A, Fender LJ, Broderick NJ, et a1. The role of the
follow-up chest radiograph in suspected non-accidental
injury. Pediatl' Radio12006; 36:216-18.
17 Kleinm an PK, Marks SC, Adams VI, Blackbourne BD. Factors
affecting visualization of posterior rib fractures in abused
infa nts. A]R 1988 ; 150:635-8.
18 Kleinman PK, Marks Sc. Spevak MR, Richmond JM. Fra ctures of
the rib head in abused infants. Radiology 1992; 185: 119-23.
19 Nimkin K, Kleinman PK, Teeger S, Spevak MR. Distal humeral
physeal injuries in chi ld abuse: MR imaging and
ultrasonography findings. Pedlatr Radio/ 1995; 25:562-5.
20 Dias JJ , Lamont AC, Jones JM. Ultrasoni c diagnosis o f
neonatal separatio n of th e distal humeral ep iphysis. ] BOHe
]Oil1t Surg (Er) 1988; 70-B:825-8.
21 Smeets AJ, Robb en SG, Meradji M. Sono grap hically detected
costo-chondra l dislocation in an abused chi ld. Pediatr Radiol
1990; 20:566-7.
22 Jaspan T, Narborough G, Punt JAG, Lowe J. Cerebral
contusional tears as a marker of child abuse - detection by
cran ial sonography. Pediat,. Radiol 1992; 22:237-45.
23 Filiatrault D, Garel L. Commentary: Pediatric blunt abdominal
trauma - to sound or not to so und? Pedia tr Radiol )995;
25:329 - 31.
24 Sivit CJ, Kaufman RA . Commentary: Sonography in the
evaluation of children following blunt trauma: is it to be or
not to be? Pediatr Radio 11995 ; 25: 326-8.
25 Ellison PH, Tsai FY, Largent JA. Computed tomography in
c hild abuse and cerebra l contusion. Pediatrics 197 8;
62:151-4.
26 Zimmerman RA, Silaniuk LT, Bruce D, et al. Computed
tomography of craniocerebral inju ry in the abused chi ld.
Radiology J 979; 130:687 -90.
27 Saulsbury FT and Alford SA. Intracranial bleeding from child
abuse: the value of skull radiographs. Perliatr Radiol 1982;
12: 175 -8.
 74 I Radiology of child abuse
28 Barnes PD, Krasnokutsky M. Im aging of the central nervous
system in suspected or alleged nona ccid ental injury, including
the mimics. Topics MR Imaging 2007 ; 18:53-74.
29 Suh DY, Dav is PC , Hopkins KL, et a1. Nonaccidental pediatric
head injUly: diffusion-weighted imaging findings.
Neurosurgery 2001; 49 :309-20.
30 Hart BL, Dudley MH, Zumwalt RE. Post mortem cra nial MR T
and autopsy correlation in suspected child abuse. Am )
Foren sic NJed Patho/1996; 17: 2 17-24.
31 Sty JR, Starsh a k R1. The role of bon e sci ntigraphy in the
eva luation of the sus pec ted abused child. Radiology 1983;
146:369-7 5.
32 Ma nd elstam SA, Cook D, Fitzgera ld M, Ditchfielcl MR.
Complementary use of radiolog ical skel etal survey and bo ne
scintigraphy in detection of bony injuri es in su spected child
abuse. Arch Dis Child 2003; 88:387-90.
33 Carty HML. Fractures ca used by child abuse. ) Bo ne Joint
Surg Br 1993; 75B:849-57.
34 Rao P, Carty H. Non-accidental injury: review of the
radiology. Clin Radiol 1999; 54: 11-24.
35 Taitz J , Moran K, O'Mea ra M. Long bone fractures in children
under 3 years o f age: is abuse being missed in emergency
department presentatiolls? ) Paediatr Child Health 2004;
40:170-4.
36 Worlock P, Sto wer M, Barbor P. Pattern s of fractures in
accidental and non-acc id ental injulY ill children: a
comparative study. Br Med) Ciin Res Ed 1986; 293:100-2.
37 Kleinman PK, Marks Sc, Richmond 1M, Blackbo urne BD.
lnflicted skeletal injury: a postmortem radiologic­
histopathologic study in 31 infants. AJR 1995; 165: 647-650.
38 Grayev AM, Boal DKB, Wallach DM, Segal LS. Metaphyseal
fractures mimicking abuse durin g treatment for clubfoot.
Pedin tr Radiol 2001; 3 1 :559-63.
39 O'Connell AM, Do noghue VB. Can classic metaphysea l lesions
fol low un co mpli ca ted caesarean section? Pediarr Radiol 2007;
3 7:488-91.
40 Kleinman PK, Marks SC, Bl ackbourne B. The metaphyseal
lesion in a bu sed infants: a radiologic-histopathologic study.
AJR 1986 ; 146:895-905.
41 Bulloch B, Schubert CJ, Brophy PD, et a1. Cause and clinical
character isti cs of rib fr ac tures in infants. Pediatrics 2000;
)05:E48.
42 Cadzow SP, Armstrong KL. Rib fr ac tu res in infants: red al ert!
The clini cal features, investig ations and child protection
ou tco mes. ) Paediarr Child HealrlJ 2000; 36 :322-6.
43 Durani Y, DePiero AD. Images in internal medicine. An n
Em erg Med 2006; 47:210-15.
44 Rizzolo PJ, Coleman PR. Neonatal rib fracture: birth trauma
or ch ild abu se 7 ) Family Practice 1989; 29:561-3.
45 Kleinman PK , Marks SC, Adams VI, Blackbourne BD. Factors
affecting visualization of po sterior rib fractures in abused
infants. A)R 1988; 150:635-8.
46 Kleinman PK, Schlesinger AE. Mechanical fa c tors associated
with posterior rib fractures: laboratolY and case studies.
Pediotr Radiol 1997; 27:87-91.
47 Kleinman PK, Marks Sc, Nimkin K, et a1. Rib fra ctu res in 3 1
abused infants: postmortem radiologic-histopathologic study.
Radiology 1996; 200:807-10.
48 Feldman KW, Brewer DK. Child abuse, cardiopulmonary
resuscitation and rib fra ctures. Pediarrics 1984; 73:339-42.
49 Sp evak MR , Kl ei nm a n PK, Belanger PL, et a1.
Cardiopulmonary resuscitation an d rib fractures in infan ts . A
post- mortem radiologic-pathologic st ud y. )AMA 1994 ;
272:617-18.
50 Dolinak, D. Rib fractures in infants due to cardiopulmon ary
resuscitation efforts. Am J Forensic iVied Parlwl 2007;
28:107-10.
... - -- --
51 Kleinman PK, Marks Sc. Vertebral body fra ctures in child
abuse: radiologic-histopathologic correlates. In vestigative
Radiology 1992; 27:715-2 2.
52 Kleinman PK, Zito JL. Avulsion of the s pinous processes
caused by infant ab use. Radiology 1984; 151:389-91.
53 Rooks VJ , Sisler C, Burton B. Cervical spine injury in child
abuse. Report of 2 cases. Pediatr Radiol 1998; 28:193-5.
54 Cullen Jc. Spinal lesions in battered babies. ) Bone Joint Surg
Br 1975; 57B:364-6.
55 Lonergan GJ , Baker AM, Morey MK, Boos Sc. Child abuse:
radiologic-pathologic correlation. Radiograph ics 2003;
23:811-45.
56 Kogut! MS, Swischuk LE, Fagan CJ. Patterns of injury and
significance of uncommon fractures in th e battered child
syndrome. Am) Roencgenol Radium Ther Nucl Med 1974;
121:143-149.
57 Kleinman PK, Shelton YA. Hangman's fracture in an abused
infant: imaging features. Pedian' Radio11997; 27:776 -7.
58 Nimkin K, Spevak MR, Kleinman PK. Fractures of the hands
a nd feet in child abuse: Imaging and pathologic features.
Radiology 1997; 203 :233- 6.
59 Carty H. Th e non-skeletal injuries of child abuse: part II the
body. In The Year Book of Paediatric Radiology, Borsod
County Teach ing Hosp ital, Miskolc, 1991 ; 3:25-34.
60 Hechter S, Huyer D, Manson D. Sternal fractures as a
manifestation of abusive injury in children . Pediatr Radiol
2002 ; 32 :902-6.
61 Ablin DS, Greenspan A, Reinhart MA. Pelvic injuries in child
abuse. Pediatr Radiol 1922; 22 :4 54-7.
62 O'Connor JF, Cohen 1. In Kleinman PK (ed.) Diagnostic Im aging
of Child Abuse. Baltimore: Wi lli ams and Wilkins, 1987, p. 112.
63 Prosser I, Maguire S, Harrison SK, et al. How old is this
fracture? Radiologic dating of fractures in children: a
systematic review. A JR 200 5; 184: 1282 -6.
64 Haviland J, Ross Ru sse ll RI. Outcome after severe non­
accidelltal head injury. Arch Dis Child 1997; 77:504-7.
65 Jayawant S, Rawlinson A, Gibbon F, et al. Subdural
haemorrhages in infants: population ba sed s tudy. BM) 1998 ;
317:1 558-61.
66 King SJ, Boothroyd AE. Cranial trauma following bi rth in
term illfa nts. Br J Radiol 1998 ; 71 :233-8.
67 Lo TYM, McPhillips M, Minns RA, Gibson RJ. Cerebral atrophy
following shaken impact syndrome and other non-accidental
head injury. Pediatric Rehabilita tion 2003; 6:47-55.
68 Lloyd DA , Carty H, Patterson M, et al. Predictive value of
skull radio g raphy for intracranial injury in children with
blunt head injUly. Lancet 1997; 349:821-4.
69 Billmire ME, Myers PA. Serious head injury in infants:
accident or abuse? Pediatrics 1985; 75:340- 2.
70 Warrington SA, Wright CM , ALSPAC Study Team. Accidents
and reS Ulting injuries in premobile infants: data from tb e
AlSPAC stu dy. Arch Dis Child 200 1; 85: 104-7.
71 Rob erton DM, Barbor P, Hull D. Unusual injUly? Re cent injury
in normal children and children with suspected non­
accidenta l injury. BMJ 1982; 285:1399-401.
72 Helfer RE, SJovis TL, Black M. Injuries resu lting when small
children fallout of bed. Ped iatrics 1977;60:53 3-5.
73 Kravitz H, Driessen G, Gomberg R and Korach A. Accid ental
fa lls from eleva ted surfaces in infants from birth to o ne yea r
of age. Pediatri cs 1969; 44 [Supp1.):869-76.
74 Stoodley N. Neuroima ging in non-accidental head injury: if,
when, why and how 7 Ciin Radial 2005; 60 :22-30.
75 McPhillips M. Non-acc identa l injulY in the yOUllg infant. ln
Rutherford M [ed.) MRJ of the Neonatal Bmin. London: WB
Saunders, 2002 , pp. 261 -71.
76 Newton RW. Inrra cranial haemorrhage and non-accidental
injury. Arch Dis Child 19 89; 64:188-190.
­

77 Whitby EH, Griffiths PO, Rutter S, et al. Frequency and
natural history of subdural haemorrhage in babies and
relation to obstetric factors. Lancet 2004; 363:846-851.
78 Looney CB, Smith JK, Merck LH, et al. Intracranial
hemorrhage in asymptomatic neonates: prevalence on MR
images and relationship to obstetric and neonatal risk factors.
Radiology 2007; 242:535-541.
79 Tung GA, Kumar M, Richardson RC, et al. Comparison of
accidenta l and nonaccidental traumatic head injury in
children o n noncontrast computed tomography. Pediatrics
2006; 118:626 - 33.
80 Duhaime AC, AJario AJ, Lewand er WJ, et al. Head injury in
very young children: mechanisms, injury types and
ophtha lmologic findings in 100 hospitalized patients younger
than 2 years of age. Pediatrics 199 2; 90: 179-85.
81 Ew ing-Cobbs L, Kramer L, Prasad M, et a l. Neuroimaging,
physica l, and developmenta l findings after inflicted and
noninflicted traumatic brain injury in young children.
Pediatrics 1998; 102 :3 00-7.
82 Hy mel KP, Makoroff Kl, Laskey AL, et al. Mechanisms,
clinica l prese ntations, injuries, and outcomes from infli cted
ve rsus noninflicted head trauma during infancy: resu lts of a
prospecti ve, multicente red, co mp arative study. Pedia trics
2007; 119:922-9.
83 Case ME, Graham MA, Handy TC, et aJ. Position paper on
fata l abusive head injuries in infants and yo ung children.
Am J Forens Med Path 2001 ; 22: 112 - 22.
84 We lls RG, Vetter C, La ud P. Traumati c low attenuation
subdural fluid co llec tions in chi ldren younger than 3 years.
Arch Pediatr Adolesc Med 2003; 157: J005 -1 0.
85 McPhillip s M. Initial and sequential MRI in non-accidental
hea d injUiy. In Minns RA, Brown JK (eds.) Shaking and other
Non-a ccidental Head Injuries ill Children. London: MacKeith
Press, 2005, pp. 262-70.
86 Ordia IJ. Strand R, Gilles F, Welch K. Computed tomography
of contusional clefts in the white matter in infants.
J Neurosurg 1981; 54:696-8.
87 Han BK, Towbin RB, de Courten-Myers G, et al. Rev ersa l sign
on CT: Effect of anoxic ischae mi c cereb ra l injury in ch ildren.
Am J Neuroradiol 1989; 10: 1191-8.
88 Kleinman PK, Zito lL, David so n RI, Raptopoulos V. Th e
subarachnoid spaces in ch ildren: normal variations in size.
Radiology 1983; 147:455-7.
89 Kapila A, Trice J, Spies WG, et a!. Enl a rged ce reb rosp inal fluid
spaces in infants with subdural hematomas. Radiology 1982;
142:669-72.
90 Libicher M, Trager J. US meas urements of the subarac hnoid
space in infants: normal values. Radiology 1992; 184:749-51.
91 Wilkinson AG, Cooke R, Tallur KK, et al. Pericereb ra l space
measurements in infants: sonographic determination. Paper
presented at Hydrocephalus 2005 Internati ona l Co nsensus
Meeting, Queenstown, New Zealand, August 2005.
92 Fessell DP, Frankel DA, Wolfson WP. Sonography of
extraaxial fluid in neurolo gica lly normal infants w it h head
circumference greater than or equal to the 95th percentile for
age. J Ultrasound !vIed 2000; 19:443-7.
93 Nimkin K. Teeger S, Wa lla ch MT, et a!. Adrenal hemorrhage in
abused children: imaging and post mortem findings. AJR
1994; 162: 661-3.
94 Ng CS , Hall CM, Shaw DG. The ran ge of v isce ral
manifestations of non-accidental injury. Arch Dis Child 1997 ;
77:167-74.
95 Barnes PM, Norton CM , Dunstan FD , et al. Abdomina l injury
Further Reading
due to chi ld abuse. Lancet 2005; 366:234-5.
96 Touloukian RJ. Abdomina l visceral injuries in battered
ch ildren. Pediatrics 1968 ; 42:642-6.
References I 75
97 Siovis TL, Berdon WE , Haller 10, et at. Pancreatitis and the
battered child syndrome: Report of 2 cases with skeleta l
involvement. Am J Roentgenol Radium Tit er Nucl iVIed 1975;
125:456-61.
98 Sivit Cl, Ingram 10, Taylor GA , et at. Post traumatic adrena l
hemorrhage in children. CT findings in 34 patients. AJR
1992; 158;1299-302.
99 KJeinman PK, Brill PW, Winchester P. Resolving duodenal­
jejunal hematoma in abused chi ldren. Radiology 1986 ;
160:747-50.
100 Fulcher AS, Das Narla L, Brewer WH . Gastric hematoma and
pneumatosis in chi ld abuse. Am J RO(,lltgenol 1990;
155: 1283-4.
101 Ablin OS, Reinhart MA. Esophagea l perforation by a tooth in
ch ild abuse. Pediatr Radial 1992; 22:339-4 1.
102 McDowell HP, Fielding Ow. Traumatic perforation of the
hypopharynx: an unusual form of abuse. Arch Dis CiJild
1984; 59:888-9.
103 KJeinman PK, Spevak MR, Hansen M. Mediastinal
pseudocyst ca used by pharyngeal perforation during ch ild
abuse. Am J Roentgenol 1992; 158:1111-1 3.
104 Ablin OS, Greenspan A, Reinhart MA. Pelvic injuries in child
abuse. Pediatr Radial 1922; 22:454-7.
105 Carty H. Case report; ch ild abuse - necklace calcification - a
sign of strangulation? Br J Radio11993; 66:1186-8.
106 Twomey EL, Naughten ER, Donoghue VB, Ryan S.
Neuroimaging findings in glutaric aciduria type I. Pedia tr
Radio/2003; 33:823-30.
107 Hartley LM, Khwaja OS, Verity CM. Glutaric aciduria type I
and nonaccidental head injury. Pediatrics 2001; 107; 174-5.
108 Shaw JCL. Copper deficiency and non-accidental injury.
Arch Dis Child 1988; 63: 448-455.
109 Sillence D. Osteogenesis imperfecta. An expanding
panorama of variants. Clin Orthop 1981; 159:11-25.
110 Pozzati E, Poppi M, Gaist G. Acute bilateral extradural
hematomas in a case of osteogenesis imperfecta congenital.
Neurosurgel)' 1983; 13:66-8.
III Gahagan S, Rimsza ME. Child abuse or osteogenesis
imperfecta: how can we tell? Pediatrics 1991; 88:987-92.
112 Wardinsky TO, Vizcarrondo FE, Cruz BK. The mistaken
diagnosis of child abuse: a three-year USAF medical center
analysis and literature review. Mil Med 1995; 160:15-20.
113 Paterson CR, Burns J, f\IlcAJlion SJ. Osteogenesis imperfecta:
the distinction from child abuse and the recognition of a
variant form. Am J Med Gellet 1993; 45:J87-92.
114 Chapman S, Hall CM. Non-accidental injury or brittle bones.
Pediatr Radiol 1997; 27: 106-10.
115 Miller MN, Hangartner TN. Temporary brittle bone disease:
association with decreased fetal movement and osteopenia.
Calc Tiss Int, 1999; 64:137-43.
116 Mendels on Kl. Critical review of 'temporary brittle bone
disease'. Pediatr Radiol 2005; 35:1036-40.
117 Rodriguez Jl, Palacios J, Ruiz A. et al. Morphological
cha nges in long bone development in fetal akinesia
deformation sequ ence: An experimenta l study in curarized
rat fetu ses. Teratology 1992; 45:213-21.
118 Knight 01 , Bennet Gc. Non-accidental injury in osteogenesis
imperfecta: a case report. J Pediatr Orthop '1990; 10:542-4.
II 9 Dunca n AA, Chandy J. Ca se report: multiple neonatal
fractures - dietary or deliberate? Ciin Radiol 1993; 48:137-9.
Kleinman PK (ed.) Diagnostic Imaging of Child Abuse, 2nd edn.
Ba ltim ore: Mosby, 1998.

 I CHAPTER 4 I
HAEMATOlOG ICAL ABNORMALITIES
THAT CAN SIMULATE ABUSE
Angela Thomas
Introduction 76
Primary haemostasis 78
Secondary haemostasis 79
Laboratory tests 81
Measurements of primary haemostasis 82
Evaluation of a bleeding patient 82
Patterns of abnormal results 86
Normal coagulation screen with a normal platelet count 90
Abnormalities of platelet number or morphology 94
INTRODUCTION
abuse does not rule out abuse. 7 When the child has bruising
Medical, nursing and other personnel who care fo r children
have a responsibility to be aware of signs and symptoms
that are suggestive of child abuse, including non-accidental
inju ry (NAI) . .Equally, however, they must also recognize
that medical and physical conditions may simu late abuse
and that appro priate measures must be taken to confirm or
rule out these conditions. Cutaneous lesions are by far the
most common presenting manifestations of child abuse, I,2
and of the conditions that may simulate ab use haematol­
ogical abnormalities are manifest usually as bruising or other
bleeding into the skin or mucosal mem branes. Over-reporting
of natural disease as NAI will occur, especially if a full
assessment of the child is not carried out. An incorrect diag­
nosis of ch ild abuse, a lthough at times unavoidable, can be
devastating for the family a nd child 3,4 and, exceptionally,
has led to parental suicide .s Another consequence of such
misdiagnosis is that a condition such as haemophilia will go
unrecognized and therefore untreated. This in itself may
lead to morbidity and even mortality. If the child is revealed
to have a genetic or serious blood diso rd er, regaining the
trust of that patient and their family may be impossible. The
mistaken diagnosis of child abuse has been described as a
form of medical abuse. 6 However, it is important to remem­
ber that diagnosis of a medical condition that can simulate
- -
Coagulation defects 96
The neonate 97
Drugs associated with bleeding 98
Bone marrow failure syndromes 99
Systemic disease associated with a bleeding tendency 100
Activation of coagulation 101
Conclusion 101
References 102
in a recognizable pattern, such as a belt or hand, then sus­
pected abuse must be repo rted regardless of the results of
laboratory tests. B•9 In addition, these children may be at
greater risk of serious injUly secondary to bleeding/,ID as
well as greater risk of bleeding secondary to an abusive
injury.l1 The physi cian who evaluates the bruised child for a
disorder of coagulation sh ould always assess the history
and physical examination to determine whether the bruises
were acqui red spontaneously, accidentally or as the result of
abuse. The pattern of bruising or associated findings may
yield important information. Knight and Bennett J2 describe
a case of abuse in a child with osteogenesis imperfecta,
which was diagnosed from the pattern of the injuries; in this
case a periorbital haematoma in association with a spiral
fracture of the humerus.
Unusual diseases may mimic abuse - misinterpreta tion of
' usual' by the inexperienced, for example a Mongolian blue
spot (see Chapter 8, Fig. 8.7, p. 152), can a lso lead to a mis­
diagnosis. 6 Advice from a paediatric haematologist, paediatric
dermatologist, paediatrician or haematologist may be crucial
in reaching the right diagnosis. Caution must be exercised
when fo llowing guidance (frequently given) that abuse
shou ld be suspected if there is significant bruising or bleed­
ing with no history of trauma or a history inconsistent with
the severity of the injury.13-16 In a ch ild with a bleeding
- -

diathesis, the severity of the bruising may be disproportion­
ate to the injury sustained or there may be a denial of any
significant injury by the parent or caregiver ; a paediatrician
who is not familiar with haematological disorders may mis­
takenly interpret this as evidence of abuse. 17
Clinical identification of the age of a bruise compared
with the age given from the history has been used to assess
the credibility of the history an d to identify whom the child
was with at the time the bruise occurred. However, there is
an increasing amount of evidence to suggest that the age of
a bruise cannot be estimated reliably.18 A bruise is caused
by blood that has escaped from damaged or leaky blood
vessels, usually capillaries, into the interstitia l tiss ues, but
the appearance of the bruise depends on several factors.
The amount of blood that escapes depend s upon the
integrity of the coagulation system, the force of the injury
and the integrity of the vessels. The location of the bruise
also determines its characteristics; for instance, periorbital
and genital bruises will appear sooner than bruises on the
extremities because the tissues are loose and the vessels
poorly supported. 19 The clinical ageing of bruises is most
frequ ently based upon colour of the bruise when compared
with an estab lished chart. 2o However, the colour of a bruise
depends not only on age, but also on amount of blood
present, location beneath skin, skin colour and amb ient
light. 21 Stephenson and Bialas 22 conducted a study on chil­
dren with bruises of known ages; photographs of the bruise
were taken at different time intervals and aged by a 'blind'
observer. They concluded that several different colours
could be present at the same time within anyon e bruise
and that bruises change colour at very different rates, even
when sustained at the same time in the same child. Ageing
of bruises from photographs is much less precise than
many textbooks imply.21 ,22 However, assessing the age of a
bruise on clinical examination may be easier, as other clues
such as tissue swelling or abrasion might be presentY This
was not the finding in a study by Munang and col­
leagues, 2J in which marked variability in colour descript ion
between observers of the same bruises, and thus estimation
of the age of the bruise, was demonstrated; less than one­
third of descriptions between two observers tallied either in
vivo or from clinical photographs. Wide variability in
bruise development and healing urges caution in ageing
bruises and therefore this should never be used as the sole
criterion for child abuse;21 the pattern of distribution of the
bruising is a key factor that must be linked to the child 's
history and stage of development. Bruises found in atypi­
cal areas in toddlers, such as the trunk, hands or buttocks,
are of concern, as are bruises in normal infants under 9
months who are not yet mobile. 24 Sibert and colleagues l8 ,25
have develop ed a scoring system to evaluate these factors
and such tools may be useful for discriminating between
abused and non-abused children.18,25 In a prospective
analysis of bruising in children with a nd without an inher­
ited bleeding disorder it has been shown that non-mobile
babies without a bleeding disorder do not bruise, but once
they start to roll they develop a few bruises that increase in
Introduction I 77
number as they begin to crawl and walk. How ever, children
with severe bleeding disorders did develop bruising when
non-mobile and had larger bruises than those with mild
bleeding disorders, who in turn had larger bruises than
those wit h no bleeding disorder. These observations indi­
cate that children with severe bleeding disorders develop
bruises before they are mobile and bruise more frequently
than a control popul ation. 26 Although the finding of
bruises of different ages coupled with an inconsistent his­
tory has been cited as a hallmark of child abuse,9 not only
is it difficult to age a bruise accurately, but also it is impor­
tant to note that a child w ith a bleeding diathesis may pres­
ent, quite normally, with bruising of differen t ages plus an
inconsistent history.
Haematological investigation of a brui sed child is
mandatory in all cases when the bruising is une xplained or
implausible, and in cases where some explanatio n is given
or found but the bleeding that results is disproportionate to
the injUly sustained. In a child who may poss ibl y have
been ab used, it is essential that investigations are as atrau­
matic as possible and yield the maximum information. In
the case of a bruised or bleeding child under these circum­
stances, whether or not a diagnosis of NAl is enteliained, a
set of screening investigations should be performed on
blood taken from a single ve nepuncture, with the labora­
tory set up and alerted to analyse the tests with the min­
imum of delay, using sma ll plasma samples. Further
invasive tests, such as bleeding time, should usually be
avoid ed at this stage. Initial screening and investigation is
aimed at the diagnosis of the commoner causes of bleeding,
and exclusion or confirmation of some of the rarer causes
for the safety and management of the child. Further inves­
tigations might be needed if no explana tion of the bleeding
is found or no admission of NAJ is made.
In order to be able to interpret the results of coagul ation
testing, the fundamentals of primary and secondary
haemostasis must be understood. Artefact can significantly
distort resul ts and lead to misdiagnosis; utmost care must
be exercised in the way the specimens are taken and han­
dled prior to processing.
Haemostasis depends upon the integrity of the tissues and
vasculature, the n umber and function of the platelets (pri­
mary haemostasisl and the formation of fibrin (secondary
haemostasis). The first of these is outwith the scope of this
chapter and will not be dealt with in any detail, exceptions
to this being Henoch-Schbnlein purpura, which is not
uncommon in children and may be mistaken for NAl, and
vita min C deficiency, which will be discussed briefly. A basic
scheme of the haemostatic mechanism is shown in Figure 4.1.
When a vessel is injured, the subendothelial collagen is
exposed, resulting in platelet adhesion, activation and
aggregation forming the primary haemostatic plug. Vascular
injury also leads to vasoconstriction and slowing of blood
flow, aiding this process. At the same time, tissue factor is
expressed on subendothelial fibroblasts and monocytes ini­
tiating the coagulation process, by which the inactive
plasma coagulation factors are converted to their active
78 I Haematological abnormalities that can simulate abuse
Coll age n Ti ssue
exposure thrombop lastin
Vasoconstriction ....- - ­
1
Platelet
1
Blood
acti vat ion coagulation
~ Th"~bi"
1
Reduced
Plait
blood flow pug
1
Clot
1
Fibrin
Figure 4.1 Haemostasis after injury. Activation of the
coagulation system and platelets is necessary for formation of a
stable fibrin plug.
counterparts by cleavage of peptide bonds. Thrombin is
generated, converting fibrino gen to fibrin , which then sta ­
bilizes the pl atelet plug. Activation offactor XIII by thrombin
allows cross-linking of the fibrin strands and further stabil­
ization. Thrombin also activates the protein C pathway,
which is inhibitory to coagulation and profibrinolysis,
allowing dissolution of the fibrin clot as healing progresses.
PRIMARY HAEMOSTASIS
Platelets must be present in ad equate numbers and function
normally in order for the primary haemostatic plug to form
effectively. Platelets are small, disc- like cells that remain in
the circulation for 7-10 days. In their quiescent state, they
ex press predominantly neutral phospholipids on their outer
membrane, which are haemostatically inert. However, when
activated, negatively charged phospholipids, predominantly
phosphatidyl serine, are exposed, providing the membrane
surface required for the coagulation reactions. On the cell
surface there are also specific protein receptors for platelet
agonists and adhesive glycoproteins, which are essential for
effective haemostasis. Storage gran ules are also contained
within the platelet cytoplasm, ex granules and dense (6)
granules, whose contents are released by fusion of their
membrane with that of the platelet.
Platelet Adhesion
The initial event in haemostasis is the attachment or ad hesion
of platelets to the vascula r subendothelium that is exposed
after injury to the non-thrombogenic vascular endothelial lin­
ing. Collagen receptors on the platelets, thought most likely to
be the glycoprotein (Gp)la-IJa complex, bind to the exposed
subendothelial coliagenY After adhesion, the platelets
undergo further interactions involving the GpIlli-llia receptor
and fibronectin, which result in the spreadi ng of the platelets
over the vascular surface. In areas of low flow or low shear
stress, this may be sufficient to keep the platelets attached .
However, in areas of high shear stress, such as those found in
capillalY beds and the arterial circul ation, further interaction
is required between the GpIb-lX receptor and von Willebrand
factor (VWF), a large, adhesive pl asma protein. Conditions
where either of these receptors is absent, for example Glan­
zmann's thrombasthenia (Gpllb-lIIa) or Bernard-Soulier syn­
drome (GpIb-lX), or where there is a lack of VWF, are
characterized by abnormal bleeding of variable severity and
will be discussed later in the chapter.
Platelet Activation
Platelet activation rap idly foll ows adhesion and results in a
change in shape of the platelets, generation of active medi­
ators an d secretion of the g ranul ar conten ts. The activation
process is hi g hly specific owing to the presence of platelet
recepto rs that recognize appropriate ago nists, the most
important of these being thrombin and ad enosine diphos­
phate (ADP). Thrombin binds to the platelet at several si tes,
including the Gplb-IX receptor as well as a sp ecific throm­
bin receptor, which is the most importa nt for activation.
The thrombin recepto r has been identified as a large seven­
domai n transmembrane protein that binds proteolytically
active thrombin, which is essential for the induction of
platelet activation. This results in a change in shape of the
platelets, platelet aggregation and secretio n of granule
contents. A second agon ist, ADP, is much weaker than
thrombin in vitro but is important in vivo as both red blood
cells and vascular tissues release ADP in response to
damage and lead to ADP-induced platelet activation and
aggregation.
Platelet Aggregation
Once platelets are activated, they aggregate, linking to each
other to form a haemosta tic plug. The binding of fibrino­
gen to the GpUb-lIla complex is essential for bridging
between acUacent platelets and is crucial to platelet plug
formation. Fibrin is also bound by this receptor, helping to
stabilize the platelet plug.
Platelet Secretion
When platelets are activated , in addition to aggregation, a
release reaction occurs in which the contents of the ex gran­
ul es and 6 granules are secreted. The constituents so
released contribute to the process of primary and second­
ary haem ostasis as well as wound repair and vascular
remodelling. The a granules release, amongst other things,
VWF (synthesi zed by the pl ate let-forming megakaryocytes)
and fib rinogen (synthesized by the liver and taken up into
 Secondary haemostasis I 79

the platelet). It is possible that these pro tei ns bind to the
platelet surface and participate in platelet adhesion and
aggregate formation. This may be particularly important in
platelet-rich plugs in which th ere is limited access to
plasma proteins. Factor V is also contained in the ex gran­
ules and is important for formation, along with factor Xa,
of the prothombinase complex on the ex posed phospho­
lipid surface of the activated platel et.
receptor that binds fibrino gen but also VWF, facilitating
aggregation. VWF is non-covalently bound to factor VIIl
and protects it from protein C proteolysis. Low levels ofVWF
result in correspondingly low levels of factor VIII, which, if
severe, can result in defects of secondalY haemostasis.
SECONDARY HAEMOSTASIS

Initiation and Amplification of Coagulation

Platelet Enhancement of Coagulation
Activated platelets accelerate coagulation several thousand­
fold. As well as releasing procoagul ant microparticles an d
factor V, activation of the platelet revea ls anionic phospho­
lipid sites on the platelet surface that are required as cofac­
tors. Two complexes of clotting fac tors, the tenase (X ase)
complex (factors lXa and VIIIa) and the prothombinase
complex (factors Xa and Va) are bound to the platelet sur­
face, thereby increasing their effective concentration and
bringing them into closer proximity. Once bound to the
platelet surface, coagulation factors are protected from
coagulation inhibitors such as antithrom bin and activated
protein C.
Von Willebrand Factor
Von Willebrand factor (VWF) is a multifunctional adhesive
protein that plays an important role in primary haemostasis.
It is secreted from endothelial cells and is essential for stable
platelet adhesion, paliicularly at high shear rates. On release,
it binds to platelet GpIb-IX, allowing adhesion and subse­
quent activation of the platelet. This exposes the GpIIb-I1Ia
Ex posure of tissue factor on damage to the vascular
endothelium results in the binding of both factor VII and
factor VIla, each of which are present in the circulation in
the quiescent state. Once factor VlJ is bound to tissue factor,
it is very rapidly activated by the tissu e factor-factor VIla
complex. This results in the activation of both factor X to
factor Xa and factor IX to fac tor IXa. Some of the factor Xa
thus formed will also bind to and activate the tissue
factor/factor VII complex but, once bound, is rapidly inacti­
vated by tiss ue fac tor pathway inhibitor (TFPI). Unbound
factor Xa stays on the cell surface and activates factor V to
factor Va, which, together wi th factor Xa, generates a small
amount of thrombin (IJa). This thrombin burst is sufficient
to activate platelets, activate fuliher factor V, activate factor
VIII by cleavage from VWF and perhaps activa te plasma
factor XI to factor XIa. The generation of thrombin by this
pathway, however, is insufficient to sustain adequate fibrin
formation because of the ina ctiva tio n of the tissue
factor-factor VIla-factor Xa complex by the binding of
TFPI (Fig. 4.2). Amplification of the coagulation pathway
comes from the sequence of reactions following the gener­
ation of factor IXa and the generation of factor XIa (Fig.
4.3). Factor IXa finds its way to the surface of an activated

x
TFPl
.~
.---'----,-_--,-.L---,--_,...~
1 II
_ ___,_vv_~-~a--,-r---. Xa - Va ~ i XI_Xla

Endothelial cell or
monocyte
TF
Vila
l.I Pl ate let
TG~\NF
"\
Va Villa

IX IXa

Activated
platelet

Figure 4.2 Tissue factor (TF) is exposed after injury and binds both factor VII and activated factor VII (Vila), which both activates factor
IX and genera tes a small amount of thrombin from prothrombin (II ) by activating factor X. The thrombin activates factors XI, Vand VIII by
cleavage from von Willebrand factor (VWF). Platelets are also activa ted and surface phosphol ipid (PL) is exposed. Tissue factor pathway
inhibitor (TFPI) binds the tissue factor-Vlla-Xa complex and inactivates it, switching off the ini tia ting pathway.
 80 I Haematological abnormalities that can simulate abuse
Endothelial cell or
monocyte
Vila
~
IX
Figure 4.3 Amplification of the coagulation system occurs after
the initial thrombin burst when factors V, VIII and XI are activated
along with platelets. Activated factor VII (Vila) by binding to
tissue factor (TF) has already generated a small amount of
activated factor IX (lXa), which binds to the exposed phospholipid
(PL) on the activated platelets along with activated factor VIII
(Villa) forming the 'tenase' complex. This activates free factor X,
which binds to the platelet surface, along with activated factor V,
forming the 'prothrombinase' complex. This converts factor II
(prothrombin) to thrombin sufficient to cause clot formation.
Generation of thrombin is sustained by continued activation of
factor XI by thrombin activating more factor IX in turn.
platelet. where it binds to the exposed anionic phospholipid
in association with its cofactor, factor V1Ila, which, generated
by the initial burst of thrombin, is already bound to the
platelet surface. This 'tenase' complex converts free plasma
factor X to factor Xa. Still bound to the platelet surface, fac­
tor Xa and factor Va, again generated by the initial burst of
thrombin , form the prothrombinase complex, which con­
verts plasma prothrombin to thrombin in amounts suffi­
cient to cause clot formation. Generation of thrombin is
sustained by the activation of factor XI to factor Xla by
thrombin, which activates more plasma factor IX to factor
IXa in addition to the factor IXa initially formed by the tis­
sue factor-factor V1la complex. Thrombin finally converts
fibrinogen to fibrin and through the activation of factor XIlI
promotes cross-linkage of the fibrin to form a stable clot
(Fig. 4.4).29 Thrombin also activates a proenzyme, thrombin
activatable fibrinolysis inhibitor (TAFI) , which downregu­
lates fibrinolysis thus slowing clot lysis.29
Classical Coagulation Pathway
Until recently, the process of coagulation has been described
as a biochemical cascade consisting of an extrinsic and an
- . -
Insoluble
fibrin
XIII • Xilia
Figure 4.4 Conversion of fibrinogen to fibrin by thrombin and
thrombin activation of factor XIII forming an insoluble clot by
cross- lin kage.
intrinsic pathway coinciding at the activation of factor X to
form the final common pathway (Fig. 4.5).30.31 The intrinsic
system is predicated on the assumption that factor XI is acti­
vated by the contact factors factor XII , prekallikrein and
high-molecular-weight kininogens (HMWKs). Factor Xla
then, as described above, activates factor IX in association
with its cofactor, factor V1Ila, which then activates factor X
and so on. In the extrinsic system, it is assumed that the
release of tissue factor and factor V11 by damaged vessels
activates factor X directly. The integrity of these assumed
pathways is measured in vitro using the activated partial
thromboplastin time (aPTI), which is based on contact acti­
vation of factor XII for the intrinsic pathway and the pro­
thrombin time (PT) for the extrinsic pathway. This scheme
sti ll works well for the diagnosis of clinical bleeding prob­
lems, although it is inadequate to explain what happens dur­
ing in vivo haemostasis. For instance, individuals with factor
XII deficiency do not bleed, those with factor XI deficiency
have a mild to moderate bleeding disorder, and those with
factor V11I or IX defiCiency have a severe bleeding disorder.
Patients with factor V11 deficiency have bleeding problems
despite having an intact intrinsic system, but for the most
part only if the level of FV1I is less than 0.10 U/mL. J2 These
anomalies are much better explained by the revised coagu­
lation process depicted in Figs 4.2-4.4.
It can be seen that because in vitro testing is used to elu­
cidate the complex in vivo system, en'ors of interpretation can
occur where prolonged clotting times may not be associated
with a bleeding diathesis. Examples of this are factor XII defi­
ciency and deficiencies of prekallikrein and HMvVKs. They
are not necessary for normal haemostasis in vivo, as they are
required neither for initiation nor amplification of coagula­
tion . The aPTT can also be prolonged if a lupus anticoagulant
is present. Phospholipid is added exogenously to the in vitro
coagulation reaction and can be easily inhibited by antiphos­
pholipid antibodies - the so-called lupus anticoagul ant. In
vivo, no such inhibition occurs because the phopholipid is
provided by the activated platelet and is protected from the
circulating antibody.
 Laboratory tests I 81

Intrinsic pathway Extrinsic path way

XII_Xlla

XI_Xla
aPTI

1
IX-IXa VII_Vila

Vllla~~
PT
Com,," p",hw", X }-::

II - Throm bin - - - ­

Fibrinogen _
1 Fibrin

Figure 4.5 The classical pathway of co agulation: it is proposed that the intrinsic system is activated by cont act with an activating
surface and the extrinsic system by tissue factor released from damaged vessels or tissue. 80th systems activate factor X which via the
final common path way results in the formation of fibrin. The prothrombin time (PT) reflects the acti vity of the extrinsic and common
8ath w ays. The acti vated partia l thromboplastin time (aPTI) is most sensitive to changes in the intrinsic pathway. II, factor II (prothrombin);
Va, activated factor V; Vila, activated factor VII; Villa, activated factor VIII; IXa, activated factor IX; Xa, activated factor X; XI a, activated
factor XI; Xlla, activated factor XII.

LABORATORY TESTS
A full blood count will yield important information regard­
ing platelet number, and some au tomated counters will
give the mean platelet volume. However, there will be no
evaluation of function, an important parameter in primary
haemostasis. Haemoglobin and white cell count may also
indicate an underlying haematological disorder such as
leukaemia or aplastic anaemia. Examination of the blood
film will reveal morphological abnormalities of platelets
such as size (Wiskott-AJdrich, Bern a rd-Souli er, May­
Hegglin, idiopathic thrombocytopenic purpura) or abnormal
granulation (Chediak-Higashi and Hermansky-Pudlak
syndromes). The basic screening tests available for coagu­
lation are the PT, the aPIT (written also as PITK) and
fibrinogen measu rement and/or thrombin time (IT) .
These laboratory investigations are designed to test the
integrity of both the extrinsic and intrinsic pathw ays of the
classical scheme of coagulation (Fig. 4.5) and remain
invaluabl e for the understanding of clot formation in vitro.
The PT reflects the integrity of the extrinsic pathw ay (fac­
t ors V1I and X), as well as factors II (prothrombin) and V of
the common pathw ay. The aPIT tests the intrinsic pathway
and is prolonged by deficiencies of factors XII, XI , IX, V11I
and X as well as prekaJJikrein and HMWKs.
Thrombin time is a functional test of fibrinogen, by which
mrombin is added to test plasma and the time to clotting is
measured. It is affected by the concentration of fibrinogen and
abnormalities of fibrinogen (dysfibrinogenaemia). Inhibitors of
the reaction include fibrinogen- and fibrin-degradation products
(FOPs) and heparin. The clotting time and the appearance of
the clot are informative. A prolonged IT is seen in hypofibrino­
genaemia, both congenital and acquired, as in, for example,
disseminated intravascular coagulation (DIC), in which there is
consumption and inhibition by FOPs and liver disease (raised
FOPs) . A prolonged IT is also seen with hypoalbuminaemia.
Heparin causes extreme prolongation of the IT, but a 'reptilase'
time, which utilizes a thrombin-like enzyme obtained from a
snake venom that is unaffected by heparin and relatively
insensitive to FOPs, is normal. It is useful therefore in differen­
tiating contamination by heparin from fiblinogen deficiency
and also in indicating if a dysfibrinogenaemia is pre-sent when
it is generally more prolonged than the thrombin time. In the
latter case, the nature of the clot is often abnormal , being
transparent and bulky as a result of abnormal fib rin polymer­
ization. This can be seen in liver disease.
Fibrinogen is measured in a variety of ways; some auto­
mated machines derive the value from the prothrombin
time, which is quick and simple. However, PT-derived
fibrinogen levels appear to show a false elevation in a variety
of clinical settings. Conversely, if the PT is very prolonged ,
a fal sely low fibrinogen measurem ent can result. The
Clauss fibrinogen assay is a functional assay, based on the
princip al that the thrombin clotting time is inversely
proportional to the fibrinogen concentration )) and is the
most reliable me thod for general use in clinical laborator­
ies.)4 The use of diluted pl as ma and a relatively high con­
centration of thrombin results in little interference by FOPs
or heparin, which are known to influence the technique. If
a PT-derived method is used, fibrinogen levels of <2.0 gil
82 I Haematological abn ormalitie s that ca n simulate abu se
or >6 .0 gil should be rechecked using the Clauss method. 35
The go ld standard method is measurement of clot weight
but this is time-consuming and not practi ca l as a screening
test. VelY lo w levels of fibrino gen «0,8 giL) wiiJ begin to
prolong both the PT and t he aPTT, especially when auto­
mated analysers are used,
MEASUREMENTS OF PRIMARY HAEMOSTASIS
In paediatrics, defects of prima ry haemostasis are more com­
mon than coagu]opathies. 36 Primary haemostasis is depend ent
upon both number and function of platelets. Other physiolog­
ical variables such as temperature,37 stress,38 anaemia,39 leu­
copenia 40,41 and the integrity of the connective tissues and
vessels also influence primary haemostasis. Many of the va ri­
ables that might cause abnonnal primary haemostasis such as
cardiovascular or renal disease, cardiopulmonary bypass sur­
gery or ingestion of aspirin-like drugs, will be obvious from
the history and examination of the child, or from the results of
initial blood tests such as full blood count or plas ma urea
level. However, platelet function and platelet-vessel wall
interaction are not eluci dated by these methods.
The bleeding time is an in vivo test and dependent upon
both haemostatic and other physiological variables outlined
above. For instance, it is prolonged in patients with the con­
nective tissue disease Ehlers-Danlos syndrome, generalized
vasculitis and scurvy (vitamin C deficiency), patients who
have a low body temperature (as can occur in prolonged sur­
gical procedures) and in those with leukaemia or uraemia.
AJthough the bleeding time can be a useful screening test to
di agnose heredita ry bleeding defects that involve platelet­
endothelium interaction, the test requires a highly motivated
and experienced operator who und erstan ds the many vari­
ables influencing the result and can inform interpretation of
the test. 36 The most widely used method is the Ivy bleeding
time 42 using a template. The test itself is invasive; small inci­
sions of a standard width and depth are made in the forean}]
which can cause scarring and sign ificant distress to the
patient, although there are specially designed templates for
children minimizing th is problem. 43-46 The norma l ranges in
children of different ages vary from those in adults and older
ch ildren and between bleeding devi ces.
Two further methods are available to screen for platelet
dysfunction: both use small volumes of whole blood, from
2 flL to 1.5 mL, a nd require special instrumentation. The first
method uses flow cyto metlY and utilizes the change in
expression of platelet membrane proteins as an indicator of
platelet activation,47 Upon activation, P-selectin is translo­
cated to the platelet surface membrane due to degranulation
of the a granules and there is increased expression of the
GpJlb-IIIa complex. Conversely, as platelets are activated,
the GpIb-IX complex is internalized and results in decreased
expression. These changes in expression ca n be detected
using different monoclonal antibodies both on resting
platelets and on platelets after stimulation by agonists in
controlled conditions.48 Thus platelet activation and reactiv­
ity can be determined. The use of whole blood in the second
method allows for the contribution of red and white blood
cells to the haemostatic process . This method measures
aggrega tion and the release reaction of pl atelets in whole
blood using the PFA-lOO system 49 ,50 Using this system,
whole, anticoagul ated blood is drawn under a constant vac­
uum into a collagen/ADP-coated or collage n/adrena line­
coated membrane, which has a small aperture in it. A
platelet plug is formed and obstructs the fl ow through the
apertu re and both the max imum velocity of flow an d closure
time are recorded. The test can be performed both on resting
platelets and after stimul ation by agonists in controlled con­
ditions. The PFA-lOO is sensitive in detecting classi ca l
defects resulting in maj or platelet dysfunction, such as
Glanzmann's thrombasthenia and Bernard-Soulier syn­
drome and also von Willebran d's disease (VWD), although a
full blood count and film will demonstrate a macrothrombo­
cytopenia in Bernard-Soulier syndrome and abnormal val­
ues of VWF antigen and/or activity will be seen in VWD, if
specifically tested as is recommended in investigation of NAI
(see beloW). False-negative results occur with the PFA-lOO in
milder platelet defects, such as storage pool disorder and
release defects, which are not detectable by t he routine labor­
ato ry tests 50 and it is not sensitive to vascular-col lagen dis­
orders. Although it may be useful as a screening tool for a
bleeding diathesis in children,51 its use in identifying those
who have a bleeding diathesis in cases of possible NAI has
not been tested. Clinicians should always perform a full
range of platelet function tests when clinical susp ic ion is
strong and when exclusion of a platelet defect is essenti al.
If pl atelet dysfunction is suspected from the pattern of
bleeding or needs to be excl uded, formal platelet function
testing can be performed. This requires larger volumes of
blood and arrangements need to be made with the labora­
tory to discuss the extent and range of tests. Screening tests
would include aggregation response to the agonists ADP,
adrenaline (epinephrine), ristocetin and collagen and
5-hydroxytryptamine release, indicating the secretory
resp onse. If abnormal ities are fo und on screeni ng tests, fur­
ther testing could include t hromboxane genera tion, platelet
nucleotides, t10w cytometry or electron microscopy to
define t he disorder more precisely. There are, however,
problems with reproducibility and the overall haemostatic
condition of the patient cannot be fully assessed by this
method as vessel-wall interaction is not taken into account.
It ca n be difficult to determine the contributio n to bleeding
made by minor abnormalities of platel et fun ction testing.
EVALUATION OF A BLEEDING PATIENT
When a child presents with bruising or bleeding, the main
differential diagnoses are physiological or accidental
bleeding. NAI or a bleeding diathesis, The sex of the

patient, age, clinical presentation, past histoty and family
histoty are all important in helping distinguish these dif­
ferent diagnoses, directing investigations and informing
interpretation of the basic screening tests.
Sex is obviously important in determining the likeli­
hood of an X-linked disease such as haemophilia A or B.
These disorders can occur in girls but, unless consanguin­
ity or Turner's syndrome is present, are vety rare. Extreme
lyonization of X inactivation can also result in girls being
affected and, although such diagnoses are unlikely, they
should be tested for full evaluation. The age at presentation
influences the likelihood of a particular cause of a bleeding
diathesis; for instance, a neonate with a purpuric rash will
have a different set of differential diagnoses from an older
child. A patient with a severe congenital bleeding diathesis
is unlikely to present for the first time in the adolescent
years. In addition, the plasma concentration of many of the
coagulation and fibrinolytic proteins is age dependent, so
normal ranges of screening tests are age dependent. It is
important for laboratories to establish normal ranges for
age using their own reagents and methods. 52
Clinical History and Presentation
The type of bleeding at presentation and the cunent history
may indicate the nature of the bleeding problem. If a child
has a bleeding diathesis, there is often a lack of history of
trauma or the bleeding is disproportionate to the injury. For
example, a boy who has bitten his tongue, which then bleeds,
stops and rebleeds on a continuing basis should be investi­
gated for haemophilia or factor XIII deficiency if initial tests
are normal. On the other hand, a child who bleeds intermit­
tently from the same nostril that can be easily controlled,
may well have a local cause, such as a superficial vessel in
Little's area. Active children, particularly as they learn to
walk, often have bruises on their shins but not usually on
non-exposed areas. If such bruising is associated with painful
joints or reluctance to use a limb, haemophilia may be the
cause. Alternatively, if the child is constitutionally unwell,
the underlying disease may be malignancy, with bone mar­
row infiltration such as leukaemia or neuroblastoma, both of
which can cause painful limbs and joints in addition to
thrombocytopenia. If a child is acutely unwell and shocked
with widespread petechial haemorrhages then DIC secondary
to sepsis, particularly meningococcal, should be considered.
Persistent mucocutaneous bleeding such as gum bleeding,
epistaxis (often bilateral) or heavy menstrual bleeding indi­
cates a platelet disorder, either of number or function, or
VWD. Drug histoty is also important, in particular whether
the child is on warfarin, heparin, sodium valproate or a non­
steroidal anti-inflammatory drug.
Past history is important in both determining whether
there has been a continued bleeding problem, whether the
problem is of more recent onset and whether the child has
had any significant haemostatic challenges. Congenital
Evaluation of a bleeding patient I 83
Figure 4.6 Haematoma formation 2 hours after a dose of
intramuscular vitamin K in a neonate with haemophilia.
bleeding problems such as severe haemophilia tend to pre­
sent early, either at birth with cephalhaematoma, intracra­
nial haemorrhage or after an intramuscular injection of
vitamin K (Fig. 4.6) or before the first birthday as the child
becomes more active and starts to crawl and toddle. Con­
genital thrombocytopenia presents early with bruising and
petechial haemorrhages in the neonatal period, although in
some conditions the initial platelet count is preserved but
deteriorates with time. Significant haemostatic challenges
include surgery, for example circumcision, tonsillectomy or
removal of teeth. Bleeding response to injuty such as biting
the tongue, sustaining a fracture or involvement in a road
traffic accident can yield important information. In girls,
increased menstrual loss may be an indicator of a bleeding
diathesis. Concomitant disease such as hepatocellular dys­
function, renal disease, malabsorption or a connective tis­
sue disease such as Ehlers-Danlos syndrome may result in
abnormal bleeding.] ·53.54 Most of the clotting proteins are
synthesized by hepatocytes, and the vitamin K-dependent
factors II (prothrombin), V1I, IX and X) are the most sensi­
tive to liver dysfunction ; however, with increasing damage
other factors such as factor V and fibrinogen are affected.
The vitamin K-dependent factors may also become signifi­
cantly depleted in malabsorption syndromes. Renal failure
and associated uraemia can contribute to a bleeding
diathesis as certain accumulating metabolites interfere
with platelet function and accompanying anaemia results
in the loss of red cells transporting centrally flowing
platelets to the vessel wall, promoting adhesion to the
subendothelium. 55 Paradoxically, nephrotic syndrome is
associated with venous thrombosis due to an imbalance of
the haemostatic system. Acute phase reactions or other
mechanisms cause high VWF and fibrinogen in conjunc­
tion with low levels of antithrombin, an endogenous anti­
coagulant, secondaty to urinaty loss. Cyanotic congenital
84 I Haematological abnormalities that can simulate abuse

heart disease can lead to thrombocytopenia, secondary to

shortened platelet survival, and hypofibrinogenaemia, sec­
ondary to poor liver function, with reduced synthesis and
clearance of clotting factor intermediates leading to low­
grade disseminated intravascular dissemination.
Family history is clearly essential to detennine the like­
lihood of an inherited disease. Children with X-linked dis­
ease such as haemophilia may give a family history of
affected males carried through the female line. However,
the family history is not always helpful as about 30 per cent
of those with haemophilia A are due to spontaneous
mutations. 56 VWD is an autosomal dominant condition
with variable penetrance, which often gives a positive fam­
ily history of bleeding but sometimes appears to skip a
generation owing to VWD modifier genes that have no
association with the VWF gene. 57 Autosomal recessive
conditions, such as factor XIII deficiency, or conditions in
which the heterozygote may only have a mild or no bleed­
ing tendency such as factor XI deficiency, usually do not
have a positive family history unless there is consanguin­
ity within the family. For example, factor XI deficiency is
found most frequently in those of Ashkenazi Jewish
descent (in whom the carrier rate is 8 per cent) and who
tend to many within a small community, resulting in a
restricted gene pool. 58

Examination of the Child Figure 4.7 Fingertip bruising in a child with Glanzmann's
thrombasthenia.
The general health and state of the child should be assessed.
In a bruised child, particular points to look for are the distri­
bution, size and age of the bruises; whether they are in any
recognizable pattern such as a hand or belt mark; and
whether there is a haematoma indicating the extent of bleed­
ing. It should be noted that fingertip bruising is not infre­
quently found in children with a bleeding diathesis and is not
pathognomonic ofNAl (Fig. 4.7). Distribution of bruising can
be the key to the diagnosis of Henoch-Schonlein purpura,
which is present on extensor surfaces and is due to a vasculi­
tis rather than a coagulopathy. Coagulation screening and a
full blood count will be nonnal and thus confusion with NAl
may arise (Figs 4.8 and 4.9).16.59.60 Presence or absence of
petechial haemorrhages will help to differentiate disorders
associated with thrombocytopenia. Such haemorrhages also
occur in situations when there is raised intrathoracic pressure
in the absence of a bleeding diathesis, such as a severe bout Figure 4.8 Symmetrical bruising on extensor surfaces seen in
of coughing or vomiting or in strangulation. In these situ­ Henoch-Schbnlein purpura.
ations, the petechiae will be seen in the distribution of the
superior vena cava area or around the eyes or mouth 61 (Fig.
4.10). A swollen, tender joint may indicate bleeding into the enlarged lymph nodes, liver or spleen usually accompany
joint, as is seen in haemophilia or, much more rarely, VWD, bruising and petechial haemorrhages secondary to acute
but tender joints may also be seen in Henoch-Schonlein pur­ leukaemia . A limp, with or without bruising, can also be seen
pura, acute leukaemia or neuroblastoma. Neuroblastoma may at presentation in this disease. A rare disease, haemophago­
present with bilateral black eyes as a result of tumour infiltra­ cytic Iymphohistiocytosis (HlH) presents with systemic signs
tion; comparatively minor injury can cause the same in and symptoms, such as pancytopenia, coagulopathy and
haemophilia (Figs 4.11 and 4.12). Systemic upset and organomegaly, but may also present with central nervous

Figure 4.9 Buttock bruising in Henoch-Schonlein purpura .
Figure 4.10 Subconjunctival haemorrhage secondary to
coughing in a child with no haemostatic abnormality.
system (eNS) manifestations. These can range from irritabil­
ity to encephalopathy and coma. Retinal and intracranial
haemorrhages can be found in these circumstances and this
condition has been mistaken for abuse. 62 Full examination
and laboratory evaluation, however, will indicate the correct
diagnosis. Disseminated intravascular coagulation with
secondary bleeding is generally seen in an ill child with signs
and symptoms of infection.
- -- - - - - -
Evaluation of a bleeding patient I 85
Figure 4.11 Bilateral black eyes after minor trauma in a chi ld
with haemophilia.
Fig ure 4.12 Black eye and mild proptosis in a child with
neuroblastoma. (Courtesy of the late Dr J. Pritchard, Edinburgh.)
Haematological Investigation
FIRST-LINE INVESTIGATIONS
Once history-taking and examination have been done, inves­
tigations can proceed. The way blood is taken and processed is
of paramount importance, as activation of the coagulation
system by a difficult venepuncture and contamination with
tissue fluids or by the presence of air bubbles will invalidate
the result. Shortening of the PT or aPIT through activation of
a specimen may mask a significant bleeding disorder such as
86 I Haematological abnormalities that can simulate abuse

Table 4.1 Common pitfalls in specimen collection and processing

, Problem Process Result

Poor ve nepuncture technique Activation of sample Prolongation or shortening of clotting times

Specimen taken from heparinized line Heparin contamination
Over- or underfilling specimen tube Ratio of citrate-plasma not 1:9
Severe polycythaemia (reduced plasma vo lume) Ratio of citrate-plasma> 1:9
Inappropriate storage/transport Activation of sample
Loss of factor activity
Thrombocytopenia
Prolongation of apn
Sho rtened or prolonged clotting times
Prolonged clotting times
Shortened or prolonged clotting times
Prolonged clotting times

haemophilia. Conversely, if blood is taken through a cannula
that has been kept patent with heparin then contamination
frequently results, giving abnormal values for many of the
coagulation tests. Venepuncture of children should be done by
an experienced operator wherever possible, at a time when the
blood can be rapidly dealt with by the laboratOly. Blood
should be taken with the minimum of venous stasis and sub­
sequent handling. Spinning, freezing and thawing blood can
cause al1efact; for example, factor XI levels may rise.6) Trans­
port to a laboratory some distance away, even if the specimen
is transported on ice, can also cause artefact, usually a lower­
ing of clotting protein values. Some common pitfalls in speci­
men collection and processing are shown in Table 4. I. Initial
tests include a coagulation screen, PT, aPIT, IT and fibrinogen
as outlined above and a full blood count and film. A fa ctor
VlU, factor IX level and VWF antigen and activity are recom­
mended in all cases of suspected NAI, as a normal or margin­
ally prolonged aPIT can be associated with a significant
decrease in factor VIII or IX levels.64 Factor XIll may also be
measured in neonates with intracranial haemonhage. If the
blood is flowing well , a few millilitres extra can save a second
venepuncture if an abnormality is found, for example a pro­
longed aPTT, and further testing is required. Bleeding time is
an invasive test and although it will demonstrate the integrity
of the platelet-vessel wall interaction, is usually unnecessary
in the early stages of investigation. It is highly operator
dependent. One of the newer tests (flow cytometry or the PFA­
100) may be helpful in the future but there is no published
work in this clinical setting and false-negative results may be
misleading.
It is sometimes helpful to investigate parents, especially
if results are equivocal or subsequent testing for clarifica­
tion of a ny abnormality requires large volumes of blood.
Identification of a child's natural parents is not always cer­
tain, or they may not be available or readily located, mak­
ing this approach impossible.
PATIERNS OF ABNORMAL RESULTS
The pattern of abnormalities obtained using first-line tests
along with the clinical presentation may indica te an und er­
lying defect/disorder. The extended testing discussed
above, which includes VWF antigen and activity, and
levels of factors VIII and IX , will add further information
a nd minimize del ay. The results of the tests will inform the
choice of fu11her tests to determine the precise diag nosis.
Isolated and Prolonged aPTI (PT Normal,
TI Normal, Fibrinogen Normal, Platelets
Normal)
Prolongation of the aPIT is probably the commonest abnor­
mality found on pelforming the coagulation screen. This may
be by only a second or two but can signal a significant bleed­
ing diathesis. Referring to the classical scheme of coagulation
(Fig. 4.5, p. 81), which is a good model for in vitro coagul a­
tion, this indicates a defect in the intrinsic pathway.
This pattern is fo und in:
1. con genital deficiency of factors VIII, IX, XI and XII,
as well as prekallikrein and HMWK ;
2. VWD, as this may result in low levels of factor VIII;
3. circulating inhibitors, e.g. lupus anticoagulant;
4. mild deficiencies of factors II (prothrombin) , V and X
may prolong the aPIT, whereas the PT rema ins norm al
(reagent dep endent);
5. heparin, either as a contaminant or a therapeutic
agent, is a common cause of a prolonged aPIT; the IT
is very sensitive to heparin and will be prolonged; a
reptilase time will be normal.
Thus the prolongation of the aPIT is due either to a
clotting factor deficiency or an inhibitor of coagulation. To
differentiate between the two, a 50:50 mix of test plasma
with normal plasma can be performed . If there is a deficiency
there should be a correction of the clotting time to within a
few seconds of normal and at least less tha n 50 per cent of
the difference between the control and test plasmas. If there
is no significant correction then the presence of an inhibitor
is suspected. This might be heparin contamination or therapy,
an acquired inhibitor of a clotting factor (velY rare, especially
in children) or the lupus anticoagulant (Fig. 4.13).
Therapeutic heparin administration or heparin contamin­
ation of the specimen can be ruled out from the history, by
checking that tlle specimen was not taken from a heparinized
 Patterns of abnormal results I 87

PT, aPIT, Fibrinogen, IT

t
... ... ..­
Isolated I PT Isolated I aPIT I PT, I aPn I PT, I aPn, Normal
1 fibrinogen (including platelet count)

. t . i t i
50:50 plasma mix '+ DIC

Likely Unlikely Likely Unlikely Severe liver

Haematological Tissuelvascular
t t i t t t t
1 VII due to: Cong 1 VII ..
Correction
'+
No correction
1 vit K 111, V, X
disease
Afibrinogenaemia
Dysfibrino­
Glanzmann's
thrombasthenia
HSP
1 vitamin C
early warfarin Reagent warfarin lv+vlll
therapy
early 1 vit K
dependent:
. i t liver LA + 111 genaemia Platelet SPD
1 XIII

Ehler's-Danlos
LA' '+ ? heparin disease
1 cx,antiplasmin

early liver mild 111, V, X Likely Unlikely (I IT. Rep time N)

1 PAI-l
disease

t t t
v\AlD 1 XI ? LA'

1 VIII reagent
(I DRVVT)

llx dependent:

hw 111, V, X
t
HM\AlK' ? factor VIII, IX or

PK' XI inhibitor
(DRVVT N)

Figure 4.13 Algorithm for initial investigation and interpretation of results in a child with a possible bleeding diathesis.
II, factor II; V, factor V; VII, factor VII; VIII, factor VIII; IX, factor IX; XI, factor XI; XII, factor XII; XIII, factor XIII; V+VIII, combined factor Vand VIII deficiency;
HMWK, high molecular weight kinonogen; PK, prekallikreth; cong, congenital; LA, lupus anticoagulant; HSP, Henoch- Schbnlein purpura; SPD, storage pool
disorder; DIC, disseminated intravascular coagu lation; vWD, von Willebrand Disease; PAI-l, plaminogen activator inhibitor-l ; H thrombin time; Rep,
Reptilase; DRWT, dilute Russell viper venom time.
'Not associated with bleeding

Table 4.2 Prolonged aPTT:differentiation of an inhibitor from a deficiency

apn 50:50 mix DRVVT Platelet neutralization Reptilase time PT

Inhibitor
Lupus anticoagulant t t t Correction N N
Heparin t t t No correction N N
Factor VIII and IX antibody t t N n/a N N

Deficiency
Factors VIII, IX, XI and XII t Correction N n/a N N
VWF t or N Correction N n/a N N
Factor II sl t or N Correction t No correction N t·
Factors V, X t Correction t No correction N t <

'May be normal if deficiency mild.

aPH partial thromboplastin time; DRWT, dilute Russell's viper venom time; n/a, not applicable; N, normal; PT, prothrombin time; 51, slightly; VWF, von

Willebrand Factor.

Note: Specific patterns of bleeding may be seen with specific diagnoses (Table 4.3).

line or put into a bottle containing lithium heparin before
being transferred to a citra ted one. It is also possible that
heparin may have been given inadvertently or with intent to
cause bleeding.55 If there is uncertainty, a reptilase time can
be performed, which, by activating fibrinogen directly, is nor­
mal in the presence of heparin. If the pattern of correction is
one of an inhibitor, and heparin has been excluded, the most
likely cause is a lupus anticoagulant (Table 4.2).
Lupus anticoagulant is a phospholipid antibody that
interferes with the phospholipid added to the in uitra test­
ing model. It affects the aPTT rather than the PT, as the
phospholipid in the former test is more dilute and usually
more sensitive than that used in the PI. FUliher tests, such
as the dilute Russell's viper venom time (DRVVT) can dif­
ferentiate this from other factor inhibitors. Russell's viper
venom causes clotting by direct activation of factor X. This
is inhibited by the lupus anticoagulant, but is not affected
by deficiencies or inhibitors of factor VIII or factor IX. A
platelet correction test can then be performed, whereby the
DRVVT is repeated but using platelet-associated phospho­
lipid rather than thromboplastin. Platelet-associated phos­
pholipid is protected from the action of the antibody and
the test revelis to normal with a lupus anticoagulant. Defi­
ciencies of factors X, V and II will also prolong the DRVVT
but these will usually give a prolonged PT and will not cor­
rect with platelet-associated phospholipid (Table 4.2).
88 I Haematological abnormalities that can simulate abuse
If the pattern of correction is one of a deficiency FVIII
and FIX should be assay ed if not already done. If these fac­
tors are normal then FXI and XII should be assayed. If the
only defect is a deficiency of FXII then this is not associ­
ated with a bleeding diathesis and will not account for any
haemorrhagic symptoms or signs. 66 Prekallikrein a nd
HMWK deficiency, which are rare, will also prolong the
aPTI but, again, are not associated with a bleeding diath­
esis. Von Willebrand factor antigen and activity shou ld also
be checked. Factor VIII is non-covale ntly bound to VWF,
which protects it from proteolysis in the plasma. Deficien­
cies of VWF may therefore be associated with reduced
levels of factor VIIl, such that the aPTT is prolonged; however,
this is not necessarily seen (Table 4.2).
The pattern of bleeding that is identified on clinical
examination may predict the most likely cause of the
abnormal test result (Table 4.3). Mucocutaneous bleeding
and/or ecchymoses are a common presentation of VWD.
Factor Xl deficiency also presents with this pattern of
bleed ing but is very rare excep t within certain population
groups such as Ashkenazi Jews, for whom the gene fre­
quency is around 8 per cent 67 Factor VIII and factor IX
deficiencies do not necessarily present with a history of
painful or sw oll en joints and, although spontaneous muco­
cutaneous bleeding is rare, bruising is common. As well as
ecchymoses with or without joint swelling and bleed ing,
factor VIII or factor IX deficiencies can present with inter­
mit te nt but persistent bleeding from mucocuta neous injury,
such as a bitten tongue or lip .
Mucocutaneous bleed ing and/or ecchymoses in a child
whose aPTI does not correct, may be secondary to an
acquired inhibitor, such as a factor VIII inhibitor, which
Table 4.3 Patterns of coagulation results and patterns of bleeding: possible diagnoses
Coagulation results Mucocutaneous bleeding
and ecchymoses
Prolonged aPTT WJD'
PT/fibrinogen/platelets normal Factor XI deficiency
Fa ctor VIII/IX inhibitor
Prolonged PT Warfarin ingestion
aPTT/fibrinogen/plate lets normal Early vitamin K deficiency
Early Iiver dysfunction
Prolonged PT and aPTT Over warfarinization
Fibrinogen/platelets normal Severe vitamin K deficiency
Over heparinization
Factor X, Factor V or prothromb in deficiency
Acquired inhibitors
Prolonged PT and aPTT Severe liver dysfunction
Decreased fibrinogen Dysfib ri nag ena emia/afibri nogena em ia
Normal or low platelets Die (including meningococcal sepsis)
'VI/VD subtype 28 is associated with reduced platelets.
APTI, partial thromboplastin time; Ole, disseminated intravascular coagulation; PT, prothrombin time; VWD, von Willebrand's disease;
2N, subtype 2N.
can result in serious bleeding, or the lupus anticoagulant,
which is likely to be an incidental finding and unconnected
with the bleeding and bnlising. The extra blood taken at
ini tial venepuncture ca n be used to carry out further test­
in g such as the DRVVT with platelet correction as described
above. If the pattern is not that of a lupus anticoagulant, a
specific inhibitor such as to factor VIII or very rarely factor
XI should be considered. In these circumstances, when the
plasma is serially diluted, the clotting times shorten , rather
than lengthen , as the inhibitor is diluted out. The aeti ology
of factor VIII inhibitors in children is uncertain but can be
associated with malig nancy or possibly recent use of high­
dose penicilli n 68 .G9 Factor XI inhibitors are very rare and
have been described foll owing viral infections, resolving
within a few weeks 70 ,71 There are only a handful of
reported cases of acquired coagulation factor inhibitors in
children.
Isolated and Prolonged PT (aPTT Normal. TT
Normal. Fibrinogen Normal. Platelets Normal)
Referring again to the classical scheme of coagulation (Fig.
4.5, p. 81), this indicates a defect in the extrinsic pathway,
that is factor VII.
This pattern is found in:
1. factor VII deficiency, either congen ital or secondary to
liver disease or vitamin K deficiency;
2. early warfarin therapy;
3. mild deficienci es of factors 11 (prothrombin), V and X
may prolong the PT, wheras the aPTI remains within
the normal range, depending on the reagents used;
Ecchymoses with or without
joint bleeding
Factor VIII deficiency
Factor IX deficiency
VWD (2N)
Factor VII deficiency
Fa ctor X deficiency
Factor V deficiency
Prothrombin deficiency
Dysfibrinogenaem ia
Afibrinogenaemia


4. lupus anticoagulant (rarely); it can be more sensitive to
reagents used in the PT than those used in the aPTT; a
mixing test will not correct.
Both warfarin and vitamin K deficiency can cause this
picture. Although they affect the carboxylation of all the
vitamin K-dependent coagulation factors (factors ll, VII, IX
and X), factor VII has the shortest half-life and is the most
sensitive to failure of that step. However, with increasing
dosage of warfarin or severity of vitamin K deficiency, the
aPTT wiH become prolonged, due to the involvement of
other factors (Figure 4.13).
Mucocutaneous bleeding and ecchymoses in association
with a prolonged PT suggest warfarin ingestion, early liver
dysfunction or early vitamin K defiCiency (Tabl e 4.3)'
although none is necessarily associated with bleeding. Vita­
min K deficiency is a recognized hazard in the neonate,
especially in those who are excl usively breast fed, and is
corrected by parenteral vitami n K at birth or repeated oral
vitamin K at birth and the early neonatal period. Malab­
sorption syndromes are also associated with vitamin K
deficiency. More severe deficiency ho wever, results in a
prolongation of both the PT and aPTT (see beloW). A trial
dose of oral or intravenous vitamin K can be given, which
shou ld begin to correct abnormalities within 4-12 hours if
deficiency is the cause.
Prothrombin defici ency is rare but usually presents with
mild mucocutaneous bleeding or with post-traumatic
haemorrhage. The aPTT may be slightly prolonged along
with the prolonged PT, but fibrin ogen is normal. Dyspro­
thrombinaemia has also been described with a similar pre­
sentation and prolonged PT.72
Ecchymoses and bleeding into joints is seen with con ­
genital factor VII deficiency, and thus a factor VII assay
can be performed once warfarin ingestion and vitamin K
deficiency have been reasonably excluded and liver func­
tion checked. A trial of vitamin K can be given as above.
Prolonged PT and apn (n Normal, Fibrinogen
Normal, Platelets Normal)
This pattern is found in:
1. vitamin K deficiency, when the PT is usually more
prolonged tha n the aPTT;
2. oral anticoagulants, when, again, the PT is usually
more prolonged than the aPTT;
3. liver disease, giving rise to multiple factor deficiencies;
fibrinogen may be abnormal in severe disease ;
4. deficiencies of factors II, V, X or combined V and VIII;
5. lupus anticoagu lant, with acquired prothrombin
deficiency.
For both of these screening tests to be prolonged either
there are multiple defects in the coagulation pathway or
there is a deficiency in the final common pathway of the
Patterns of abnormal results I 89
classical scheme (Fig. 4.5, p. 81). Inhibitors can also pro­
duce this pattern; in paliicular, over-heparinization can
lead to a markedly long aPTT, but with only a minor pro­
longation of the PI. Mixing experiments using the PT may
be useful if there is no history 0 f anticoagulant use or obvi­
ous cause of vitamin K deficiency, for example parenteral
nutrition or prolonged antibiotic use. Inhibi tors of coagu­
lation factors may be specific, for example to factor V or
factor II (prothrombin). or may affect more than one factor.
However, they all have the characteristic that as the plasma
is serially diluted, the clotting times shorten rather than
lengthen as would be seen in a simple deficiency.
Mucocutaneous bleeding and ecchymoses are seen in
severe vitamin K deficiency, over-warfarinization and
over-heparinization. Severe liver dysfunction is associated
with a reduced fibrinogen level. Acquired inhibitors of
coagulation can cause very severe bruising, muscle bleed­
ing, and gastrointestinal or urogenital bleeding, but are
very rare. Factor X, factor V and prothrombin deficiencies,
which are also rare, can all present with mucocutaneous
bleeding and bruising of varying severity. Factor X defi­
ciency can also present with haemalihroses, as can factor
V and prothrombin deficiency (less commonly). Specific
factor assays will need to be performed to differentiate
between them.
This pattern of coagul ation tests can be seen in associ­
ation with the lupus anticoagulant and acquired prothrom­
bin deficiency. However, sponta neous bleeding is rare and
mixing tests are indicative of lupus anticoagul ant.
Prolonged PT and apn, Long n with Low
Fibrinogen (Platelets Normal)
This pattern is found with:
1. high level s of heparin (with lower levels, TT is long but
fibrinogen is normal);
2. hypo- and afibrinogenaemia; dysfibrinogenaemia;
3. systemic hyperfibrinolysis;
4. some cases of liver disease.
When the fibrinogen level falls below 0.8 giL (measured
by a functional assay), as well as a prolonged TT, the PT, and
to a lesser extent the aPTT, also become prolonged. This
occurs with hypo- and afibrinogenaemi a and with the
majority of dysfibrinogenaemias. The reptilase time, which is
based on clotting induced by release of the fibrinopeptide A
fragmen t from fibrinogen, is also prolonged, especially with
the dysfibrinogena emias. The reptilase time is not affected
by heparin and so is useful in determining whether there is
contamination by this anticoagulant or not. The platelet
count is normal. Bleeding in those with afibrinogenaemia
and fibrinogen levels of less than abo ut 0.5 giL is lifelong
and varies in severity between patients. Ecchymoses, muco­
cutaneous haemorrhage and haemarthroses have all been
described. Patients with dysfibrinogenaemi a are usu ally
90 I Haematological abnormalities that can simu late abuse
asymptomatic with bleeding (when it do es occur) often
limited to epistaxis, menorrhagia and mild-to-moderate post­
traumatic bleeding, including surgely.
Prolonged PT and aPIT, Long IT with Low
Fibrinogen and Platelets
This pattern is found in:
1. disseminated intravascu lar coagulation ;
2. so me cases of severe liver disease.
Children with DIe secondary for instance to infection ,
such as meningococcal septicaemia, are usually clinically
sick, with evidence of in fection in addition to ecchymoses
and mucocutaneous haemorrhage. Th e coagulation scree n
will be deranged bu t the platelets will be low. Sequential
measurements of coagulation and full blood count may
show a continuing fall in both platelet count a nd fibrino­
gen concentration. Evidence of increased fibrin degrad­
ation will support the diagnosis a nd meas urem ent of the
D-dimer level is probably th e most reliable. D-Dimers a re
formed followin g the di gestion of cross-linked fibrin by
plasmi n and are raised in DIe. Fibrinogen degradation
produ cts (FDPs) can also be measured but are not specific
to cross-linked fibrin and can be raised in other conditions.
Severe liver dis eas e may also give a pattern of prolonged
PT, aPTT with a low fibrinogen and lo w pl atelets secondary
to hype rsplenism, a lthough these can be normal. Clinical
exa mination will di fferentiate the problem as well as add­
itional tests of liver fun ction .
Prolonged PT and aPIT, Platelets Low (IT
I\lormal, Fibrinogen Normal)
This pattern is fo und in:
1. mass ive transfu s ion ;
2. some cases of chronic liver disease, for exampl e
cirrhosis.
The clinical findings and liver function tests will clarify
the cause.
NORMAL COAGULATION SCREEN WITH
A NORMAL PLATELET COUNT
Several conditions need to be considered ifthere is Significant
bleeding and screenin g investigations are nom1al. The com­
monest is VvVD but factor XIII deficiency, platelet storage
pool disorder and Glanzmann's thrombasthenia can produce
these results, although very rarely. Disorders of fibrinolysis
such as cxra ntiplasm in deficiency and plasminogen activator
inhibitor-l (PAl- 1) defiCiency may also predispose to bleed­
ing. In these rare instances, clot solubility assays, factor XIII,
PAl-l and Ctrantipl asmin assays or platelet function studies
may be necessa ry. Abnormalities of collagen vascular dis­
eases and vascular integrity such as Henoch-Schonl ein pur­
pura will also have normal investigations (see Chapter 4,
p.94).
Von Willebrand Disease
Von Willebrand disease is a bleeding dis order secondary to
either a quantitative or qualitative abnormality of VVI/F.
Diagnosis hin ges on demonstrating these defects. Von
Willebrand factor is a multimeric, high-molecular-weight
glycoprotein that is synthesized both by endoth elial cells
an d megakaryocytes, and plays an impol1ant paI1 in pri­
mary hae mostasis. It is essential for platelet adhesion to the
vessel wall, which, having been secreted from endothel ial
cells, it initiates by bindi ng to the GpIb-IX receptor. This
ex poses the GpIlb-llla receptor complex to wh ich it binds
a lon g with fibrinogen, facilitating platelet aggregation ,
es pecially in high-shear conditions. It acts as a carrier pro­
tein for circulating FVI llc to wh ich it is non-covalently
bound, protecting it from proteolysis by protein C and thu s
its rapid clearance from the plasma. Decreased levels of
VWF or reduced binding of V\I\fF to factor VIlle are accom­
panied by correspondingly low levels of circulating factor
VIlIc. It can be seen therefore that both qu antitative and
qualitative defe cts of VWF may affect both primary
haemostasis (resulting in mucocutaneous bleeding) and
secondary hae mostasis (resulting in haema to ma formation
a nd , more rarely, hae marthroses). Several types and sub­
types of VWD have been described; however, the id entifi­
cation of the latter requires specialized testing. Di agnosis
of the specific subtype, although importan t for correct
treatment, is not necessary for the purposes of identifying
if a chUd has a propensity to bl eed. The importan ce of
screening for VWD lies in the fa ct that, excluding type 3
VWD, it is the commonest inherited bleeding disorder, with
a prevalence of between 0.1 per cent and 2 per cent 73 ­ 75
and that the coagulation screen and full blood count can be
normal. Levels of VWF vary with environmental factors,
such as stress and exercise. To minimize the risk of misdiag­
nosis, VWF antigen and function must be measured in
samples obtained on at least two occasions with consistent
resu lts. 76.77 However, from a pragmatic point of view, ini­
tial tests will guide the clinician in early management but
should be repeated at a later stage w here possible.
Type 1 V\I\fD is the commonest form of the disease (80
per cent of cases) an d is cha racteri zed by a quantitative
defect. It is inherited as an autosomal dominant but with
marked variability of both phenotypic penetrance an d
exp ressivity. This may be because th ere are VWD modifier
genes unassociated with the VWF locus as well as the VWF
gene S7 .78 A bleeding diathesis secon dary to V\I\fD can usu ally
be id entified from evidence of a fa mily history of bleeding,
a clinical histOIY and la boratory demonstratio n of VWF

deficiency. However, without documentation of all three of
these features, a diagnosis of type 1 VWD can be more dif­
ficult. Type 2 VWD is inherited as an autosomal dominant
condition when there is a qualitative defect in the VWF. For
both type I and type 2 diseases, the clinical picture is one
of mucocutan eous bleeding, most commonly epistaxis,
gum bleeding and bleeding from superficial cuts and
wounds. Petechi ae are rarely a feature, with none reported
in a study of 1257 patients 79 ,Bo but in 11.5 per cent of
patie nts reported in a study by Silwer. BI The haemorrhagic
tendency is velY vatiab le and depends On the type and
severity of the disease. In many patients with type 1 or 2
disease there may be no history of bleeding and caution
should be exercised when attributing bruising in a child
with suspected NAI automatically to VWD on the basis of
laboratory testing.? VWD may induce increased bleeding
wi th trauma, but not precipitate spontaneous bleeding.
Diagnosis of a bleeding diathesis does not exclude NAl.
and the histolY and clinical findings remain important
contributors to determining the cause of bleeding7
The results of the screening tests usually reveal a normal
platelet count, although mild thrombocytopenia may be
found in subtype 2B or platelet-type-pseudo VWD (see
beloW). The PT is normal, whereas the partial thromboplas­
tin time may be mildly prolonged and dependent upon the
plasma level of factor VIllc (Table 4.5). In mild type 1 dis­
ease, t he bleeding time may be normal or prolonged and
thus may not be helpful in the diagnosis and should be
avoided,n However, an in vitro bleeding time using the
platelet function analyser PFA-lOO shows good sensitivity
to the defect in primary haemostasis that occurs in VWD
and may be helpful if the device is available. 50 ,5 1 Tests used
for the primary diagnosis of VWD include a factor VIlIc
level and VWF antigen (VWF:Ag). Factor VIllc half-life is
regulated by VWF and is frequently reduced in VWD,
although factor VJlIc levels are not necessarily reduced. 77
Table 4.4 Causes of bleeding in a well child
Coagulation screen and platelets
Normal
Common VWD
Henoch-Schiinlein purpura
Uncommon Glanzmann's thromba sthe nia
Platele t storage pool disorder
Fa ctor XIII deficiency
PAI-l deficiency
a:2-ant iplasmin deficien cy
ITP, idiopathic th rombocytopenic purpura: PAI-l, plasminogen activator inhibitor-l ; VWD, von Wi llebrand's disease.
Normal coagulation screen with a normal platelet count I 91
VWF:Ag is decreased in type 1 VWO and decreased or nor­
mal in type 2. More than 80 per cent of all patients with
VWD will have an abnormal v'vVF:Ag result BO but, if done
alone, this will miss a proportion of patients with normal
antige n levels but abnorm al function, i.e. type 2 VWD, In
addition, in some cases the levels will be borderline and
may require repeat testing. 77 In those patients for whom
the diagnosis of VWD is strongly suspected or needs to be
excluded, results of several different tests shou ld be
analysed to y ield the maximum information from which to
draw conclus ions. Further tests include VVllF function,
measured using a platelet-based ristocetin cofactor (RiCo F)
and VWF collagen binding assays. These assays improve
the abi lity to detect type 2 variants and more clearly define
type 1 VWD. Other assays including agarose gel elec­
trophoresis of the VWF multimer pattern, further platelet
agglutination studies and VWF-factor VIllc binding assays
are required to subclassify VWD but are not always neces ­
salY to establish the basic diagnosis of VW0 77 (Table 4.5).
An important point to note is that VWF levels vary accord­
ing to blood gro up and also in crease as part of the acute
phas e response. The mean plasma VWF:Ag level is about
30 per cent lower for those with blood group 0 than those
with blood group A, and 39 per cent lower for those of
blood group AB.B2 Whether the ABO group has any effect
on the specific activity ofVWF is unresolved. Variation due
to the acute phase response means that those with VWD
may have intermittently normal levels and , therefore, a
singl e normal level does not exclude the diagnosis. There is
no evidence that venepuncture alone results in raised lev­
els of VWF, but it is very likely that venepuncture per­
formed under stressful circumstances can result in an
increase in factor VIIIc and VWF. 77 Unfortunately, children
are often stressed when blood is taken and, if this occurs,
results shou ld be interpreted with this in mind and the fact
recorded in the notes, As mentioned above, VWF:Ag and
Abnormal
ITP
Ha emophi lia A or B
Vitamin K deficiency
Wa rfarin or heparin
Congenital platelet abnormality
(excluding Glanzmann's thrombasthenia)
Deficiencies of factors II, V, VII, X, XI
Dysfibri nogenaem ia
Afibri nogenaem ia
92 I Haematological abnormalities that can simulate abuse
Table 4.5 Laboratory testing of von Willebrand's disease
VWD apn VIII:c Platelets
Type 1 Nor l' N or t N
Type 2A Nor l' N or t N
Type 2B N or l' N or t t
Type 2N l' t N
Type 2M N N N N
Type 3 l' l' t t N t t
N, normal; RiCoF, ristocetin cofactor; VWD, von Willebrand's disease; VWF:Ag, von Willebrand factor antigen.
function must be measured in samples obtained on at least
two occasions. One subtype, type 2B, is characterized by an
increased affinity of the abnormal VWF for the platelet.
This often results in a mildly reduced platelet count with
large forms seen on the blood film 83 (Table 4.5). The platelet
counterpart to this can be seen in platelet-type-pseudo
VWD, in which the platelet Gplb-IX complex has an
increased affinity for normal VWF and produces the same
clinical and laboratory picture B4 Differentiation between
the two can be made by mixing experiments.
A variant of VWD, type 2N, can also present in this way
with a prolonged aPTT on testing. Factor VIII is unable to
bind to the abnormal VWF and factor VIII levels are reduced
(Table 4.5). The condition can be confused with mild
haemophilia A, particularly as VWF levels and activity can
be normal. However, the differentiation between mild
haemophilia and VWD type 2N is not necessary for the
immediate management of the child . It is inherited as an
autosomal dominant and can be confirmed at a later stage
once family history and investigation has taken place. A fur­
ther variant, type 2M, has a type 2 pattern of response, i.e.
normal VvVF:Ag and reduced activity but, unlike the other
type 2 variants, it has a normal multimer pattern (Table 4.5).
Type 3 VWD has an autosomal recessive inheritance
and is very rare. It results in a severe bleeding disorder with
markedly decreased or undetectable VWF/Ag and activity
and has factor VIII levels in the region of 0.01-0.05 IU/mL.
Acquired von Willebrand syndrome has been described,
although very rare, and has similar laboratory findings to
congenital disease. 8 5 The severity of the bleeding tendency
varies from mild to severe and has been found in associ­
ation with several disease states including systemic lupus
erythematosus 86 and congenital cardiac defects.8o
Platelet Storage Pool Disorders
These are a collection of disorders that are characterized by
failure of secretion of the contents of the platelet granules on
stimulation. This may be due to either absence of the gran­
ule contents (storage pool deficiency) or to failure of the
VWF:Ag RiCoF Further tests
t t Multimers normal
N or t t Multimers abnormal
t N or t Multimers abnormal
Platelet aggregation with low-dose ristocetin
N N Multimers normal
Abnormal factor VIII binding
t Multimers normal
t 1 Virtually absent multimers
secret ion mechanism. They are often associated with rela­
tively mild bleeding states and diagnosis may be delayed
until later in childhood, or even in adult life. Common pre­
sentations include easy bruising and possibly prolonged
bleeding from superficial cuts, epistaxis and menorrhagia.
Screening tests are normal, therefore specific platelet func­
tion testing, including platelet aggregation and secretion, are
required to confirm the diagnosis. Again, however, the diag­
nosis of platelet storage pool disorder does not exclude NAI,
and the history and clinical findings are very important.
Specific syndromes such as Grey platelet, Hermansky­
Pudlak, Chediak-Higashi and Wiskott-AIdrich syndromes,
which are associated with other defects, and specific blood
film appearances in addition to the st orage pool disorder are
described under separate headings.
Deficiencies can occur in either the 0. granules (Grey
platelet syndrome) or the 6 granules or both. Abnormalities of
platelet aggregation may be detected but can be normal in
milder cases. The definitive diagnosis depends on the demon­
stration of the deficiency either biochemically (using tests of
platelet secretion), by electron microscopy or both.
Failure of the secretion mechanism occurs in aspirin
ingestion and cyclooxygenase deficiency. Defective aggre­
gation and secretion is seen to specific agonists. Most
patients with congenital cyclooxygenase deficiency do not
have an accompanying bleeding diathesis, perhaps because
the defect is balanced by impairment of prostaglandin gen­
eration in the vascular endothelial celIs.87
Defects have been shown in platelet testing t o weak
agonists only and described as weak agonist response
defect (WARD) . The clinical significance of this is very
uncertain especially since defective responses to weak agon­
ists such as epinephrine are found in normal platelets and
results should be interpreted with caution.
Factor XIII Deficiency
Factor XIII is responsible for the stabilization of the fibrin
clot by polymerization 88 and also inhibition of fibrino­
lysis by the binding of <:X2-plasmin inhibitor to the fibrin. 89
 Normal coagulation screen with a normal platel et count I 93

Deficiency of factor XIII results in reduced clot stability

and is associated with delayed haemorrhage and poor heal­
ing. A classical presentation is at birth, with delayed sepa­
ration of the umbilica l stump, sometimes taking over 4
weeks to separate, and continued bleedin g from the site. 90
There is delayed and repeated bleeding from superficial
wounds when a clot is seen to form normally at the site of
the wound only to break down 24 hours later with resumed
bleeding. Intracranial haemorrhage is a considerable risk,
often resulting in death. 91 To screen for factor XIII defi­
ciency, solubility of the clot is tested in urea or monochlo­
racetic acid. If soluble, further immunochemical tests can
confirm and quantitate the deficiency.

Antiplasmin Deficiency

The enzyme plasmin digests fibrin clot as healing takes

place and is regulated by Ctrantiplasmin. Defici ency of this
enzyme results in a severe bleeding disorder that can be
inherited. 92 ,93 Patients present with symptoms similar to
factor XIII deficiency, with delayed bleeding after trauma
and mucocutaneous bleeding as well as joint haemor­
rhages. Coagulation screening tests are norm a l but tests
measuring fibrinolytic activity, such as the euglobulin lysis
time, are usually abnormally shortened and C\:2-antiplasmin
levels are reduced.
Figure 4.14 Multiple bruises on the trunk and arms of a child
with Glanzmann's thrombasthenia.
Plasminogen Activator Inhibitor-l Deficiency

Deficiency of PAI-l can cause a hereditary bleedin g dis­

order, with haemorrhage occurring most usually after sur­
gery or trauma. 94 - 96 The coagulation screening tests are
normal with a shortened euglobulin lysis time. Levels of
PAI-I antigen and activity are reduced.

Glanzmann's Thrombasthenia

In Glanzmann's thrombasthenia, the platelets lack an intact

Gpl/b-lIIa complex, the fibrinogen receptor essential for
aggregation. 97 ,9B The clinical features of this condition are
those typical of primary haemostatic bleeding: petechiae,
purpura (Figs 4.14 and 4.1 5) and mucous membrane bleed­
ing, including epistaxis which can be life-threatening, men­
orrhagi a and gastrointestinal haemorrhage. 99 There is also
significant haemorrhage in response to minor trauma (Fig.
4.16) and major trauma is life-threatening. It is inherited as
an autosomal recessive and there is usually no family history
of bleeding but the child may have a history of easy bleeding
and bruising from birth. It cannot be distinguished on clini­
cal grounds from other severe platelet defects or from severe
VWD. Screening tests are normal and the diagnosis can be
made on platelet function tests that show lack of aggregation
to all agonists including ADP, thrombin and collagen. The Figure 4.15 Multiple bruises on the buttocks and thigh s of a
PFA-lOO shows very prolonged closure times with both child w ith Glanzmann's thrombasthen ia.
94 I Haematologica l abnormalities that can simu late abuse
Figure 4.16 Bleeding after a sma ll scratch in a child w ith
Glanzmann's thrombastheni a.
cartridges. Ad hesio n to connective tissue occurs normally.
The bleed ing time is also markedly prolonged.
Several conditions can prese nt with bleeding manifest­
ations when the primary problem is vascular integrity or
connective tissue disease and laboratory scre ening res ul ts
are normal.
Henoch-Schonlein Purpura
This condition causes purpura secondary to an immunoglobu­
lin A (IgA)-mediated vasculitis and has been mistaken for
child ab use. J6 ,59 The screening tests of coagulation are nor­
mal, or the platelet count may even be slightly raised. 16 ,100
The purpura typically begins as an urticatial wheal, which
gradually changes to the colour of a bruise (palpable pur­
pura), occurring on the extensor sutfaces of the body. An
important differential feature fro m child abuse is the sym­
metry of the bruising, particularly when it is extensive (Figs
4.8 and 4.9, pp. 84 and 85).60
Vitamin C Deficiency
Vitamin C deficiency (scurvy) is rare but can be manifes t in
infan ts betw een the ages of 6 months to 2 years who have
a poor dietary intake, especially if milk and fruit juice is
boiled. It has also been rep orted in chil dren with very
restricted diets. IOI ,102 The bleeding tendency is due to loss
of vascular integri ty wi th collagen deficiency, and results
in peri follicular haemorrhage, subperiosteal, orb ital or sub­
dural haemorrhage. Purpl e, spongy swelling of the gums
where teeth have erupted may occur, and there may be
melaena and haematuria. Bone disease is also manifest
with osteoporosis, metaphyseal white lin es, submetaphy­
sea l lucency and subperiosteal haemorrhage. 103 The diag­
nosis is made on clinical and radiographic gro unds, by
assessmen t of vitamin C intake and by measurement of
leucocyte ascorbic acid levels. Treatment with asco rb ic acid
stops the bleeding manifestations.
ABNORMALITIES OF PLATELET NUMBER OR
MORPHOLOGY
Congenital
In severa l congenital platelet disorders the platelets are
not only low, but also morphologically ab normal. In
add itio n, some a re associated with dysfunctional platelets.
In practice, as long as a full blo od count an d film are
examined, the diagnosis of a haematological abnormality
sho uld be easily made. Many of these abnormalities are
very rare, how ever, and precise di ag nosis may require fur ­
ther testing including pl atelet function tests and electron
microscopy.
WlSKOTT-ALDRICH SYNDROME
This co ndition occurs via X-linked recessive inheritance
and therefore is seen almost exclusively in boys. It is char­
acterized by thrombocyto penia with small platelets (Fig.
4. 17) (which a re dysfunctional), eczema an d combined
immunodeficiency.104 Bleeding episodes usually present
with in the first fe w years of li fe and have a patt ern typical
of a primary haemostatic disorder, namely mucocutaneous
bleeding and bruising.
BERNARD-SOULIER SYNDROME
In this condition there is moderate-to-severe thromb o­
cytopenia, with associated pla telet dysfunction due to
absence of the Gpl b receptor lO5 and giant platelets on
examination of the blood film (Fig. 4.18) . The type of
bleeding seen is typ ical of tha t of a primary hae mostatic
defect and can be severe. The haematological picture can
be confused with idiopathic thrombocytopenic purpura
(ITP), in which the low platel et co un t may also be associ­
ated wit h gia nt forms . Howeve r, there is no functional
defect in ITP and the bleeding is genera lly less severe.
HERMANSKY-PUDLAK SYNDROME
This syndrome is associated with a lack of platelet-dense
granules. It is inherited in an autosomal recessive man ner

Figure 4.17
Wiskott-Aldrich syndrome, show ing thrombocytopen ia and a
sma ll platelet.
Figure 4.18
Bernard-Sou lier syndrome, show ing thrombocytope nia and giant
platelets. Neutrophils and lymphocytes are normal.
and is characterized by tyrosinase-positive oculocutaneous
albinism and the accum ulation of ceroid in the tissues. \06,107
Bleeding is usually mild to mod erate.
CHEDIAK-HIGASHI SYNDROME
The most striking morphological abnormality in this syn­
drome is the presence of giant organelles in leucocytes and
Abnormalities of platelet number or morphology I 95
Figure 4.19 The peripheral blood of a patient with
Chediak-Higashi syndrome, show ing a neutrophil with giant
abnormally staining granules.
oth er granule-co ntaining cells (Fig. 4.19). There may be neu­
tropenia and later thrombocytopenia. The pla telets show a
storage pool defect on testing, and bleeding is usually mild to
moderate and less severe than that seen in Hermansky-Pudlak
syndrome. lOS Partial oculocuta neous albinism also occurs.
MAY-HEGGLIN ANOMALY
This is a rare, autosomal dominant disorder in which th rom bo­
cytopeni a and giant platelets are seen. Pale blue-staining
inclusion s, similar to Dohle bodies, are also visible in the
cytoplasm of neutrophils, eosinophi ls, basophils and mono­
cytes (Fi g. 4.20) .1 09 Functional abn ormalities of the platelets
have been demonstrated. llo Bleeding is seen in nearly one­
half of the pa tients and is of a primary, haemostatic nature.
OTHER MACROCYTOPATHIES
There are several other condi ti ons in which macrothro mbo­
cytopenia and leucocyte cytoplasmic inclu sions are a feature
and these include Fech tner's syn drome, with associated renal
failure, and the Sebastian platelet sy ndrome, witho ut renal
failure. 109 Alport's syndrome, in which there is sensorineural
deafness, haematuria, cataracts and progressive renal fail ure,
can be associated in some cases with macrothrombocyto­
penia and platelet dysfunction (Epstein's syndrome) III or
macrothrombocytopenia with norma l pla tel et func tion
(Eckstein's syndrome).lJ 2In these conditions, bleeding can be
mild to moderate bu t can be significant after tra uma. The
progressive renal failure seen in some of these sy nd romes
will also con tribu te to the bleeding di athesis.
GREY PLATELET SYNDROME
In this extremely rare recessive condition the platelet 0:
granu les are red uced or ab se nt and th e pla telets thus
appear agranular on a Wright-Giemsa stain. There is mild
thrombocytopenia and occasi onal large forms are seen. !Os
Bleed ing is usually mild.
96 I Haematological abnormalities that can simulate abuse
Figure 4.20 The peripheral blood of a patient with
May-Hegglin anomaly, showing thrombocytopenia, a giant
platelet and a small pale blue neutrophil inclusion (Dahle bod y).
Acquired
---~
Acquired defects can be either quantitative or functional, the
former being the more common. Acquired thrombocytope­
nia should be easily identified from a full blood count and
film. Further investigations may be appropriate dependent
upon initial results and clinical examination.
IM MUN ETH ROM BOCYTOPENIA
Idiopathic lor immune) thrombocytopenic purpura (lIP) is
often an acute, self-limiting, benign disease in childhood,
presenting most commonly with skin purpura and mucocuta­
neous bleeding. Intracranial haemorrhage is extremely rare
and is often associated with additional risk factors. 1lJ-116 The
child is usually well or may have an accompanying viral
infection and the diagnosis is made on the basis of a normal
clinical examination, apart from the bleeding manifestations.
The presence of petechial haemorrhages accompanying the
bruising is an important clinical finding. A full blood count
shows an isolated thrombocytopenia and large platelet forms
may be seen on the blood film. Wheeler and Hobbs report
three cases of ITP 117 and Harley two cases,B for which, without
a platelet count, the diagnosis was initially missed and a mis­
taken d iagnosis of abuse made.
MYELODYSPLASIA
In this condition platelet numbers are often low and may
be morphologically abnormal. The condition will be dis­
cussed later under 'Bone marrow failure syndromes'.
COAGULATION DEFECTS
Haemophilia A and B
Haemophilia A and haemophilia B are the most common of
the severe inherited bleeding disorders. They are due to a
functional deficiency of coagulation factors vm and IX,
respectively, and are clinically indistinguishable from each
other. As X-linked conditions, the severe forms occur almost
exclusively in males. They are the commonest inherited
bleeding disorders to present neonatally (Smith 1990 11B ) and
90 per cent of those with severe disease wiII have presented
with bleeding by the age of 1 year. Neonatal presentations
include cephalhaematoma, excessive bleeding from a heel
stab or venepuncture, or bleeding at the site of intramuscu­
lar vitamin K (Fig. 4.6, p. 83). Only occasionally does umbil­
ical stump bleeding occur. Intracranial haemorrhage (ICH)
has been reported in neonates with severe haemophilia,
especially, but not only, if there has been a long, difficult
labour or instrumental deliveryY9-1 2J Diagnosis in these cir­
cumstances is crucial, as immediate treatment is required.
After the immediate neonatal period, the baby is unlikely to
bleed unless injury occurs. This may be comparatively triv­
ial, such as a fall from a sofa onto a padded surface, but the
bruising or bleeding that results will be disproportionate to
the injury (Fig. 4.21). The child may present with tongue and
mouth bleeding as he begins to put objects in to his mouth,
especially if he is learning to walk and falls with such an
object in his mouth. As he learns to walk haemarthroses will
occur, resulting in a limp or reluctance to use a limb. Soft tis­
sue bleeding and bruising ca n lead t o the mistaken diagno­
sis of chi ld abuse, particularly as there may be no history of
significant injury and bruises of different ages may be pres­
ent. Although an inherited disease, approximately one-third
of cases arise as a result of a new mutation, 56 a positive his­
tory is only elicited in around 50 per cent of cases. 122 Several
reports describe children who have been misdiagnosed as
victims of NAI before full evaluation of the possible causes
of the bleeding has been undertaken. B,117, 123 In haemophilia
A or B, a coagulation screen will show isolated prolongation
of the aPTT, which will correct on 50:50 mixing with normal
plasma. Specific coagulation assays wi ll identify the defi­
cient factor, and appropriate treatment and advice can then
be instituted. Moderate and mild haemophilia present later.
In the case of mild haemophilia, presentation is usually only
after trauma . The aPTT assay is not very sensitive to mild,
but clinically significant, decreases in factor VIII or IX

Figure 4.21 Haematoma and bilateral black eyes in a chi ld with
undiagnosed haemophilia after a fall from a height of 40 em on to
a eush ion.
levels. In a survey as part of the National External Quality
Assessment Scheme (NEQAS) in the United Kingdom, test
plasma w ith a factor vm level of 0.33 fU/ml was sent for aPlT
testing to 572 laboratories. In total, 605 tests were carried out
and the overall median aPTI ratio was 1.25, with a median
aPTI of 38.1 seconds. 64 Increases in aPTI of only a second
or two should not be ignored.
Haemophilia C
Factor XI deficiency is found mostly but not exclusively
amongst the Ashkenazi Jewish population and is the com­
monest of the rare coagulation deficiencies after haemophil­
ias A and B. 58 •67 It is inherited as an autosomal recessive
co ndition and there is often no positive family history, unless
there is consanguinity within the family. Heterozygotes may
have a mild bleeding tendency and have levels of factor XI of
between 0.15 and 0.70 fU/mi. Homozygotes have levels
below 0.15 IU/mL and generally have more severe bleeding
problems although bleeding tendency is not exclusively
determined by factor XI levels. 67 Bleeding is usually mucocu­
taneous in nature with easy bruising, epistaxis and menor­
rhagia. Haemarthroses are uncommon. Bleeding is manifest
after trauma or surgery. A coagulation screen will show an
isolated, prolonged aPTI with normal factor VllI and factor
IX levels. A factor XI assay will show reduced levels.
Rare Coagulation Deficiencies
These include deficiencies of fibrinogen (factor 1), prothrombin
(factor 11), factor V, factors V + vm, factor VIl and factor X.
They occur with a frequency of 1 in 500000 to 1 in
2 million. In congenital fibrinogen deficiency, levels of
The neonate I 97
0.4-0.5 gIL are haemostatic and in prothrombin deficiency lev­
els need to reach 0.2-0.3 fU/mi. for factor V, factors V + VIII,
factor VIl and factor X haemostatic levels are achieved at
around 0.10-0.15 fU/mL. This should be borne in mind when
attributing bleeding to reduced levels of these factors. 124,125
Deficiencies of Factor XII, High-l\I1olecular­
Weight Kininogen and Prekallikrein
------
Deficiencies of these proteins result in a prolongation of
the aPTI but are not associated with a bleeding tendency.
They are important for in vitro haemostasis but not for
in lJivo haemostasis. 66 .67
THE NEONATE
Special cons ideration needs to be given to interpretation of
haemostatic parameters in the neonate, which differ from
those in older children and adults. As the system matures
in the growing child, so parameters change towards those
of the older child and adult. Age-related normal ranges are
therefore necessary and sequential ranges are required in
the first few weeks and months of life. There are also some
bleeding problems that might manifest themselves in the
neonatal period, either because they are specific to that
period or because they are congenita l and present after a
haemostatic challenge.
Most haemorrhagic conditions presenting in the neonatal
period are accompanied by abnormal screening tests, the
majority of them are acquired. The exceptions to this are
severe platelet function defects in the presence of normal
platelet numbers such as Glanzmann's thrombasthenia and
factor XIII deficiency. Presentation and description of both
these conditions have been outlined above. Dysfibrino­
genaemias rarely present in this period but in general are
detectable by abnormal fibrinogen measurements or throm­
bin time. By contrast, severe hypofibrinogenaemia or afib­
rinogenaemia frequently presents in the neonatal period
with bleeding from the umbilical cord, gastrointesti nal
haemorrhage or cerebra l haemorrhage. In afibrinogenaemia
both PT and aPTI are markedly long. Von Willebrand dis­
ease presents, exceptionally, in the neonatal period, as levels
of VWF are physiologically increased and high-molecular­
weight multimeri c forms, which are more active, are prefer­
entially increased in this period. 126 These physiological
changes result in patients with VWD having adequate levels
for primary haemostasis. Those with type 3 VWD, however,
are at risk of bleeding, as both vv\1f and factor VlII are
markedly reduced and screening tests wi ll be abnorma l.
Thrombocytopenia in the Neonate
This will lead to purpura and, in particular, a petechial rash.
Differentiation from NAl can be made from examination of

98 I Haematological abnormalities that can simulate abuse
the blood count and film. The causes will include the rare
inherited thrombocytopenias described above, immune
causes such as neonatal alloimmune thrombocytopenia
(NAlT) or maternal ITP and non-immune causes. Non­
immune thrombocytopenia is seen in congenital intra­
uterine infection such as that caused by cytomegalovirus
(CMV) or the human immune deficiency virus (HN), and in
neonates who are sick from various causes, including sep­
sis and neonatal asphyxia, congenital aplastic processes as
described above or infiltration of the marrow by leukaemia,
for example. History and examination are important in
determining the precise cause.
Neonatal Alloimmune Thrombocytopenia
Although NAIT classically presents at or soon after birth it
should not be mistaken for NAl. The classical presentation is
one of bleeding, which includes intracranial haemorrhage in
an otherwise well infant. Neonatal alloimmune thrombo­
cytopenia occurs as a result of transplacental passage of
maternal IgG antibody directed against the fetal platelets
owing to expression of a paternal antigen that the mother
does not possess. The commonest cause of this is antibody
directed against the HPA-Ia antigen expressed on the platelets
of about 98 per cent of the population. This is significant
because the antibody, when bound, blocks the GpIfb-lIIa
receptor on the platelet and leads to a platelet function defect
as well as moderate to severe thrombocytopenia. Although
skin and mucocutaneous bleeding are frequent, the most ser­
ious complication is that of intracranial haemorrhage, which
occurs in up to 20 per cent of neonates with this condition;
around one-half of these bleeds occur in utero. The degree of
thrombocytopenia is most marked in the first day of li fe and
platelet numbers gradually rise over the next 2-4 weeks as
the level of the antibody declines in the infant's circulation. It
is essential that the condition is correctly identified so that
appropriate treatment may be given to the child and advice
given to the mother for future pregnancies.
Vitamin K Deficiency
Vitamin K is essential for the carboxylation of coagulation
factors II, VJI, IX and X. Without this step, these factors are
inert and incapable of activation in the coagulation
process. Vitamin K deficiency misdiagnosed as child abuse
has been described in several reports, for example Wheeler
and Hobbs 117; Wetzel et al 127 in the neonatal period (haem­
orrhagic disease of the newborn, HDN) and by Carpentieri
et al 128 and Kaplan 129 later in life. Presentation of HDN can
be in the immediate neonatal period and is usually second­
ary to maternal ingestion of vitamin K antagonists such as
warfarin or anticonvulsants. Classical HDN presents at
between 2 and 5 days after birth, with purpura and gas­
trointestinal haemorrhage in an otherwise well infant.
Mucocutaneous bleeding can occur, but intracranial
haemorrhage is uncommon. Vitamin K prophylaxis wil l
not have been given and the infant is often almost exclu­
sively breast-fed. Factor VJI has the shortest half-life of the
vitamin K-dependent coagulation proteins and therefore a
coagulation screen will initially show a prolongation of the
PT. As factors II, IX and X then fall, so the aPTT prolongs.
A deficient state is demonstrated by correction of the coagula­
tion times after a 50:50 mix of test with normal plasma.
These screening tests are not specific and further investiga­
tion of individual factors and the carboxylation state of
prothrombin will confirm the diagnosis. A therapeutic trial
of vitamin K will shorten the PT significantly in this condi­
tion and may be the most practical approach.
Late HDN occurs after the first week of life and is most
common between weeks 2 and 8 after birth; however, there
have been reports of it occurring as late as 15 weeks after
bilih.1JO A significant proportion of neonates (50 per cent)
presents with intracranial haemorrhage and its associated
morbidity and mortality. A history of exclusive breast­
feeding and either lack of vitamin K at birth or a single oral
dose, is often elucidated. However, it is also seen with other
underlying conditions that res ul t in malabsorption of vita­
min K, such as antibiotic therapy. Laboratory investigation
reveals the same results as for classical HDN.
DRUGS ASSOCIATED WITH BLEEDING
Warfarin
Ingestion of warfarin, whether accidentally or for therapeutic
purposes, will lead to a haemorrhagic di athesis, the extent
depending upon the amount taken. Bruising is the common­
est manifestation but mucocutaneous bleeding, including
haematuria and gastrointestinal bleeding, can also occur. If
investigated, a coagulation screen will show a prolonged PT
with normal or slightly prolonged aPTT and normal platelets.
When a large amount of warfarin has been ingested, the aPTT
will be prolonged, but not to the same extent as the PT. Such
an overdose may be accidental, but it must be remembered
that a child with fabricated or induced illness (FII) may pre­
sent with bleeding due to warfarin poisoningYI.132
Heparin
Heparin must be given parenterally to exert its antithrombotic
effect, which is mediated through antithrombin. Owing to
anti-factor Xa and antithrombin effects, the capacity to gen­
erate thrombin is both decreased and delayed. Heparin can
cause excessive bruiSing but also more serious bleeding,
including CNS and retroperitoneal bleeding. Heparin can also
induce thrombocytopenia. If standard heparin has been given,
a coagulation screen will show both prolonged aPTI and TI.
The reptilase time will be normal. Low-molecular-weight
heparin is most senstively detected using an anti-Xa assay.

Platelet Dysfunction Secondary to Drugs
Many drugs have been associated with platelet dysfunction
in vitro but either do not have a clinically signifi cant effect
in vivo or are used under strict medical supervision;1 33 such
examples include high-dose penicillin and heparin.
Sodium Valproate
The anticonvulsant sodium valproate may also cause
thrombocytopenia and minor coagulation abnormalities
due to an effect on platelet function. Richardson 134
describes a study in which platelet functio n was shown to
be abnormal in 6 out of 23 patients on sodium valproate
but who had normal bleeding times. A previous report had
shown a prolonged bleeding time in four out of five
patients, one of whom was thrombocytopenic. A later
prospective study showed that one-third of children stud­
ied had a fall in platelet count, with the lowest reaching
35 X 10 9 /1. 135 Subsequent studies have confirmed that
sod ium valp roate-associated thrombocytopenia occurs in
12 per cent to 18 per cent of patients , with th e platelet
counts correlating inversely with age of the patient and
dosage of the dru g. 136 - 138 Whether the mechanism is
immune or due to impaired platelet production has not
been clearly e]ucidated . 139 ,\40 There are a few other drugs
that have clinically relevant antithrombotic and haemor­
rhagic effects.
NON-STEROIDAL ANTI-INFLAMMATORY DRUGS
Aspirin is rarely given to children because of the risk of
Reye's syndrome. There are certain specific indications such
as Kawasaki 's disease, for whic h there is a risk of thrombosis
and aspirin is given specifically for its ant ipl atelet effects. 141
Accidental aspirin ingestion or poisoning occurs. Inges tion
of aspirin is the commonest cause of platelet dysfunction in
the population as a whole. Aggregation is impaired due to
deficien t thromboxa ne A2 formation by aspirin's inhibition
of cyclooxygenase. The defect lasts for the lifespan of the
platelet and causes a mild bleeding tendency. It can con­
tribute significantly to the bleeding seen in those with coag­
ulation defects such as haemophilia; maternal asp irin within
10 days of delivery may contribute to neonatal bleeding.14 2
O'Hare and Eden 7 describe a case of child abuse in which
aspirin resulted in abnormal platelet function tests an d led to
severe bleeding from a minor injury. Other anti-inflamma ­
tory drugs such as ibuprofen that might be used in children
are rarely associated w ith bleeding problems.
ANTI-PLATELET DRUGS
Some drugs a re g ive n specifically for their antithrombotic
effects but a re rarely used in ch ildren. They include prosta­
cyclin and prostacyclin analogues, ticlopidine and dipyri­
damole.
- - -- - - -- -
- - -- - --
Bone marrow failure syndromes I 99
BONE MARROW FAILURE SYNDROMES
Inherited Bone Marrow Failure Syndromes
It is outwith the remit of this chapter to discuss bone mar­
row failure syndromes in detail, but Freedman and Doyle 143
have reviewed this topic. When there is significant thro mbo­
cytopenia, presentation with bruising and bleeding is pos­
sible. Many of these conditi ons have accompanying
physical abnorm alities as well as haematological ones. For
example, Fanconi's (aplastic) anaemia is often accompa n­
ied by short stature, pigmentary skin changes and abnor­
malities of the upper limbs. 144 Othe r conditions in this
category includ e Shwach man-Diamond syndrome, dysker­
atosis congenita and thrombocytopenia with absent ra dii
(TAR) . Amegakaryocytic thrombocytopenia presents in
infancy with isolated thrombocytopenia due to red uced or
absent marrow megakaryocytes. An initial diagnosis of ITP
may be made (see below)' but with failure of resoluti on of
the thrombocytopenia within 3 weeks a bone marrow asp ir­
ate is recommended 145 and the diagnosis made. Overall , 45
per cent of patients will go on to develop ap lasti c anaemia.
Aplastic Anaemia
In this condition, a child may present with pallor, overt infec­
tion or bleeding problems typical of a primary haemostatic
defec t. Past history is usually unrem arkab le and physical
examination normal. A fuJ I blood count will show a combina­
tion of anaemia, leucopenia and thrombocytopen ia. The
diagnosis can be confirmed on bone marrow aspirate and
biopsy.
Leukaemia
Leukaemia can present primarily with bleeding problems,
but the clinician is usually ale11ed to the diagnosis as the
child is often unwell and examination may reveal
hepatosplenomegaly and lymphadenopathy. A full blood
count may show a pancytopenia or a naemia a nd thrombo­
cytopenia with a normal or high wh ite cell count. Gener­
ally, blast cells are seen on the blood film and diagnosis
ca n be confirmed by bone marrow examin ation. This con­
dition is rarely misdiagnosed as NAI, although McClain
et a!. 146 describe a case of acute lymphoblastic leukaem ia
diagnosed at autopsy when abuse was suspected.
Myelodysplasia
Acqu ired storage pool defects can acco mpany the thrombo­
cyto penia of myelodysplasia and result in significant
skin purpura and mucocutaneous bleeding (Fig 4.22).133
Diagnosis is genera lly straightforward, with a full blood
-----
--
100 I Haematological abnormalities that can simulate abuse

(al
count showing pancytopenia commonly with a red cell
macrocytosis and teardrop poikilocytes; neutrophils may
show Pelger-Hue! forms. Bone marrow cellularity is normal
or increased, often with increased reticulin, and one or more
cell lines may be dysmorphic. An iron stain to detect ring
sideroblasts and chromosomal analysis is also useful. Early
in the disease process, however, mild thrombocytopenia
alone may be present, with a disproportionate bleeding ten­
dency due to the functional defect. Other investigations such
as fetal haemoglobin (HbF) level and plasma lactate dehy­
drogenase measurement may indicate the correct diagnosis.
Vitamin B12 or folate deficiency must be excluded.

Bone Marrow Infiltration

Malignant conditions with bone marrow infiltration can

result in thrombocytop enia and subsequent haemorrhage.
Neuroblastoma in particular can present with unilateral or
bilateral black eyes due to tumour infiltration (Fig. 4.12, p.
85). Lipid storage diseases, such as Gaucher's disease, can
infiltrate the marrow and cause splenomegaly with subse­
quent hypersplenism. In this condition other coagulation
abnormalities have been reported, including factor IX and
factor Xl deficiencies and abnormalities of platelet func­
tion. 147 - 149 Clinical bleeding is generally mild and in most
cases is that expected for the degree of thrombocytopenia.

(el SYSTEMIC DISEASE ASSOCIATED WITH A

BLEEDING TEI\IDEI\ICY

Although defects of both coagulation and platelet function

are described in systemic disorders, the primary condition
underlying the defect is nearly always manifest and there­
fore consideration of the bleeding defect can be taken into
account if the question of possible child abuse arises.
Johnson 19 points out that children with chronic and stress­
produ cing diseases may be at increased risk of abuse from
their parents or carers.

Renal Failure

Bleeding in children with end-stage renal disease is usually

from mucosal membranes and is related to the level of
uraemia. I SO Qualitative defects of platelet function are well
described and are one of the most important contributors
to impaired haemostasis. Mild thrombocytopenia can also
occur and anaemia is invariabl e; both of these conditions
will contribute to the bleeding tendency.

Figure 4.22 Bruising with no history of trauma in a child with Liver Failure
mild thrombocytopenia and platelet dysfunction due to a
congenital myelodysplasia. The liver is the main site of synthesis for most haemostatic
proteins. An exception to this is factor V1II, which has

signi ficant extrahepa tic synthesis . Liver failure results in
impaired synthesis of coagulation proteins and reduced
clearance of activated haemostatic factors. Thrombo­
cytopenia can occur as well as imp aired platelet function. 49
Owing to the reduction of both procoagulant and anticoag­
ulant proteins and activation of both the coagulation
and fibrinolytic systems, thro mbosis, as well as bleeding,
may occur. Factor VII has the shortest half-life of the coag­
ulation proteins and therefore disappears from the
circulation first. The result is an initial prolongation of the
PT, foll owed by prolongation of the aPTI, and, lastly,
reduced fibrinogen as liver func tion declines. In the
absence of DIC, factor VIII is normal owing to its extrahep­
atic synthesis. Bleeding manifestations include bruising
and petechiae, mucosal haemorrhage and gas trointestinal
haemorrhage.
Malabsorption Syndromes
Malabsorption of fat-soluble vitamins, including vitamin K,
can occur in cystic fibrosis ,128 biliary atresia and obstruc­
tive jaundice. Broad-spectrum antibio tics may alter the
normal intestinal flora and result in malabsorption of
vitamin K. Bleeding manifestations are those of vitamin K
deficiency.
ACTIVATION OF COAGULATION
Disseminated Intravascular Coagulation
Disseminated intravascular coagulation is due to patho­
logical activation of the coagulation system and can be
precipitated by a variety of disease processes. Both endothelial
injury and release of tissue factor can trigger the activation,
resulting in both circulating thrombin and plasmin. Throm­
bin generates fibrin fro m fib rinogen, which is deposi ted in
small vessels and causes microvascular thrombosis. Con­
sumption of coagulation factors occurs during this process,
leading to the classical changes in coagulation parameters
of a prolonged PT, prolonged aPTI, decreased fibrin ogen
and reduced platelets. A bleeding tendency results.
Increased circulating plasmin digests both fibrinogen and
fibrin, resulting in raised FDPs and D-dimers (see p. 91).
These interfere with both fibrin polymerization and platelet
function, thus adding to the bleeding tendency. In children,
DIC is usually acute and occurs in an ill child; it can be sec­
ondary to infection, trau ma (such as crush injuries or burns)
and malignancy (such as leukaemia and li ver disease). A
more chronic, compensated form can occur, with the fol­
lowing features : (1) less marked derangement of the coagu­
lation parameters when the platelet count may be only
moderately reduced; (2) plasma fibrinogen is often normal
or slightly elevated; and (3) the PT and aPTI may be within
normal limits. In such patients, the diagnosis is made on
- - - - -
- ~~
Conclusion I 101
evidence of microangiopathy on the blood film (red
cell fragments and reduced platelets) and increased
levels of FDPs and/or D-dim ers . A chronic localized form of
DIC is seen with giant haem angiomas; this is known as the
Kasabach-Merritt syndrome.
Head Injury
Head injury in children can be associated with coagulatio n
abno rmalities secondary to DIc.152.153 In adults, a coagu ­
lopathy has been associated with delayed brain injury sec­
ondary to bleeding. 154 ,155 The study by Hymel et al 151 of
265 children with head injury showed that those with
parenchymal brain damage were more likely to have a
coagulopathy manifested by prolongation of the PT than
those without. Of the children who had died from abusive
head injury, which had caused parenchymal brain damage,
94 per cent had prolongations of their PT and 63 per cent
had evidence of activated coagulation. Hymel and col­
leagues concluded that it was highly unlikely that the
coagulation abnormalities were due to an underlying
haemorrhagic diathesis. It is important to recognize th is as
a secondary phenomenon , not a primary one, so that inves ­
tigation into possible abuse is not terminated on spurious
gro unds.
Cyanotic Congenital Heart Disease
Abnormalities of coagulation and platelets can occur in
cyanotic congenital heart disease. 156 ,157 The exact mech­
anis m producing the coagulopathy is not known. Hypofib­
rinogenaemia secondary to poor liver function and reduced
synthesis and clearance of clotting factor intermediates can
lead to low-grade disseminated intravascular dissemin­
atio n. Thrombocytopeni a may result from both DIC and
shortened platelet survival.
CONCLUSION
In a child presenting with haemorrhagic symptoms, a his­
tory that includes details of ethnic origin, consangu inity
and family history, drug history and clinical examination is
essential. Initial screening tests should include aPT, aPTI,
TI and fibrinogen, a factor VIII and factor IX level and a
VWF antigen CVWF:Ag) and activity (RiCoF). If the history,
examination and investigations do not account for the
bleeding then further investigation might include other
specific assays, such as factor XIII or platelet function test­
ing. Advice on ful1her testing or interp retation of findings
should be sought from a haematologist who has experience
in this field, especiall y if there is a disparity between the
clinical and laborato ry findings. If the case is likely to
102 I Haematological abnormalities that can simulate abuse
receive a formal legal challenge then even the rarest of
causes may require exclusion. Diagnosis of a bleeding
diathesis, especially if associated with a mild phenotype,
does not exclude NAI and when these are found concur­
rently the child will be at even greater risk.
REFERENCES
Stewart GM, Rosenberg NM. Co nditions mistaken for child
abuse: part II. Pa ediatri c Emergen cy Care 1996; 12:2 17-221.
2 Gallenoh H, Oppenheim W. Th e battered child syndrome
revisited. Clin OrtllOpaed Related Res 1982; 162:11-19.
3 Wexler R. Wounded il1lIocents: the Real Victims of tile War
Against Child Abuse. Buffalo, NY: Pro metheus Books, 1990.
4 Vorenberg E. Di ag nosin g child abuse; th e cost of getting it
wrong. Arch Dermatol 1992 ; 128:844-845.
5 Anh NT. 'Pseudo battered child' syndrome. JAMA 1976;
236:2288.
6 Kirschner RH, Stein RJ. Th e mistaken diagnosis of child
clotting. Science 1964: 145:1310-2.
abuse: a form of medical ahuse? Am J Dis Child 1985 ;
32 Peny OJ. Factor VlI deficiency. Br J Haematol 2002;
139:873-87 5.
118 :689-700.
7 O'Hare AE, Eden OB. Bleeding disorders and non-accidental
injury. Arch Dis Ch ild 1984; 59:860-864.
8 Harley JR. Disorders of coagulation misdiagnosed as
nona cc id enta l b!1lising. Pedialr Emag Care 1997; 13 :347 -349.
9 Maguire S, Mann MK, Sibert J, Kem p A. Are there patterns of
b!1lising in childhood which are diagnostic or suggestive of
abuse? A systematic review. Archs Dis Child 2005; 90:182 -186.
10 Johnson CF, Coury DL. Bruising a nd hemophilia: accid ent or
139X. Norwich: Her Majesty's Stationery Office.
child abuse? Ch ild Abuse and Neglect 1988 ; 12:409-415.
36 Sutor A. The bleeding time in pediatrics. Semin Thromb
11 Smith. S.M, Hanson. R. 134 batte red children: a med ical and
Hemostasis 199 8 ; 24:531-543.
psychological study. British IVIedical Journal 1974 ; 3:666-670.
12 Knight DJ, Bennet Gc. Non-accidental injury in osteogenesis
imperfecta: case report. J Pediatr Orthop 1990 ; 10:542-544.
13 Altieri MF. Child abuse. When to be suspicious and what to
do the n. Postgrad Med 1990; 87:153-162.
14 Ladebauche P. Childhood trauma. When to suspect abuse. Reg
Nul'S J 1997; 60:38-42.
15 Roberts R. When you suspect abuse: what next? The
Practitioner 1997; 241:322-326.
16 Daly KC, Si egel RM. Henoch-Schonlein purpura in a child a t
risk o f abuse. Arch Ped iatr Ado/esc IVIed 1998; 152:96-98.
17 Lapo sata ME, Laposata M. Children with signs of abuse: when
is it not child abuse? Am J c/in ParI1O/2005; 123(Suppl. I):
S 119- I 24.
18 Sibert J. Bruising, coagulation disorder and physical child
abuse. Blood Coagulation Fibrinolysis 2004; 15(Supp\. I):
S33-39.
19 Johnson CF. Inflicted injury versus accidental injury. Pediatr
c/in North Am 1990; 37:791-814.
20 Wilson EF. Estim a tion of the age of cutaneous contusions in
chi1.d abuse. Pediatrics 1977; 60:750-75 2.
21 Schwartz AJ, Ricci R. How accu rately can bruis es be aged in
abused children? Lite rature re v ie w and synthesis. Pediatrics
1996; 97:254-257.
22 Stephenson T, Bialas Y. Estimation of the age of bruising.
Arch Dis Child 1996; 74:53-55.
23 Munan g LA, Leonard PA , Mok JY. Lack of agre e me nt on
colour description between clinicians examining childhood
bruising. J Clin Forensic Med 2002; 9:171-174.
24 Sugar NF, Taylor JA. Feldm a n KW. BIllis es in infants and
toddlers: those who don 't c ruise ra rel y bruise. Puget Sound
Pediatric Resea rch Network. Arch Pediatr Adolesc Med 1999;
153 :399-403.
25 Dunstan FD, Guildea ZE, Kontos K et a\. A sco ring system for
bruise patterns: a tool for identi fying abuse. A rch Dis Child
2002; 86:330-333.
26 Hamilton M, Jenkin s B, Dunstan F et a\. Prospective analysis
of bruising in children with and without an inherited bleeding
di so rder. Blood 2006; 108:304a-305 a.
27 Kunicki TJ, Nugent OJ, Staats SJ et a \. The human fibroblast
cl ass \[ extracellul ar ma trix receptor mediates platelet
adhesion to collagen and is identical to the platelet
glycoprotein la-lia compl ex . J BioI Chern 1988;
263 :4516- 4519.
28 Roberts HR, Monroe OM, Oliver JA et a\. New er concepts of
blood co agulation. Ha ell10ph ilia 1998; 4:331-334.
29 Bazjar L, Morser J, Neisheim M. TAFI, or plasma
procarb oxpeptidas e B, couples the coagulation and
fibrinolytic casca des through the thrombin-thrombomodulin
complex. J Bioi Chem 1996 ; 271: 16603-1 6608.
30 MacFarlane RG. An enzyme cascade in the bl oo d clotting
mechanism and its function as a biochemical amplifier.
Nature 1964; 202 :498-499.
31 Davie EW, Ratnoff 00. Waterfall sequ en ce for intrinsic blood
33 Clauss A. Rapid physiological coagulation methods in
determination of fibrinogen. Acta Haemato/ (Basel) 19 57;
17 :237 -246.
34 Mackie IJ, Kitchen S, Machin SJ, Low e GDO. Guidelines on
fibrinogen assays. Br J Haemato/2003; 121 :396-404.
35 Medical Dev ices Agency. MDA Evaluation Report : fibrinogen
assay reagents and methods. MDA 2000/01 ISBN184J82
37 Valeri CR, MacGregor H, Cassidy G, Tinney R, Pompei F.
Effec ts of temperature on bleeding time and clotting time in
normal male and female volunteers. Crit Care Med 1995;
23 :698-704.
38 van den Burg PJ, Hospers J .E, van Vliet Met a\. Changes in
haemostatic factors and activation produ c ts after exercise in
healthy subjects w ith different ages. Thromb Ha emostas is
1995; 74:1457-1464,
39 Lowe GDO. Blood viscosity and cardiovascular disease.
Th romb Haemostasis 1992; 67 :494-498.
40 Jakob E. Sutor A. Interaction of platelet aggregation and
gra nulocyte function (Abstract). Ann Haemato/ 1991; 62:78.
41 sohngen D, Hattsteill E. Heye ll A et aI. Hematological
parameters in fluencing th e Thrombostat 4000. Semin Thromb
Hemostasis 1995; 2J(Suppl. 2):20-24.
42 Ivy AC , Shapiro F, Melnick P. Th e bleeding tendency in
jaundice. Surg Gynaeco/ Obstet 193 5; 60:781 - 784.
43 Lethagen 5, Kling S. New bleeding time devices with
retractable blades evaluated in children, hea lthy volunteers
and patients with prolonged bleeding time. Thromb
Haemostasis 1993; 70 :595-597.
44 Fe usner JH. No rmal and a bnormal bleed in g times in neo nates
and yo ung children utilizing a full standardized template
technic. Am J Clin Pa t11011980 ; 74:73-77.
45 Andrew M, Castle V, Mitchell L, Paes B. Modified bleeding
time in the infant. Am J Hemato/1989; 30:190-191.
46 Andrew M, Paes B, Bowke r J , Vegh P. Evaluation of an
automated bleeding time device in the newborn.
Am J Hematol 1990; 35:275-277.
47 Michelson AD. Flow cytometlY: a clinical test of platelet
function (Rev iew). Blood 1996; 87 :4925-4936.
48 Michelson AD. Platel et function in the newborn. Semin
Thromb Hemostasis 1998; 24:507-512.

49 Kni:il1er R, Weissbach G, Kuhlisch E. Platelet function tests in
childhood. Measuring aggregation and release reaction in
whole blood. Semin Thromb Hemostasis 1998; 24:513-521.
50 Ha rrison P, Robinson M, Leisner R et al. The PFA-IOO: a potentia l
rapid screening tool for the assessment of platelet dysfi.lnction.
Clinical Labomtmy Ha ematology 2002: 24:225-232.
51 Harriso n P. The role of PFA-IOO® testing in th e in vestigation
and management of hae mo stati c defects in children and
adults. Br] Haematol2005; 130:3-10.
52 Andrew M, Vegh P, J ohnston M et al. Maturation of the
h em atopoie tic system during childho od . Blood 1992; 80:
1998-2005.
53 Lane DA, Scully MF, Thomas DP et al. Acquired
dysfibrinogenae mia in acute and chronic li ver disease.
Br] Haemato/ 1977; 35 :301-308.
54 Weigart Ai, Schafer AJ. Uremic bleeding: pathogenesis and
th era py. Am] !VIed Sci 1998 ; 316 :94- 104.
55 Mannucci PM, Gi angrande PLF. Acquired disorders of
coagulation. In Bl oo m AL, Forbes CD, Thomas DP,
Tuddenh am EG D (eds) Haemostasis and thrombosis. London:
Churchill Livings ton e, 1994, vol. 2, pp. 249-268.
56 Giannelli F, Green PM. The molecular basis o f haemophilia A
and B. Bail/ieres Clin HaematoI1996; 9:21/-228.
57 Casaiia P, Martinez F, Haya S, Espinos C, Aznar JA.
Significant linkage and non-linkage of type I von Willebrand
disease to th e von Willebrand fac to r gene. Br 1 Haematol
2001; 115:692- 700.
58 Bolton-Maggs PHB, Young Wa- Yin, McC raw AH et al.
Inh e ritan ce and bl eedin g in factor XI deficiency.
Br] Ha ematol 1988; 69 :521-528.
59 Brown J, Melinkovich P. Henoch-Schonlein purpura
mi sd iagnosed as suspected child abuse: a case report a nd
literature review. ]AtVIA 1986; 256:617-618.
60 Menter A. Symmetrical purpura doesn't mirror ch ild abuse.
lAMA 1987; 257:486.
61 Perrot LJ. Masque ecc hymotique. Specific or nonspecific
indicator for abuse. Am] Forens ic iVIed 1989; Pathology
10:95-7.
62 Rooms L, Fitzgerald N, McClain KL. Hemophagocytic
lymphohistiocytosis masquerading as child abuse ;
presentation of three cases and revi ew of cen tral nervous
system findin gs in hemophagocytic Iy mphohistiocy tosi s.
Pediatrics 2003 ; 111:636-640.
63 Pearson RW, Triplett DA. Factor XI assay results in the
CAP su rvey (1981). Am] Clil1 Pathol 198 2; 78(Suppl.
4) :615-620.
64 UK NEQAS for blood coagulation Survey !l8, September 1999.
65 Martin CM, Engstrom PF, Barrett O. Surreptitious self­
administration of heparin. ]AMA 1970; 212:475-476.
66 Blanchette VS, Dean J, Lillicrap D. Rare co ngen ital
hemorrbagic disorders. In Lilleyman J, Hann I, Blan chette V
(eds) Pediatric Hematology, London: Churchill Livingstone ,
1999,611-628.
67 Seligsohn U. Facto r XI deficien cy. Tinomb I-Iaemostasis 1993;
70: 68-71.
68 Green D, Lechn er K. A survey of 215 non-hemophilic pa tients
with inhibitors to fa cto r VlII. Thromb Ha el110stasis 1981;
45 :200- 203.
69 Nakashima K, Miyahara T, Fuj ii S et al. Spontaneously
acquired fa cto r VllJ inhibitor in a 7-ye ar- old girl. Acta
Haematol 1982; 68:58-62.
70 Reece EA, Clyne lP, Romero R, Hobbins J.e. Spontaneous
factor XI inhibitors: seven additional cases and a review of
the lite rature. Arch In t Med 1984; 144 :525-9.
71 Beck DW, Strauss RG , Kisker CT, Henriksen RA. An intrinsic
coagulation pathway inhibitor in a three year old child.
Am] Ciin Patho11979; 71:470-472.
References I 103
72 Shapiro SS, Martin ez J, Holburn RR. Congenital
dysprothrombinaemia: an inherited stru ctural disorder of
human prothrombin.] Ciin In vest 1969; 48:2251-2259.
73 Miller CH, Lenzi R, Breen e. Prevalence of von Willcbrand's
disease among US adults. Blood 1987; 69:454-459.
74 Werner EJ, Broxson EH, Tucker n, Manaiac LF, Giroux DS,
Abshire Te. Prevalence of vo n Willebrand disease in children:
a multiethnic study. Blood 1991; 78:68a.
75 Castaman G, Fredericic AB, Rodeghiero F, Mannucci PM.
Von Willebrand's Disease in tb e year 2003: towards complete
identification of gene defects for th e correct diagnos is and
treatment. Haematologica 2003; 88 :94-108.
76 Abildgaard CF, Su zuki Z, Harrison J et al. Serial studies in
von Willebrand disease: variability versus 'variants '. Blo od
1980; 56:712-716.
77 Laffan M, Brown SA, Collins PW et al. The dia gnosiS of
von Willebrand disease: a guideline from the UK Hae mophilia
Centre Doctors' Organisation. Haemophilia 2004; 10: 199-217.
78 Castaman G, Eikenboom JCJ, Bertina RM, Rodeghiero F.
In co nsistency of associa tion between type I von Willebrand
di sease phenotype and genotype in families id ent ified in an
epidemiological investigation. Tinomb Ha emostas is 1999;
82: 1065-1070.
79 Frederici AB, Mannucci PM. Actual management of von
Wil!ebrand disease: first report on 880 cases of the Ita lian
regis tlY of vWD. Blood 1997; 90:32a.
80 Frederi ci AB. DiagnosiS of vo n Willebrand disease. Semill
Thromb Hemostasis 1998; 24:654-660.
81 Silwer J. Von Willebrand 's disease in Sweden. Acta
Pa ediatrica Scandanovia 1973 ; 238:1-159.
82 Gill JC, Endres-Broo ks J, Bauer PJet al. The effec ts of ABO
blood groups on the diagnosis of vo n Willebrand disease.
Blood 1987; 69:1691-1695.
83 Ruggeri ZM, Zimmerman TS. Variant von Willebrand 's
disease. Characteri zation of two subtypes by analysis of
multimeric composi tion of FVllljvWF in pl asma and platelets.
] Ciil1 fl1/Iest 1980; 65: 1318-1325.
84 Mill er JL, Ca stella A. Pl atelet-typ e von Willebrand disea se:
characterization of a new bleeding disorder. Blood 1982;
60:790-794.
85 Rinder MR, Richard RE, Rinder HM. Acquired von Willeb rand's
disease: a concise review. Am] Hematol1997; 54:139-145.
86 Sim one N, Co rnet JA. Abildgaard CF. Acquired von
Willebrand ' syndrome in systemic lupu s erythe ma tosus. Blood
1964; 31:806-812.
87 Pareti Fl, Mannucci PM, D'Angelo A et aL Congenital
deficiency of thromboxane and prostacyclin. La ncet 1980;
1:898-901.
88 Duckert F. Jung E, Shmerling DH. A hitherto undescribed
co ngen ital hemorrhagic diathesis probably due to fibrin
stabilizing fact or deficien cy. Throl17b Dia thesis Ha em orrl1agic
1960 ; 5:179-186.
89 Sakata Y, Aoki N. Si gnitl ca nc e of cross-linking of a lpha
2-pJasmin inhibitor to tlbrin in inhibition of fibrinolysis and
in hemostasis.] C/in Il1vest 1982; 69:536-542.
90 Duckert F. Documentation of the pl as ma fa ctor Xlli
deficiency in man. Anl1 New York Acad Sci 1972;
20 2:190-199.
91 Lorand L, Losowsky MS, MiJoszewski KJM. Human factor
XIII: fibrin s tabilizin g factor. Progress Hemostas is Tl1romb
1980; 5:245-290.
92 Saito H. Alpha 2-plasm in inhibitor a nd its deficiency states.
1 Lab Ciin Med 1988; 112:671-678.
93 Lind B, Thorse n S. A novel missense mutation in the
human plasma inhibitor (alpharantiplasmin) gene
associated with a bleeding tendency. Br] Haematol 1999;
107:317-322.
104 I Haematological abnormalities that can simulate abuse
94 Schleef RR, Higgins DL, Pillemer E, Levitt LJ. Bleeding diathesis
due to decreased functional activity of type I plasminogen
activator inhibitor.) Clillinvest 1989; 83:1747-175 2.
95 Dieval J, Nguyen G, Gross S et al. A lifelong bleeding disorder
associated with a deficiency of plasminogen activator
inhibitor type J. Blood 1991; 77:5 28- 53 2.
96 Fay WP, Parker AC, Condrey LR, Shapiro AD. Human
pla sminogen activator inhbitor-l (PAl-I) defici e ncy:
characterization of a large kindred wi th a null mutation in the
PAI-l gene. Blood 1997 ; 90:204-208.
97 Nurden AT, Caen JP. An abnormal platelet glycoprotein
12 2 Conway JH, Hilgarter MW. Initial presentations of pediatric
pattern in three cases of Glan z mann's thromba sthenia.
hemophiliacs. Arch Pediatr Adolesc Med 1994; 148:589-594.
Br) Haematol 1974 ; 28:253-260.
123 Schiver W, Brueschke EE, Dent T. Family practice grand rounds:
98 Phillips DR, Agin PP. Platelet membrane defects in
Gl a nzmann 's thrombasthenia. Evidence for decreased amounts
of two maj or glycoproteins.) Gin Invest J977; 60:535-545.
99 Gl a n zman n E. Hereditare hamorrhagische Thrombasthenie.
12 5 Bolton-Maggs PH , Peny OJ , Chalmers EA et a!. The rare
Ein Betrag zur Pathologie der Blutplattchen. )alzrbuch
coagulation disorders : review with guidelines from the UK
Kinderheilkunde 1918 ; 18 : 1-42.
haemophilia Doctors Organisation. Hael110philia 2004;
100 Sauslbury FT, Kesler RW. Thrombocytosis in Henoch­
Schonlein purpura. Clinical Pediatr 1983; 22:185-87.
101 Weinstein M, Babyn P, Zlotkin S. An orange a day keeps the
doctor away: scurvy in the year 2000. Pediatrics 2001; 108:55.
102 Chatproedprai S, Wananukul S. Scurvy: a case report. ) Med
Assoc Thailand 2001; U4(Suppl. J):S 106-lJO.
103 Stewart GM , Rosenberg NM. Conditions mistaken for child
abuse: part I. Paediatr Emerg Care 1996; 12:116-12J.
104 Aldri ch RA. Steinberg AG, Campbell DC. Pedig ree
demonstrating a sex-linked recessive condition characterized
by draining ears, eczematoid dermatitis and bloody diarrhea.
Pediatrics 1954; 13 :1 33-139.
105 Clemetson KH, Clemetson JM. Molecular abnormalities in
Glanzman 's thrombasthenia, Bernard-Soulier syndrome and
platelet-type von Willebrand's disease. CUTl' Opill Hematol
1994; 1 :388-393.
106 Hermansky F, Pudlak P. Albinism associated with
hem orrhagic diathesi s and unusual reticulated pigmented cells
in the bone marrow: a report of tVIO cases with histoch emical
studies. Blood 1959; 14:162-169.
107 White JG. Structural defects in inherited and giant platelet
disorders. Adv HlIm Gen et 1990; 19:133-234.
108 White JG. Inherited abnormalities of the platelet membrane
and sec retory granules. Human Pa th ology 1987 ; 1U: 123-139.
109 Greinacher A. Muelle r-Eckh a rdt C. Hereditary types of
thrombocytopenia with giant platelets and inclusion bodi es in
the leukocytes. BIt//' 1990; 60:53-60.
110 Djaldetti M, Creter 0, Bujanover Y, Elian E. Ultrastructural
and functional studies of the platelets in patien ts with
May-H egglin anomaly. Haemafologica 1982; 67:530-538.
III Epstein CJ, Sahud MA, Piel CF et al. Hereditary
macrothrombocytopenia , nephriti s and deafness. Am) Med
1972; 52:299-310.
112 Eckstein JD , Filip OJ , Watts JC. Hereditary thrombocytopenia,
deafness and renal disease. Ann Intern Med 197 5 ; 82:639-645.
113 Lilleyman JS. Intracranial haemorrhage in idiopathic
thrombocytopenic purpura. Arch Dis Child 1994; 61:251-253.
114 Lilleyman JS. Intracranial haemorrhage in chronic childhood
ITP. Pedia tric Hematology and On cology 1997; 14:iii-v.
115 Lilleyman JS. Managem e nt of childhood idiopathic
thrombocytopenic purpura. Br) Haematol 1999; 105: 871-875.
115 Bolton-Maggs PHB, Moon I. Assessment of UK practi ce for
management of acute childhood idiopathic thrombocytopenic
purpura against published guidelines. Lancet 1997;
350:620-623.
117 Wheel er OM, Hobbs CJ. Mistakes in diagnosing non­
accidental injury: ]0 years' experience. Medical Practice
1988; 296:1 23 3-123 6.
118 Smith PS. Co ngenital coagul a tion protein deficiencies in the
perinatal period. Semin Perinatal 1990 ; 14: 384-392.
119 KJetzel M, Miller CH, Becton DL et 31. Post delivery head
bleeding in hemophili c neonates. Am ) Dis Clli/d 1989;
143: 1107-1110.
120 Ljung R, Lindgren AC, Petrini P, Tengborn L. Normal vaginal
delivery is to be recommended for haemophilia ca rrier
gravidae. Acta Paediatrica 1994; 83 :609-61 J.
J21 Kadir RA, Economides DL. Obstetric management of carriers
of haemophilia A. Haemophilia J997; 3:81-86.
hemophilia.) Fam Pmct 1982; 14:661-662,664, passim.
124 Peyvandi F, Mannucci PM. Rare coagulation disorders.
Thromb Haemostasis 1999; 82:1207-1214.
10:593-628.
126 Katz JA, Moake JL, McPherson PO et al. Relationship between
human devel opment and disappearance of unusually large
von Willebrand factor multimers from plasma. Blood J989 ;
73: 1851-1858.
127 Wetzel RC, Slater AJ, Dover GJ. Fatal intramuscular bleeding
misdiagnosed as suspected non-accidental injury. Pediatrics
1995; 95:771-773.
128 Carpentieri U, Gustavson LP, Haggard ME. Misdiagnosis of
neglect in a child with bleeding disorder and cystic fibro sis.
South Med) 1978; 71 :854-855.
129 Kaplan SJ. Pelcovitz 0, Salzinger S et al. Adolescent physical
abuse: risk for adolescent psychiatri c disorders. Am J
Psychiatry 1998 ; 155: 954-959.
130 Sutor AH , Dagres N, Nied erhoff H. Late form of vitamin K
bl eeding in Germany. Klinische Padiatrie 1995; 207:89-97.
131 White ST, Voter K, Perry J. Surreptitious warfarin ingestion.
Child Abuse Neglect 1985; 9: 3 49-352.
132 Souid AK, Korins K, Keith 0 et al. Unexplained menorrhagia
and he m a turia: a case of Munchausen's syndrome by proxy.
Pedin tr Hel17otol Oncol 1993; 10:245-248.
133 Warkentin TE, Kelton JG. Aquired platelet disorders. In Bloom
AL, Forbes CD, Thomas DP et al (eds) Haemostasis and
Thrombosis. London: Churchill Livingstone, 1994, pp. 767-815.
134 Richardson SG, Fletcher OJ, Jeavons PM, Stuart J. Sodium
valproate and platelet function. BM) 1975; 1 :221-2 22 .
135 Barr RD, Copeland SA, Stockwelll\llL et a l. Valproic acid and
immune thrombocytopenia. Arch Dis Childh 1982; 57:681-684.
136 Ko CH, Kong CK, Tse Pw. Valproic acid and thrombocytopenia.
Hong Kong Med ) 2001; 7: J5-21.
137 Conley EL, Coley KC, Pollock BG et al. Prevalence and risk of
thrombocytopenia with valproic acid: experience at a
psychiatric teaching hospital. Pharmacoth erapy 2001;
21:1325-1330.
138 De Beradis 0, Campanella 0, Matera V et a l.
Thromb ocytopenia during valproic acid treatment in young
patients with new-onset bipolar diso rder. ) Ciin
Psychopharma col2003; 23:451-458.
139 Ganick OJ, Sunder T, Finley JL. Severe hematolog ic toxicity
of valproic acid: a report of four patients. Am) Pedintr
Hematol Oncol 1990; 12:80-85.
140 Gesundheit B, Kirby M, Lau W et al. Thrombocytopenia and
megakaryocyte dysplasia: an adverse effect of v alproic acid
treatment.) Pediatr Hematol Onco12002; 24:589-590.
141 Durongpisitkul K, Fururaj VJ, Park JM, Martin CF. The
prevention of coronary artery anuerysm in Ka wasaki disease:
a meta-analysis on the efficacy of aspirin and
immun oglobulin treatment. Pediatrics 1995 ; 96: 1057- 106J.

142 Stuart MJ, Gross SJ, Elrad H, Graeber JE . Effects of
acetosalicylic- acid ingestion on materna l and neon ata l
haemostasis. N Engl J Med 1982 ; 307:909-912.
\43 Freedman MH, Doyle JJ. Inh erited bone marrow failure
syndro mes. In Lilleyman J, Hann I, Blanchette V (eds) Pediatric
Hematology. London: Churchill Livingstone, J 999, 23- 49.
144 Young NS, Alter BP. Clinical features ofFanco ni 's anae mia. In
Aplastic Anaemia, Acquired 011d fuherited. Phila delphia, PA:
WB Saunders, \ 994 , pp: 275-309.
\45 Eden DB, Lilleyman JS. Guidelines for the management of
idiopathic thrombocytopenic purpura. Arch Dis Child 1992 ;
67: 1056-1058,
146 McCla in JL, Clark MM, Sandu sky GE, Undiagnosed, untrea ted
acute lymphobl astic leukaemi a presenting as susp ected child
abuse. J Forensic Sci 1990 ; 35:735-739,
147 Berrebi A, Ma lnick SDH, Vorst fJ , Stein D, Hi gh incidence of
factor Xl deficiency in Gau cher's disease. Am J Hema lol
1992; 40:153-161.
148 Humphries JE, Hess CE, Gaucher's diseas e and acquired
coagu lopathy. Am J Hemato l 1994; 45:347-353,
149 Kelsey H, Chri stopoulos C, Gray AA. Mach in SI. Acquired
pseudo-pseudo Bernard-Soulier syndrome comp licating
Gaucher'S disease. J c/in Palh ol 1994; 47:162-165,
References I 1 05
150 Leititis JU, Brandis M, Critical care in uraemic children,
Pediatric Nephrology \ 992; 6:88-95,
151 Kelly D, Summerfi eld J, Haemostasis in liver disea se, Semil1
Liver Dis 1987; 7:182-\91.
\ 52 Miner ME, Kaufman HH, Graham SH, Haar FH, Gilden berg PL.
Disseminated intra vascu lar coagul ation fibrin olytic syndrom e
following head injulY in children: frequency and prognostic
implications, J Pedialr 1982; 100:687-691.
153 Hymel](P, Abshire Te, Luckey DW, Jenny C. Coa gulopathy
in pediatric abusive head trauma. Pe(iialrics 1997;
99:371-375.
154 Stein SC, Young GS, Talucci RC et at. Delayed bra in injury
after h ea d trauma: sign ificance of coagulopathy. NeurosurgelY
\992 ; 30:160-165.
155 Kaufman HH, Moake JL, Olson JD et al. Delayed and recurrent
intracran ial hematolnas related to disseminated intravascular
clottin g and fibrinoly sis in head injury, Neurosurgery 1980;
7:445- 449 ,
156 Wedemeyer Ai, Ed son JR, Krivit W. Coagulation in
cyanotic congenital heart disease. Am J Dis Child 1972;
124:656- 660,
157 Waldman JD, Czapek EE , Paul MH et al. Shorten ed platelet
survival in cyanotic heart disease. J Pedialr 1975; 87:77-79.
I CHAPTERS I
BIOCHEMICAL INVESTIGATIONS
ON POST-MORTEM SPECIMENS
Denis R Benjamin
Introduction 106
General evaluation 107
Hypoxia 109
Inflammation 109
Anaphylaxis 109
Infection 110
Dehydration and electrolytes 110
INTRODUCTION
fluids decreases, compounds may dissociate from proteins
The diagnostic use of biochemical tests on post-mortem
specimens dates back to the earliest development of the
technology that enabl ed us to measure the blood concen­
trations of various compounds. I It is safe to assume that
almost evelY analyte measured during life has been
investigated in post-mortem samples. Unfoliunately, the
changes associated with death can profoundly affect the
concentration of many intermediary metabolites. As a ll
the dynamic energy systems normally maintaining various
concentration gradients begin to fail or run out of sub­
strate, compounds rapidly equi librate across the various
body compartments and cell membranes. Such changes
occur in the first few minutes to hours after death. Indeed,
many such changes may already be under way before a
formal declaration of death by the medical a ttendan ts.
With time, as cells die, intracellular compounds are
rel eased into the surrounding tissue and circulation, com­
pounding t he difficulties in the interpretation of results.
This is well illustrated by the onset of haemolysis in the
circulation. The rupture of red cells and t he release of
the haemoglobin and al l the intracellular enzymes in to the
bloodstream often signa l the end of the useful period for
the measurement of many metabolites. Not only does the
concentration change in unpredictable ways, but also the
' matrix ' in which a particular ana lyte is normally measured
may be sign ificantly a ltered and this can have an important
imp act on the analytical method. As the pH of vario us body
Time of death (post-mortem interval) 111
Endocrine disorders 112
Genetic metabolic disorders presenting as sudden
unexpected death 114
Technical considerations at the time of autopsy 117
References 120
and other ligands thus changing their measured concentra­
tion. This is especially important in the interpretation of drug
levels. A recent study, which included a large compendium
of drugs detected in post-mortem blood, examin ed the pos­
sible factors affecting their concentrations. 2 Hormones and
other compounds may be similarly affected.
Some compounds are relatively stable for prolonged
periods of time. These are of the most value in post-mortem
assessment. Others show relatively predictable changes,
increasing or decreasing with some definable relationship
to the post-moliem in tervaL Numerous efforts have been
made to use these for the assessment of the time since death,
or if the time were accurately known, to use some 'correc­
tion factor ' or equation to calculate what the ante-mortem
value would have been. Both uses are fraught with diffi­
culty as the many factors influencing the rate of change are
usually not known, nor can they be readily controlled.
Other compounds show erratic and unpredictable behaviour
with increasing time after death, which is of almost no ben­
efit in post-mortem diagnosis. For many analyies, changes
in blood concentration are so non-specific that their meas­
urement offers no signifi cant diagnostic information.
Despite the obvious drawbacks of the post-mortem
changes, the investigation of sudden, unexpected death
in infants a nd children can be substantially aided by the
judicious use of post-mortem biochemical testing on a
variety of body fluids a nd tissues. The recognition, over the
past decade, of a large number of inherited metabolic

defects, which may not be preceded by overt or obvious
clillical disease, has greatly ex pand ed the differential diag­
nosis of such cases. A high index of suspicio n will ensure
[ha t these are not missed at the time of the autopsy, as
diagnosis has imp0l1ant implications for the family and
future pregnancy planning. If ca refully so ught, clues are
frequently uncovered in the his tO lY or on the gross and
microscopic exam ination alerting the pathologist to a pos­
sible underl ying disease. Newer technologies, such as ta n­
dem mass sp ectrometry, now permi t ready scree ning for a
wide variety of metabolic defects, an d molecul ar biologica l
techniques can be used for specific genetic disord ers. These
have considerably exp anded our abi lity to establish diag­
noses from post-mortem material. It is pruden t to coHect all
of the approp ri ate samples, en sure that they are suitably
stored an d then decide on how to proceed, once the various
lines of evidence unfold.
All the caveats pe11aining to post-mortem biochemistry
in adults apply in children . Most of the studies relating to the
changes in con centration of various metabolites after death
have been perfo rmed in ad ults, although a number have
been validated in children. Table 5.1 lists the common fac­
tors influencin g the interpretation of bioch emical resul ts.
The literature relating to thana tochemi stry is relatively old.
Many of the studies were carried out in the early yea rs, soo n
after clinical patho logy techni ques became available. Many
have no t been rigo rously repeated with curren t technology
and this has created potential problems. For example, flam e
photometry, whi ch was the sta ndard for electrolyte measure­
ments in the 1950s and 1960s, has been largely repla ced
with ion-selective electrodes. Each analytic method is sensi­
tive to its own particular set of interferences, and one cannot
necessarily extrap olate results between different methodol o­
gies. For this reaso n, reference ranges published in the older
literature may no longer have currency.
For each analyte these variables need to be known and
appropriately considered. Because of the large number of
possibilities, great precision is frequently not possible. Inter­
pretation of the test results often requi res considerable judg­
ment and certainty of diagnosis is seldom achieved ,
especially with the measurement of traditional intemlediary
metabolites and electrolytes Even so, these difficulties should
Table 5.1 Factors affecting the results of biochemical
investigations
Post-mortem interval
Temperature at which the body has been maintained
Sample type (e.g. blood, vitreous humour, urine, cerebrospinal
fluid, etc.)
Sample site (e.g . left or right heart, peripheral blood)
Ana Iytica I method
Interfering substances, e.g. drug s, alcoh ol
Resuscitative attempts
Mode of death (e.g. rapid, prolonged, presence of hypoxia, etc.)
General eva lu ation I 107
not militate against the thoughtful use of selected biochemi­
cal tests. On the other hand, the use of molecular biologic
tech niques and tandem mass spectrometlY has opened pow­
erful new avenues for the definitive diagnosis of a host of dis­
orders in infants and children who die unexpectedly.
Failure to appreciate the impact of those factors influ­
encing results of post-mortem biochemical investigations,
or to perform adequate cont ro l studies, has spawned a large
literature in which numerous claims have been made for
the pathogenesis of certain diseases, only to be refuted by
subsequent studies. Nowhere is this more eviden t than in
the investigation of sudden infant death sy ndrome (SIDS,
or criblcot death). The 'scientific' literature is filled with
such poorly designed studies . There has been a characteris­
tic pattern of publication in this field . The first couple of
articles usua lly note an interesting chem ical observation in
a small number of infa nts. This is claimed to be sp ecific
a nd sensitive. A few articles validatin g the observation
may follo w. But soon the naysayers arrive, poin ting out
similar findings in a number of other settings. The speci­
ficity of the original fin ding co mes under se rious scrutin y
as more 'controls' are added and the impact of post­
mortem in terval and other factors are assessed. And not
infrequently, the test or obselvation that appea red so hope­
ful at its inception fades into obscurity.
This chapter only covers those analytes or compounds for
which interpretation is reasonably possible and there is some
demonstrated clinical application . Readers are referred to a
number of excellen t general reviews of post-mortem bio­
chemistlY if they require information about analytes that
are not dealt with in this discussion. The following are nat
covered in this chapter: lactate, pyruvate, ammonia, serum
enzymes, lipids, trace metals, immunoglobulins,) methaemo­
globin,4 myoglo bin,5.6 thiamine,? and thyroid and parathy­
roid hormones 8- 14 as well as the other hormones.) It is indeed
un fortunate that four of the most important indica tors of
metabolic diseases in infants and children, namely, elevated
ammonia, low glucose, abnormal pyruvate an d lactate are so
unreliable in the post-mortem pelio d. Other fluids, such as
pelicardial 15 and synovial16 have had such limited study that
fe w ge neralizations about their value in the day-to-day prac­
tice of pediatlic pathology can be made.
In considering the role of biochemical testing in the
post-mortem period in pediatric fore nsic cases, tests can be
di vid ed into two broad gro ups. Some tests are us eful for the
assessment of t he general health of the child, whereas oth­
ers focus on specific clinical disorders. Table 5.2 lists the
role of these anci lla ry pro cedures in pediatric cases.
GENERAL EVALUATION
Nutritional Status
The nutritional statu s of the deceased is an important con­
sideratio n in many pa ediatric au topsies. This may have
108 I Biochemical investigations on post-mortem specimens
Table 5.2 Role of biochemical testing in paediatric forensic cases
General evaluation
Nutritional status
Hypoxia
Inflammation
Anaphylaxis
Infection
Dehydration
Time of death (post-mortem interval)
Specific disarders
Endocrine disorders
Diabetes
Adrenal and/or pituitary insufficiency
Renal failure
Inborn errors of metabolism
Fatty acid oxidation defects
Other mitochondrial disorders
Urea cycle defects
Aminoacidopathies
Organic acid urias
Abnormal carbohydrate metabolism
Miscellaneous
implications in regard to parental neglect and/or abuse, or
to a chronic, but clinically unsuspected, underlying dis­
ease. Evaluation of the physical characteristics and stan­
dard anthropometric measurements should be carefully
performed and compared with the appropriate standards
for age, sex and ethnic group. These will be abnormal in
many chronic diseases, including endocrinolog ica l, meta­
bolic, immunological and renal disord ers. From the labora­
tory standpoint there are only a small number of tests that
can be utilized to assess both recent and long-term nutri­
tional status. Obviously, none of the dynamic or functional
tests that is helpful during life can be used.
RECE NT NUTRITIONAL INTAKE
Knowing the recent food intake history of the deceased
sometimes has important diagnostic implications. Perhaps
the best serum protein for evaluating recent protein and
calorie intake is prealbumin. Prealbumin, also known as
thyroxin-binding prealbumin or transthyretin, is a transport
protein with a very short half-life (2 days). It is synthesized
by the liver and has a small pool size. Either protein or calo­
rie deprivation will result in a decrease in the serum level of
prealbumin in 3 to 4 days.17,18 As such it is an indicator of
recent nutritional intake rather than an accurate assessment
of nut ritional status. Any infant or child who has been
recently sick or who may have had significant anorexia
could have a depressed prealbumin level. There are no good
studies of prealbumin in the immediate post-mortem inter­
val. Retinal-binding protein, also produced by the liver has
an even shorter half-life (approximately 12 hours). It has no
advantages over prealbumin for the assessment of recent
nutritional intake.
Both these proteins are very sensitive to the presence of
the acute inflammatory response. Hepatic synthesis rapidly
decreases in the face of any inflammatory stimulus, be it
from infection, trauma, neoplasia or autoimmune disease.
For this reaso n prealbumin should always be combined with
some measure of the acute phase response. At this time
C-reactive protein is the recommended test (see below).19
LONG-TERM NUTRITIONAL STATUS
The long-term nutritional status is much more problematic
to assess from a biochemical standpOint and the laboratory
probably offers few significant advantages over anthropo­
metric studies. When the body is adapting to the decrease in
nutrients, including total calories and protein, there is a phase
of compensation during which a variety of hormonal adapta­
tions and changes occur. These are difficult enough to evalu­
ate under ideal circumstances during life, and become even
more so with the post-mortem changes in blood chemistry.
Similarly, many tests of immune function become abnormal
in the chronically malnourished infant and child but these are
almost impossible to evaluate after death.
Selected serum proteins have long been used as surro­
gate measures of nutrition. Albumin is a non-specific and
relatively insensitive measure of nutritional status. Its
serum concentration is dependent on a variety of factors.
Post-mortem levels in the serum are close to ante-mortem
values unless there is significant haemolysis. The problem
is not with any post-mortem changes, but with t he ma ny
factors influencing serum albumin concentration during
life . The serum value is the net result of the protein intake
and hepatic synthesis, degradation and loss. Each of these
is subject to many influences other than protein intake. It
should also be remembered that in cases in which caloric
insufficiency exceeds the protein deprivation (i.e. maras­
mus) serum protein levels might be normal.
Because of the long half-life of albumin in the circula­
tion (18-20 days) and its large extravascular pool, changes
in serum values are slow to develop. Decrease in serum
albumin develops late in the course of malnutrition and
only in the more severe cases. In interpreting the albumin
level the following factors must be kept in mind:
• Serum albumin is dependent on t he hydrational state
of the patient, so that dehydration in a malnourished
child may result in relatively normal albumin level.
• Both synthesis and catabolism are dep endent on
hepatic function and the acute infl a mmatory response.
The presence of abnormalities of either will
significantly change albumin levels.
• Renal and gastrointestinal losses of protein can
profoundly decrease the albumin levels.
If all of these can be excluded then low levels roughly
correlate with the severity of chronic malnutrition.

HYPOXIA
appropriate clinica l indications, albeit quite non-spec ific.
A laboratory test for the detection of ante-mortem hypoxia
would be a great boon for the forensic pathologis t. The
question arises in a diverse array of circumstances, varyin g
from the suspicion of suffoca tion and strangulation to
drowning and fire victi ms and to natural causes of death,
such as SIDS. Needless to say, although many tests have
been investigated, none has proven to be reliable or satis­
factory. This is hardly a surprise as so many mechanisms of
death involve some element of hypoxia as one of the com­
mon final pathways a nd the rapidity of the post-mortem
change on most of the candidate chemical compounds pre­
cludes their use after more than a few minutes . For exam­
ple, post-mortem oximetry, with blood being samp led from
va rious sites in the circulation, was no t fou nd to be partic­
ularly helpful or reliable in defining the cause of death in
one study of 214 ad ult cases .20 Other compounds have also
been investigated from this standpoint,21 such as lactate,
which increases in response to anaerobic glyco lysis. How­
ever, there is such wide variation in the res ul ts that inter­
pretation in th e indivi dual case is all but impossible.
Normally, lactate rises quite promptly in the blood - within
a day reaching leve ls of up to 70 times greater than the
ante-mortem va lues .22 The rise of lactate in the vi treous
humour is not quite so dramatic, but is sti ll substantial. 2J
Hypoxa nt hin e is formed from adenosi ne monophosphate
in the presence of hypoxia. Its appearance in the vitreous
humour has been used as indicatio n of ante-mortem
hypoxia. For examp le, some investigators have hypothesized
that a SIDS death may be preceded by hypoxic episodes of
varying duration. 24 One study investigated the hypoxan­
thine levels in four g roups of patients including a group of
SlDS victims, ch ildren with congenital heart disease, babies
who succumbed from respiratory distress and a control
group of accident victims in whom there was no evidence of
hypoxia. A sizeable fraction of the infants and children in
the SIDS, respiratory distress and congenital heart groups
had elevated levels of vitreous hypoxanthine.25 This obser­
vation has been challenged by other stud ies, which have all
failed to corroborate the results. 26 - 28 At this time there is
conflicting evidence to support a significant difference in the
hypoxa nthine concentration between SIDS and other causes
of death in infants. 29 ,)O Most recently, the meas urement of
vascular endothelial growth factor (VEGF) in cerebrospinal
fluid (CSF) has been used to assess the incidence of hypoxia
in SlDS victims.)1
INFLAMMATION
It is freque ntly useful to know if there is a pre-ex isting
infl amm atory state at the time of death. This 'acute phase
response' commonly accompanies infections, autoimmune
conditions or maligna ncies. The erythrocyte sedimentation
rate (ESR) has been the tradition al screening test with the
Anaphylaxis I 1.09
Performing a satisfactory ESR test after death is not possi­
ble with all the changes in coagulation and fibrinolysis. In
the last two decades, C-reactive protein has found a very
useful role as an alternative or a compl ementary indication
of inflammation; it can be used as a replacement for the
ESR in the majority of clinical situations or in addition to
it. C-reactive protein has many advantages over the ESR, as
it is not influenced by many of the other proteins that
affect the ESR and it is independent of the shape, size and
number of the red cells. It is produced by the liver as part
of the acute phase response and is a sensitive marker of
inflammation . It increases dramatically in the serum, rising
to over 10 times the baseline value within 12 hours.)2,))
None of the other proteins that also increase in response to
inflammation, such as a -acid glycop rotein (orosmucoid)
a nd tra nsferrin , offers any advantages over C-reactive pro­
tein in terms of rapidity of change, ease of technical meas­
urement or stability in the serum afte r death. A C-reactive
protein level that is greater than 0.8 mg/dL is good indica­
tion of a pre-existing inflammatory state. The higher the
valu e, t he greater the chance of a bacterial infection,
alth ough elevations in C-reactive protein should always be
considered quite non-specific. A number of studies have
confirmed the relative stability of this protein in the imme­
di ate post-mortem period. )4,)5 A number of recent investi ­
gations have also confirmed its value.)6,)7
A great deal of recent work, from Europe in general and
France and Germany in particular, has focused on the use
of pro calcitonin as a marker of bacterial infection .)8,)9
There is as yet no consensus of opinion on whether this is
more useful t han C-reactive protein. Its behaviour in post­
mortem specimens has recently been studied,4o and some
of the changes associated with tempera ture and storage
have also been documented .41
ANAPHYLAXIS
An acute anaphylactic reaction is a possible mechanism of
death in a small minority of paediatric cases. There is usually
a past history of severe hyp ersensitivity to a particul ar all er­
gen, such as peanuts or shellfish. However, bee venom and
other allergens may also be responsible. Generally the clini­
cal features of overwhelming anaphylaxis are evident,
although in some cases they can be so acute or confusing as
to go unrecognized. A number of tests have been promoted
for the diagnosis of anap hylaxis. Histamine is released from
both tissue mast cells and the circulating basophils. It has a
very short half-life and so is of no use in the post-mortem
diagnosis of anaphylaxis. Of the other tests, serum tryptase,
released from degranul ating mast celis, has attracted the most
attention. This enzyme is not released from the circulating
basophils. A number of studies in living patients, involvi ng
cases of witnessed anaphylaxis, have demonstrated raised
levels of serum tryptase in association with an increase in
110 I Biochemical investigations on post-mortem specimens
allergen-specific immunoglobulin E (IgE). In the early studies
of post-mortem specimens the resuHs appeared quite encour­
aging; however, these involved only a small number of
cases. 42.43 Subsequent evaluation in a group of 49 patients
known to have died from other causes, with no evidence of
anaphyla xis whatsoever, showed that many samples had
levels of tryptase above the usual detectable levels in the
serum.44 Of these, five subjects had levels of tryptase that
were greater than 10 ng/mL, which was regarded as a positive
result, and one had an extremely high value. The mechanisms
for these increases were quite unclear. There was no correla­
tion with the post-mortem interval. Further studies have
demonstrated raised levels in a variety of circumstances. 45
At this time it is advisable to utilize both the senlm
tryptase level and an increase in the allergen-specific IgE
for the diagnosis of anaphy laxis. A recent study concluded
that as many of 13 per cent of deaths in adults may be
accompan ied by the activation of mast cells 46 There has
been one formal investigation into the lise of tryptase in
sudden unexpected deaths in infants. This study suggested
that mast cell degranulation was more evident in infants in
t he SIDS groups than in the infants in whom a cause of
death was found. 47 This was a small study that has not
been corroborated in a systematic survey involving a larger
number of control subjects, with careful attention to the
post-mortem interval and mechanism of death; this issue
has yet to be resolved. 48.49
INFECTION
If there is any suspicion of infection then the appropriate cul­
tures should be obtained. The problems of interpreting post­
mortem bacterial cultures are well known , eve n if standard
techniques are followed. Culturing multiple sites may be use­
ful or separately sampling each ventricle of the heart might
provide additional infomlation that assists in the interpreta­
tion of positive cultures. 5o Viral cultures can be very inform­
ative and the use of new polymerase chain reaction (PCR)
assays for the detection of many human pathogens is becom­
ing more widespread. These should always be considered in
cases of possible myocarditis and meningit is. Over the years
we have been struck by the poor correlation between the
gross appearance of the heart and the microscopic presence
of significant myocarditis. For this reason we tend to err on
the side of conservatism and obtain viral studies in most
cases of unexpected death. The inherent sensitivity of these
molecular diagnostic techniques raises the difficult issue of
the clinical significance of a positive finding. We are only just
beginning to learn about our 'normal' microbial flora as
defined by PCR methodology.
Assays for specific bacterial tox ins are described and
have been quite useful in selected cases. These are not gen­
erally available in most clinical laboratories. If a particular
infection is suspected from the gross and microscopic find­
ings then the pathologist should seek out the appropriate
laboratory for assistance in the diagnosis. For example,
toxic shock syndrome has been identified as a cause of
unexpected death in a child , with the identification of the
staphylococcal toxin TTST-1 in brain tissue. 51
The presence of bacterial endotoxin has been sought in
a variety of clinical situations. The usual test is based on
the limulus lysate assay, of wh ich there are a number of
modifications. The level of endotoxin is not significantly
affected by the post-moJiem interval in the first 48 hours
after death, nor does it correlate with the blood culture
results. Detectable endotoxin has been noted in a many dif­
ferent causes of death, including blunt injury as well as
severe infection. However, it is interesting to note that it
has not been found in a study of SIDS victims. 52
DEHYDRATION AND ELECTROLYTES
The question of possible dehydration is a frequent consider­
ation , especially in paediatrics when the caregivers may
underestimate the extent and severity of vomiting/diarrhoea
and there is inadequate fluid replacement. Dehydration
develops more rapidly in infants and young children than it
does in adults and is often unrecognized by the parents. The
chemical diagnosis of dehydration and the accompanying
electrolyte abnormalities have been well studied. The ability
of cells to maintain their normal concentration gradients of
ions is highly energy dependent and this rapid ly decreases
after death. Within minutes, intracellular ions, such as potas­
sium ions, begin to equilibrate across the cell membrane.
After an hour the serum potassium may well be six times
the ante-mortem level, even in the absence of haemolysis.
Once haemolysis develops, the pot assium concentration
soars even higher. Therefore, serum potassium should n ever
be used to assess the state of dehydra t ion as it increases
geometrically after death. The appearance of haemolysis is
quite variable, depending on the temperature and the storage
of the body. It can take as long as 48 hours for haemolysis
to be evident. Serum sodium concentrations' tend to
decrease more slowly after death compared with the rapid
and very significant changes with potassium. Sodium
decreases at a rate of approximately 0.9 mEq/L per hour
after death;5J chloride decreases as well , with an average
fall of 0.97 mEq/L per hour. However, there is so much vari­
ation in sodium and chloride ion levels from among
subjects that they have not been useful for the assessment
of post-mortem interval. In addition, serum is not the ideal
body fluid for the assessment of dehydration unless blood
can be obtained very soon after death.
Sodium and chloride, as well as osmolality, are much
more stable in the vitreous humour. Almost all studies of
electrolytes have relied on the measurement of their con­
centrations in v it reous humour,2J, 54,55 which are similar
in both adults and children, 56- 58 The sodium and chloride
levels in the vitreous humour change in parallel with tbe
serum and every study has shown very good correlation

between the known concentration of these during the ter­
minal hours of life and the post-mortem values in the
vitreous humour. The levels of these ions are the most
reliable for the assessment of hypernatraemic dehydration.
Measured with ion-se lective electrodes, concentrations of
sodium of > 165mEq/L and chloride >125mEq/L are
excellent indi cations of hypernatraemic dehydration. The
blood urea nitrogen (BUN) may also be significantly ele­
vated. If renal failure has developed, the creatinine will
also be increased. Although severe dehydration is the usual
cause of hypernatraemia, either as a result of a disease such
as gastroenteritis, or neglect by the caregivers, other possi­
bilities must be kept in mind. Excess loss of water may
occur through the skin and lungs in infants and children
with fever, or there may be excessive renal excretion, such
as occurs in diabetes insipidus (both central and nephro­
genic), osmotic diuresis (e.g. chronic renal failure, hyper­
glycaemia) or hypercalcaemia. Both accidental salt
poisoning and inte ntional salt poisoning have also been
described in chi ldren. 59 Claims that a significant fraction of
infants dying of SIDS have significa nt electrolyte distur­
bance based on vitreous chemistry60 are almost certainly
overestimated.
In certain cases of dehydration, when fluid witho ut suf­
ficient electrolytes is given to the patient to drink, such as
water or apple juice, hyponatraemic and hypochloraemic
dehydration can occur (e.g. excessive vomiting as might
accompany pyloric stenosis) .61 The vitreous sodium is
generally less than 135 mEq/L and the chloride is less than
95 mEq/L. Hyponatra emic dehydration is much less fre­
quent than the hypernatraem ic form. Both adrenal insuffi­
ciency (see below) and salt-losing nephritis are causes of
hyponatraemia in the dehydrated patient. Other causes of
hyponatraemia and hypochloraemia should also be consid­
ered, including lethal water intoxication due to intentional,
forced water ingestion. In the oedemato us patient, low
sodium frequently accompanies congestive cardiac failure,
hepatic cirrhosis and the nephrotic syndrome 6 2 However, it
must be remembered that not all cases of dehydration are
accompanied by significant electrolyte disturbances.
Coe J,6J has described four basic patterns of electrolyte
abnormalities in the vitreous humour in forensic cases in
which they are diagnostically helpful:
1. The hypertonic dehydration pattern, which is the
common pattern in most cases of severe dehydration
leading to an unexpected death. In this state, both
sodium and chloride are elevated, usu ally to a similar
degree, and the urea nitrogen may be mildly to
moderately increased.
2. The uraemic pattern, in which sodium and chloride are
relatively normal, while both the creatinine and the
urea nitrogen are elevated (see below).
3. The hypotonic or low-salt pattern, in which both the
sodium and chloride are low, as is the potassium
concentration; the potassium concentration should
be less than 15 mEq/L.
Time of death (post-mortem interval) I 111
4. The decomposition pattern, in which the sodium and
chloride are both low but the potassium is markedly
elevated (>20 mEq/L).
As can be deduced, the hypotonic and decomposition
patterns are only distinguished by the potassium concen ­
tration. For this reason it is important to measure the vitre­
ous potassium level as part of any electrolyte panel in order
to separate these t\ovo pathogenetic mechanisms, Vitreous
potassium has been well studied, as it shows a linear
increase after death and has been utilized for the assess­
ment of the post-mortem interval 64 (see below).
Coe and Apple 65 have studied the influence of the
method of measurement on the values of vitreous elec­
trolytes and showed very sizeable differences between tra­
ditional flame photometry and ion selective electrodes. It is
very important therefore for each laboratory performing
these assays to establish their own unique reference ranges,
and fo r the pathologist to be knowledgeable about the
methodology used for satisfactory interpretation.
Studies on the osmolality of vitreous fluid are quite lim­
ited; in normal subjects it has been repolted to range from
305 mOsm/kg 66 to 346 mOsm/kg.67 No systematic studies of
v itreous osmolality in the post-mortem period have been
performed in children covering a wide enough spectrum of
pathologies. The anion gap, a useful measurement during
life for the interpretation of electrolyte and acid-base dis­
turbances, plays no role in post-mortem chemistry,
TIME OF DEATH (POST-MORTEM INTERVAL)
Over the years, in addition to all the gross pathological evi­
dence and the temperature of various body organs, other
methods have been sought to more accurately define the
time of death. 68 .69 No other question, apart from the cause
of death, has provoked so many studies or spilled so much
ink, The answer to this forensic question is often crucial to
the entire case. Tests on blood constituents have largely
been abandoned in favour of a few ana lytes in vitreo us
humour. These have included such compo unds as inorganic
phosphorus,7o amino nitrogen, non-protein nitrogen,
ammonia and creatinine. 7I Levels of amino acids also rise
sharply after death, although there is considerable varia­
tion between the various amino acids in terms of the rate at
which their levels rise. None of these analytes has been
proven to be sufficiently rei iabl e or reproducible for the
estimation of the post-mortem interval on a routine basis.
In the last tvvo decades, potassium and hypo xanthine have
gained the most favour for the assessment of the post­
mortem intervaI. 64 ,72,7] In the uncomplicated adult case, both
the potassium and hypoxanthine concentrations in vitreous
humour, show a linear increase in the early post-mortem
period. 74,75 However, the rate of change after the first 24
hours is different from the initial rate and this is seldom taken
into account. It has been suggested that hypoxa nthine shows
112 I Biochemical investigations on post-mortem specimens
less varia tion in the early post-mortem period, especially in
the first 12 hours. 76 A variety of studies have investigated the
relationship of these two analytes to the post -mol1em interval
and the effect of temperature. Also, equations have been con­
structed to estimate the time of death, ordinarily based on
regress ion analyses from cases in which the time of death was
precisely known 77 The most recent example suggests the use
of both measurements 78 and the formulae constructed from
this study are as follows : post-m0l1em interval (hours) =
4.32 X (potassium [mmol/Lll - 18.35; post-mortem interval
(hours) = 0.31 X (hypoxanthine [mmol/Lll + 0.05. The final
estimate is based on the mean of these two calculations.
Needless to say there are a variety of methodological
and conceptual problems with these estimations. Recent
refinements in the formulae have not been systematically
tested for routine practice in paediatrics. They are far from
precise, with considerable variation from case to case. They
have not replaced body temperature in the early post­
mortem period . Their use in adults shou ld be considered
once the body reaches ambient temperature. The rate of
increase in both ana lytes is not perfectly linear over time
and this has been ignored in all the regression equations
published to date. Moreover, the mechanism of death and
the length and extent of ante-mol1em hypoxi a have not
always been well controlled for in the study popu­
lations. The environmental temperature at wh ich the body
remains after death has a major impact on the rate of
change of potassium. Finally, the number of studies that
have included paediatric cases is very limited. There is evi­
dence that potassium increases much more rapidly in the
vitreous humour of the infant than in that of the
adu]t. 3 ,58.79
Some of the problems associated with hypoxanthine are
discussed above in the section on hypoxia .
ENDOCRINE DISORDERS
Diabetes lVIellitus
GLUCOSE
Estimation of the glucose level is important for the detec­
tion of a number of disorders. It is obviously critical for the
determination of uncontrolled or untreated di abetes melli­
tus, during which the severe hyperglycaemia is usually
accompanied by significant ketoacidosis. When decreased,
glucose is one of the key metabolites indicative of a possi­
ble metabolic disease, or may be associated with starvation,
abuse and neglect. Glucose measurements have been well
investigated after death. The specimen type and its proper
selection is crucial for evaluation. If blood is taken from the
right side of the heal1 or the inferior vena cava, glucose lev­
els may be very high. This is a result of hepatic glycogenol­
ysis. With time, the glucose diffuses into the surrounding
vessels. On the other hand, very low levels of glucose from
such a sample maybe indicative of hypoglycaemia, depend­
ing on the length of the post-mortem interval. Note, how­
ever, that it has been well shown that high levels of glucose
may occur in peripheral blood sampled from a distal
extremity even in patients who are not diabetic.8o .8 ! This has
been observed in a variety of deaths due to different etiolo­
gies and may be the result of a terminal surge of noradren­
aline (norepinephrine) and adrenaline (epinephrine) and/or
the iatrogenic effects of resuscitation. 82 In interpreting the
glucose level, the terminal therapy received by the patient
should be carefully scrutinized, including all the adminis­
tered drugs and the intravenous fiuids. The use of car­
diopulmonary resuscitation outside the hospital has risen
dramatically in many Western countries, often being
administered by paramedical personnel, and such interven­
tion can significantly impact the glucose values. In some
communities it is the rare patient who avoids the chemical
'last rites'.
Some investigators, such as Coe,] feel very strongly that
a diagnosis of diabetes mellitus should never be based exclu­
sively on a blood glucose measurement. The level of glucose
in the CSF (rather than in blood) is more reliable as it is less
affected by the terminal conditions that can so dramatically
infiuence the serum. On the other hand, the increasing use of
vitreous fiuid has largely supplanted CSF as the sample of
choice for the diagnosis of hyperglycaemia.
In the normal situation, the vitreous humour glucose
level falls after death due to continuing glycolysis. In most
uncontrolled diabetics the glucose will remain above
200 mg/dL. Ketones invariably accompany the hypergly­
caemia and can be detected by the standard qualitative lab­
oratory procedures in the vitreous fiuid (or peripheral blood).
Diabetes has been diagnosed in embalmed cadavers.8] Some
have recommended the use of both lactate and glucose
measurements on the vitreous humour, in addition to the
measurement of ketone bodies in the serum, such as acetone,
for the diagnosis of diabetes. 82 However, a cl ose examina­
tion of the data shows no significant benefit of the addi­
tional lactate measurement. The circumstances and
environm enta l tempe rature around the time of death may
significantly infiuence the rate of change of the vitreous glu­
cose. Rapid cooling, as might occur in drowning (cold water
immersion) reduces the rate of glycolysis, such that the glu­
cose may be higher for a given post-mol1em in terval. This
may be compounded by the stress response accompanying
such deaths. 84 ,8 5
In recent years, glycosylated haemoglobin (haemoglobin
A Ic) has been used as a reliable marker of long-stand ing
hyperglycaemia. It is stable in the post-mol1em blood and is
a refiection of the average blood glucose leve l over the pre­
ceding 3 months. It is important to recall that there may be
other mechanisms for long-standing hyperglycaemia. For
exa mple, one study of 28 cadavers demonstrated raised
haemoglobin A Ic in all five patients with known diabetes. 86
In addition, seven subjects with malignancies, who had been
receiving steroid therapy, also had elevated levels. Other

studies have confirmed the value of measuring glycated
haemoglobin as an assessment of hyperglycemia. B7 - B9
Fructosamine, which is a measure of non-labile glycated
serum proteins, has also been used during life to monitor
the effectiveness of diabetic therapy. It too is an indication
of the mean blood sugar level over the preceding weeks.
One of the advantages of fructosamine is that it is influ­
enced neither by the presence of any abnormal haemoglo­
bins, nor the dynamics of red cell turnover in patients who
may have haematological problems. However, there are
only a few post-mortem studies on fructosamine. 90. 91
Although sudden unexpected death in childhood due to
undiagnosed diabetes mellitus is very rare, it has been
reported. 92 •93 In one particular case an ll-year-old girl died
suddenly after a 2-day history of a minor respiratory tract
infection. The autopsy was performed 3 hours after death and
documented a vitreous humour glucose level of 606 mg/dL,
blood acetone level of 24 mg/dL, marked glucosuria
(2073 mg/dL) and positive urinary ketones. Her liver con­
tained microvesicular fat and the pancreas showed typical
changes in the islets. 94 The combination of an elevated glu­
cose in the vitreous humour and acetone in the blood has
proven to be quite reliable for the diagnosis of unexpected
diabetes mellitus. Sudden death has also been well described
in children who have been recently diagnosed and are under­
going initial therapy.
The diagnosis of hypoglycaemia in post-mortem sam­
ples is much more problematic than the detection of hyper­
glycaemia. This remains a serious problem in paediatric
forensic cases as hypoglycaemia is such an important clue
for a variety of metabolic diseases. Ongoing glycolysis
generally results in a progressive decrease in glucose level
in all body fluids, including CSF and vitreous humour.
However, the dynamics of this decrease are quite unpre­
dictable. Attempts to combine glucose measurements with
lactate and other metabolites have not proven to be satis­
factory for routine diagnosis. At this time there is no
acceptable method for the detection of hypoglycaemia.
KETO NES
The detection of ketone bodies (acetone, acetoacetic acid
and (3-hydroxybutyric acid) is important for the diagnosis
of a variety of conditions. Infants and children have a ten­
dency to generate ketone bodies more readily than adults,
especially after a relatively short period of fasting. Children
shift their energy production to utilize fat quite promptly,
generating the ketone bodies in the process. The dipstick
test, which is based on the reaction of acetone and ace­
toacetic acid with nitroprusside in an alkaline buffer to
produce a purple colour, is quite sensitive. detecting as few
as 5 mg/dL in the urine. Unfortunately, the detection of
ketones in body fluids is rather non-specific and should be
interpreted with caution. It accompanies starvation/fasting,
vomiting and exercise, as well as any impairment of carbo­
hydrate metabolism, most notably diabetes mellitus in
----=-- -- ­
- .'
Endocrine disorders I 113
which it is associated with hyperglycaemia. On the other
hand, the finding of ketosis with hypoglycaemia is seen
in some of the glycogen storage disorders, such as Von
Gierke's disease. Perhaps even more important than the
finding of ketones in an infant or child with sudden death
is the absence of ketosis, especially if there is any evidence
of hypoglycaemia. The lack of ketones should trigger the
investigations for a possible (3-fatty acid oxidation disor­
der. Hypoglycemia with negative ketones is also present in
nesidioblastosis and other states of hyperinsulinism.
Addison's Disease: Adrenal Insufficiency
Sudden death in infants and children as a result of an
Addisonian Gisis is well described. 95 - 99 It may be due to
congenital adrenal hypoplasia or destruction of the adrenal
glands as a result of a number of different pathological
processes. These latter cases are more likely to occur in the
childhood and teenage years, 100 whereas the congenital forms
generally are present in infancy and may mimic SlDS. JOI
The gross and microscopic appearance of the adrenal
glands normally establishes the diagnosis. The usual meta­
bolic derangement accompanying an Addisonian crisis is
usually reflected by low sodium and chloride concentration
in the vitreous humour and very elevated potassium, often
in the range of the 'decomposition' pattern described
above. As salt wasting and dehydration are frequent fea­
tures, levels of urea nitrogen, and even creatinine, may be
elevated. If urine is available, the elevated sodium level
should be evident. This combination of dehydration ,
hyponatraemia and an elevated level of sodium in the urine
is generally indicative of adrenal insufficiency or salt­
losing nephritis. If both the adrenal glands and kidneys are
normal then one of the many causes of inappropriate
secretion of antidiuretic hormone should be entertained.
Random cortisol values will be markedly decreased in case
of adrenal insufficiency, regardless of the time of day. Cor­
tisol concentration in the blood is relatively stable for up to
18 hours after death, especially if the body is cooled within
4 hours of death.102 Reference values are method, age and
sex dependent, so that interpretation depends on compari­
son with the specific laboratory 's reference values.
Renal Failure
Renal failure in childhood may be quite occult, with few
clinical indications. Sometimes the family or the attending
physicians may not readily detect the growth failure asso­
ciated with declining renal function. Unexpected death in
such situations is quite uncommon but can occur. Gener­
ally the gross and/or microscopic appearance of the kid­
neys and the urinary system will establish the diagnosis.
The severity of the renal impairment can also be gauged by
114 I Biochemical investigations on post-mortem specimens
measuring the concentration of urea nitrogen and creati­
nine in the blood and vitreous humour. Urea nitrogen is
remarkably stable in these body fluids 56 ,IOJ-)05 and the
diagnosis of renal failure can readily be established. IOG , IU7
Of all analytes, urea nitrogen is perhaps the most stable,
showing no diagnostically significant changes over a
period of up to 5 days post-mortem. J08
Creatinine measurements in both blood and vitreous
humour are also quite stable,109 although some investiga­
tors have described modest increases in creatinine after
72-96 hours. 106,107 Levels in the vitreous humour are
slightly lower than serum values 54 but parallel the changes
that occur in renal fai lure. In well-established renal failure,
both the urea nitrogen and creatinine levels will be
markedly elevated. In states of acute dehydration and pre­
renal failure, the level of creatinine will be normal and
the renal nitrogen will be elevated. The ratio of urea
nitrogen to creatinine is a useful estimate of prerenal ver­
sus renal failure.
GENETIC METABOLIC DISORDERS
PRESENTING AS SUDDEN UNEXPECTED
DEATH
The list of inborn errors of metabolism presenting as sud­
den unexpected death continues to expand. These can be
broadly grouped into two large categories. The first group
involves enzyme defects leading to an increase of inter­
mediary metabolites proximal to the defect. One or more of
these compounds becomes toxic as they accumulate, pro­
ducing the clinical symptoms or unexpected death during
an acute metabolic event. Examples in this category
include many of the organic acidurias, aminoacidopathies,
urea cycle defects and carbohydrate intolerances. The sec­
ond category involves energy metabolism, based primarily
in the mitochondrion. The major organs affected by these
defects are the liver, skeletal muscle, heart and central
nervous system (CNS). A number of metabolic pathways
have been incriminated, including fatty acid oxidation,
mitochondrial transport and metabolism , glucose and
glycogen metabolism, and the lactic acidaemias. A variety
of clues shou ld be sought to help point the way to the most
appropriate work-up, as it is seldom possibl e to investigate
all the possible alternatives. Table 5.3 is helpful in directing
the physician and the pathologists to the appropriate cate­
gory for further work-up of infants who present with a n
acute ill ness in infancy. This is a very heterogeneous group
of disorders, so it is to be exp ected that there are a variety
of clinical phenotypes and presentations. While this chap­
ter focuses on the rare event of a sudden unexpected death
in an infant, the vast majority of cases do present during
life with one of the more frequent clinical manifestations.
During life the common laboratory signs suggestive of
an acute metabolic decompensation associated with some
of the inborn errors of metabolism include hypoglycaemia ,
an increased anion gap (due the accumulation of unmea­
sured acids), hyperammonaemia, ketonuria, lactic acidosis
or a respiratory alkalosis. Unfoliunately, neither ammonia
nor lactic acid measurements are reliable after death and
cannot be used. They tend to increase in both the serum
and the vitreous humour in a fairly lin ear fashion in the
first 24 hours, but there are no systematic studies of post­
mortem ammonia levels in infants and children evaluating
their role in diagnosis. Organic acid ana lysis in the urine or
v itreous humour is valid after death. Many changes occur
in the plasma amino acids (including an elevation of glut­
amine and alanine) and the neurotransmitters GABA and
(1-alanine. This makes post-mortem amino acid measure­
ments difficult to interpret.
Table 5.4 lists a collectio n of metabolic diseases that are
associated with 'energy production' and that have reportedly
caused sudden unexpected death. This list will undoubtedly
grow as defects in each step of the complex pathways
leading to the generation of adenosine triphosphate are
identified.
Fatty Acid Oxidation Defects
Tn the past couple of decades, a large num ber of defects
involving the 0 - oxidation of fatty acids have been recog­
nized. lll ,116-119 Some of these can be entirely asymptomatic
and present with unexpected deaths in infancy, usually
after a year of age. Some have occurred earlier under clin­
ical circumstances that closely mimic SIDS. 79 ,120- 132 It is
the unusual case that is truly confused with a classic or
typical SlDS. In most instances eli nical clues suggest that
the infant was not quite normal prior to death. Often the
death is later than the usual 2- to 6-month range of most
SIDS babies. Clinical indicators include lethargy, hypotonia
and an acrid odour to the breath (likened to the pungency
of 'smelly feet'). There is enough overlap between classical
SIDS and some case of acute metabolic decompensation to
justify consideration of this group of disorders always.
Certainly, any sudden unexpected death under the age of
5 years should raise the suspicion of a fatty acid oxidation
defect. The reported incidence of metabolic disorders
contributing to the number of SIDS v ictims or sudden
unexpected death after the age of 1 year varies very widely
in the available literature. The true incidence is unknown
b ecause there have been no satisfactory, systematic and
comprehensive studies with a large enough number of
cases. Moreover, metabolic defects are continuously being
uncovered and few studies have investigated the really
wide spectrum of currently known possibilities. Acquisition
bias afflicts some of the studies, as does poor case selection
and widely valying definitions of what constitutes a diag­
nosis. Having said all this there is no doubt tbat metabolic
diseases can and do result in sudden unexpected deaths in
infants and children who were otherwise fe lt to be grow­
ing, feeding and developing normally.
 Genetic metabolic disorders presenting as sudden unexpected death I 115

able 5.3 Genetic metabolic disorders which may present acutely during infancy

Acute presentation Major disorder Abnormal clinical test Disease example Confirming genetic
in infancy test
MSUD Plasma AAs
(l ethargy, coma, acidosis, Amino acid disorder NKH glycinemia Organic acids
hepatomegaly)
Tyrosina em ia-I Tissue enzymes
GTC Plasma/urin e AAs
Rule out: Urea cycle disorder NH 3t Argininosuccinate Orotic acid
• hypog Iycaem ia Citrullinaemia Tissue enzym es
• sepsis acidosis Methylmalonic
• CNS bleed Organic acids
Organic acid disorder NH3t (7) Propio nate/isova leric
• congestive heart Tissue enzym es
urine ketones Glutaric-II
• toxins
Renal loss Plasma/urine
HC0 3 loss 'Iytes
GI loss pH Et HC0 3

PDH def Pyruvate/lactate

Lactic acidosis lactate t Mitochondrial S Tissue enzymes
mitochondria
MERRF, MELAS Mu scle bx, DNA
Glucose/lactate
Glycogen storag e glucose ... GSD-I Et II Chol/trig/uric a.
Physical exam:
Liver bx, DNA
• norma I phenotype
• enlarged liver/spleen Organic acids
MCAD
• dysmorphic Fatty acid oxidation urine ketones ... DNA for MCAD, LCHAD
LCAD
Tissue enzymes

+ urine-reducing RBC gal-I-PUT

Galactosaemia Galactosaemia
substance DNA genotype
X-rays
Zellweger syndrome
Laboratory test: VLCFA/plasma log ens
Peroxi so mal disease Pseudo-Zellweger
CBC: Urine: Liver bx
syndro me
'Iytes pH Tissue enzymes
glucose spec. grav. X-rays
lactate ketones Tissu e enzymes
NH 3 GM-l gangliosidosis
Lysosoma l disease MPS excretion
MPS disorders
Oligosaccharides
EM of skin

CI~S, central nervous system; 'Iytes, electrolytes; MSUD, maple syrup urine disease; NKH, non-ketotic hyperglycinaemia; OTC, ornithine
transcarbamilase; GI, gastrointestinal; Glutaric-II, glutaric aciduria type II; PDH def, pyruvate dehydrogenase deficiency; MERRF, myoclonic epilepsy
with ragged red fibres; MELAS, mitochondrial encephalomyopathy, lactic acidosis, stroke-like episodes; GSD, glycogen storage disease; MCAD, medium­
chain acyl-CoA deficiency; LCAD, long-chain acyl-CoA deficiency; LCHAD, long-chain 3-hydroxyacyl-CoA-dehydrogenase; S, syndrome; MPS,
mucopolysaccharide; AAs, amino acids; RBC, red blood cell; bx, biopsy; VLCFA. very long cha in fatty acids ; EM, electron microscopy.

A minor illness, a short peliod of fas ting, an immuniza­
~on or other stress might trigger the acute metabolic decom­
pensation. There are a number of rather ordinary situations
in which physiological fasting might occur and not be obvi­
ous to the parents. The two classic examples occur at the
time of weaning and when an infant begins to sleep through
~he night. It is not unco mmon for in fants to present with
sy mptoms at such times and it is useful to specifically
~equest this history from the family. In a recent case of car­
:1itine transporter deficiency, it was surmised that the fasting
stress caused by poor breast-feeding without any supple­
mentalY formula feedings, and possibly the vegetarian diet
of the mother, might have triggered the metabolic crisis
leading to a neonatal death.11 4
The biochemistry of this group of disorders is complex
and incompletely understood. It involves the transfer of
fatty acids into the mitochondria via the carnitine transport
mechanism and their oxidation to produ ce adenosine
triphosphate (ATP). The oxidation of fa tty acids is an impor­
tant energy source in many organs, apart from the brain,
but especially the heart and skeletal muscles. In the normal
situation, fatty acids enter the cell, are activated to their
acyl-eoA esters, linked to carnitine by the action of carni­
tine palmitoyl transferase 1 and II, and then translocated
116 I Biochemical investigations on post-mortem specimens
Table 5.4 Disorders of eneigy metabolism
Disorder
Medium-chain acyl-CoAdeficiency
Long-chain acyl-CoA deficiency
Short-chain acyl-CoA deficiency
Long-chain 3-hydroxyacyl-CoA dehydrogenase deficiency
Carnitine palmitoyl CoA tranferase II defect
Mitochondrial acyl-carnitine/carnitine translocase defect
Cellular carnitine uptake defect
Trifunctional protein defect
Multiple acyl-CoA dehydrogenation defects (glutaric aciduria type II)
Carnitine transporter deficiency
Electron transport flavoprotein subunit ex defect
Electron transport flavoprotein subunit ~ deficiency
Electron transport flavoprotein-OQ deficiency
Mitochondrial ATPase
Cytochrome oxidase deficiency
across the inner mitochondrial membrane into the mito­
chondrial matrix. The carnitine is removed and the acyl­
CoA fatty acid then enters the 0-oxidation cycle. Repetitive
rounds of dehydrogenation by a number of different dehy­
drogenases (short, medium and long chain) depending on
the length of the fa tty acid, convert them to acetyl-eoA.
This then enters the respiratory (electron) chain that gener­
ates the ATP.
The diagnosis of abnormalities in 0-fatty acid oxidation,
which may be due to a defect at any of the many points in
the process, usually relies on the measurement of organic
acids in the urine, carnitine in the serum and the profile of
acyl-camitines. The in terp retation of the organic acid pro­
files on the urine as well as the acyl-carnitine profile requires
considerable experience and skill, and this is beyond the
scope of this chapter; those interested in pursuing the topic
further should consult the suggested references. 13J-lJ5
Depending on the results, more definitive tests involve the
measurement of specific enzyme activities or direct detection
of a gene defect.
The use of the serum measurement of dodecanoic acid
was initially promoted as a sensitive and specific screening
test for the common fatty oxidation defects. JJ6 Soon after
the original publication, issues around its specificity sur­
faced and it was reported to be much less specific than
originally hoped. 137 Until this issue is resolved with addi­
tional studies and a satisfactory cut-off value estab lished,
this method of screening cannot be relied upon. However,
it may alert the pathologist to those cases need ing addi­
tional work-up.
One of the most commonly identified disorders in this
group is medium chain acyl-CoA dehydrogenase deficiency
(MeAD). This is inherited as an au tosomal recessive, and the
most frequent mutation accounting for 90 per cent of cases
Reference(s}
110-112
111
111
111
111,113
111
111
111
114
111
111
111
115
is the G958 allele. JJ8, 1J9 Medium chain acyl- CoA dehydro­
genase defiCiency accounts for the majority of cases of sud­
den death due to a defect in fatty acid oxidation. It usually
presents during infancy with lethargy, a Reye-like syn­
drome, accompanied by hypoketotic hypoglycaemia. As
many as 40 per cent of infants are said to die during their
first episode. Generally, the liver shows considerable fatty
change, although there is great variation from case to case.
Moreover, lipid accumulation in the liver is quite non­
specific. Although much less common, the mutation for
long ch ain 3-hydroxacyl-CoA dehydrogenase deficiency is
also well describ ed. 140
Tests for the diagnosis of 0-fatty acid oxidation defects
include:
J. Urine: organic acids by gas chromatography/mass
spectroscopy to identify specific glycine conjugates;
urine acyl-carnitine profile to identify the specific
acyl-carnitine conjugates. If urine is not avai lable then
consider collecting either bile or vitreous humour. Both
bile l41 and vitreous humorlJl have been used
successfully to establish a diagnosis.
2. Liver: enzyme analysis for both medium- and long­
chain acyl-eoA dehydrogenase activity. 142 Immediate
freezing of tiss ues greatly retards the post-mortem
decrease in activity. However, the short-long chain
ratio remains constant so that the difference between a
true enzyme deficiency and post-mortem degradation,
in which both values will be decreased, can be
distinguished. An alternative approach to enzyme
activity is the analysis of various metabolites obtained
from the liver after methanol extraction and
ultracentrifugation. 14J This has been successfully
utilized for the detection of a number of the defects,
even in livers that were only obtained 72 hours after
 Technical considerations at the time of autopsy I 117

death. This technique assays a large number of the Table 5.5 Other genetic metabolic disorders claimed to be
farry aci d metabolites, together with glucose and associated with sudden death
lactate. The pattern of metabolites present can
Carbohydrate metabolism
sometimes establish a specific diagnosis.
Galacatosaemia
J. D A: medium-chain acyl G985 mutation - a PCR
Hereditary fructose intolerance
assay is available;144 coenzyme A dehydrogenase
Glycogen storage disease, type"
de ficiency (MCAD) - the mutation has been detected in
Fructose-l ,6-diphosphatase deficiency
D A obtained from a variety of different sources,
including frozen blood, llOdried blood spots,145 Urea cycle defects
archived paraffin embedded material, 146 and liver. 147 Carbamylphosphate synthetase
Detection of the mutation associated with long-chain Ornithine carbamyltransferase
J-hydroxyacyl CoA dehydrogenase (G 1528C) is also Argininosuccinate synthetase
possible. 14o Newborn screening cards are an excellent Argininosuccinate lyase
source for this testing. Many DNA tests can be carried
Amino acids
out directly from the newborn screening blood spots,
Tyrosinaemia
as can acyl-carnitine profiling.
Isovaleric acidaemia
~. Cultu red fibroblasts: globa.l [9,1O- 3Hl myristic acid
Non-ketotic hyperglycinaemia
oxidation assay - detects at least nine different fatty
Maple syrup urine disease
acid oxidation defects;117 specific enzyme analysis.
Organic acidaemias
Propionic acidaemia
Other Genetic Metabolic Disorders Methylmalonic acidaemia
3-Hydroxy-2-methylbu tyric aciduria
:'l,number of disorders involving amino acid metabolism Glutaconic acidaemia
may cause unexpected death, especially in the first Isovaleric acidaemia
6 months of life. Some of these are also associated with the Multiple carboxylase deficiency
ab normal excretion of certa in organic acids. Of these, Glutaric acidu ria type II
maple syrup urine disease is an excellent example. 148
Miscellaneaus
Table 5.5 provides a list of potential enzym e deficiencies that
Lysine protein intolerance

might cause sudden death; as will be noted, these involve

Biotinodase deficiency (ref. 152)

'ery disparate metabolic pathways. Some of the disorders

Glycerol kinase

may be classifIed as both an organic acidaemia and an

Menkes' disease (ref. 153)

a minoacidopathy. It must also be pointed out, however,

that this list has been comp iled from a variety of sources,
~e \V of which include the clinical or pathological details of
lhe individual conditions or cases.ll8.149-151 Finding the
pri mary documentation for all these claims has proved
quite difficult. In reality, sudden unexpected death is a very
ncommon manifestation of very rare diseases. Almost all
cases would have been suspected or diagnosed during the
patient's life.
TECHNICAL CONSIDERATIONS AT THE
TIME OF AUTOPSY
The following recommendations have been compiled from a
'ariety of sources. 3, 128.1 49.154 The extent and thoroughness of
ampling will depend on available resources, the relationship
between the forensic pathologist and the reference biochem­
:cal laboratory and the circumstances surrounding the
autopsy. Although some investigators recommend a stan­
dard protocol for all unexplained infant deaths this has gen ­
erally been limited to research investigations, especially
I .. hen there is specific expertise and interest in many of the
uncommon inborn errors of metabolism.
Vitreous Humour
Vitreous humour is obtained by puncturing the eye at the
outer ca nthus, using a 7- to 20-gauge needle and a small
syringe. Because of the viscous nature of the fluid, the fluid
flows slowly. Gentle suction should be applied, accompanied
by a large measure of patience. Vacuum tubes should not be
used as they generate too much force. An effort should be
made to collect all the avai lable fluid, which may be as little
as 0.5 mL in a neonate or as much as 3 mL in the young
adult. Fortunately, technology today allows a number of
measurements on these small volumes. The vitreous must be
centrifuged and the supernatant is then used for the analy­
sis. The fluid should be perfectly clear, with no evidence of
clouding or discolouration. No significant differences have
been shown in samples drawn simultaneously from each
eye, although this contention has been challenged. IS S Any
pre-ex isting eye disease that affects the vitreous humour
might confound the results. Fortunately, such diseases are
extremely rare in paediatrics. The recommendation to
118 I Biochemical investigations on post-mortem specimens
remove all the fluid is based on some expelimentaJ data that
demonstrated regional differences in the concentration of
certain analytes, Because of this, repetitive sampling of the
vitreous humour is not possible, Replacing the vitreous
humour with a volume of water or saline similar to the
amount aspirated can restore the shape of the eyeball by
maintaining its turgidity, a courtesy generally appreciated
by the mortician.
Vitreous humour can be used for the following tests; glu­
cose, ketones, sodium, urea nitrogen and organic acids. The
differences in levels of organic acids in the blood and vitre­
ous humour have been investigated. 111.156 Amino acid analy­
sis is also satisfactory if the vitreous humour can be obtained
in the first 24 hours after death and is rapidly depro­
teinized. 139 With time there is a gradual decrease in the
sodium and glucose concentration, whereas urea (measured
as urea nitrogen) shows a slight rise. The level of magnesium
is quite age dependent and also rises with time. The rate of
decrease of sodium is approximately 0.5-0.65 mEq/l per
hour. Calcium shows no consistent or predictable changes
after death. Abnormalities in calcium metabolism have not
been successfully diagnosed after death. Glucose declines at
a rate of around 0.21-1.6 mg/dl/hour. Urea rises in the vit­
reous humour at a rate of 0.081-0.62 mg/dL/hour. 56 Bicar­
bonate values are usually low in the vitreous humour, which
may be an artefact due to exposure of the sample to the air
with loss of carbon dioxide. Such is apt to occur when a
small volume of vitreous humour is placed in a large con­
tainer and if there is a long delay before the specimen is
analysed. Other compounds that have been measured in the
vitreous humour include amino acids,157 hormones l50 and
lactate. 23
Blood Collection
There is considerable variation in the concentration of cer­
tain analytes between the right and left side of the
heart. 159.J 60 Moreover, the influence of post-mortem hepatic
metabolism, such as glycogenesis, can affect a number of
carbohydrates. As the largest body of clinical biochemical
data is based on peripheral blood analysis during life, it
makes the most sense to use this source whenever possible.
Femoral or subclavian vein puncture will most accurately
reflect the ante-mortem values.
Blood should be centrifuged as soon as possible and the
serum separated and stored at - 20· C or -70· C. The longer
the serum is in contact with the formed blood elements, the
more difficult interpretation becomes. 16J Haemolysis will
affect the measurement of many analytes and should be
avoided. If red cells are going to be analysed (e.g. for the
presence of a haemoglobinopathy) then some blood should
be anticoagulated with ethylenediaminetetraacetic acid
(EDTA) in the standard fashion. Collecting blood in EDTA
or heparin is always valuable, as the specimen can be
immediately centrifuged and the plasma separated. Certain
laboratories prefer plasma for some of the analyses_ It is
important that the laboratory's reference ranges clearly
specify the type of specimen , as there may be considerable
differences between whole blood, serum and plasma, Blood
collected in EDTA can be used for the preparation of DNA.
Blood collected in heparin or acid-citrate-dextrose (ACD)
can be utilized for the preparation of white cells for the
analysis of lysosomal enzymes,
Urine Collection
The most convenient way to obtain urine is to perform a
bladder puncture once the abdomen has been opened. This
will ensure that the urine is not contaminated. In the event
that the bladder is empty, attempts can be made to collect
a small quantity of urine by puncturing the renal pelvises.
Other techniques, such as catheterization are more com­
plex, especially in small infants, as the appropriate size of
catheter is often not available. Expressing the urine from a
full bladder is also possible, although it is more difficult to
collect than a simple bladder tap. Urine should be stored in
sterile plastic or clean glass tubes at -70· C, in aliquots of
1-5 ml, until analysis.
Urine may not always be available. In one study of
infants dying of SIDS a substantial majority (60 per cent)
had no urine on opening the bladder. 132 If urine cannot be
aspirated from the renal pelvis then Bennett et al 131 recom­
mend swabbing the surface of the bladder with a cotton
swab, storing the swab at -20·C and then submitting it as
the specimen for organic acid analysis . However, most
laboratories will not have the expertise to process such
samples effectively.
In the event that the urine cannot be shipped to the ref­
erence laboratory in the frozen state, a few drops of 6N
hydrochloric acid or chloroform can be added as a preser­
vative in order to prevent the growth of bacteria. However
it is preferable to maintain the specimen frozen until the
time of analysis.
Urine can be utilized for the measurement of ketone and
organic acids. If urine is not frozen rapidly then the insta­
bility of some of the urinary organic acids will result in the
disappearance of some of the key groups, such as the
oxoacids. There are many potential artefacts and pitfalls in
the post-mortem analysis of urinary organic acids. The
most common and serious is bacterial contamination.
Freezing or inhibiting the growth with acidic compounds
such as hydrochloric acid will avoid this. Bacterial over­
growth can render the entire analysis meaningless. The key
for the successful interpretation of urinary organic acids is
close communication between the pathologists and the lab­
oratory worker who is performing the assay. Knowledge of
the infant/child's diet and medication history may be cru­
cial. For example, certain acids such as adipic, furoic and
tartaric acid may have a dietary origin. 'Nutramigen', a
common supplementary formula, may be associated with

an increase in 5-oxoproline. Infa nt feeding formulas that
a re supplemented with MCT oil will result in elevations of
the dicarboxylic acids: adipic, suberic and sebaric. Many
pharmaceuticals profoundly affect the analysis, although
this can generally be accounted for. Unfortunately, there
a re only a velY few organic acid disorders that can be diag­
nosed from a single urine analysis, even if one pays very
close attention to the pattern of additional metabolites.
Fibroblast Culture for Enzyme Analysis
If there is any hint of a possible metabolic disease an
attempt should be made to establish culture of fibroblasts.
This is generally successful with the cell culture media
available nowad ays for at least 48 hours after death. The
sk in should be cleaned with an alcohol-based disinfectant,
and well dried before excising a couple of2- to 4-mm J por­
lion of skin. These can usuaJJy be taken along one of the
sta ndard autopsy incisions. The specimen must contain
adequate amounts of dermis. Some investigators have used
:;"chilles tendon biopsies for the source of the fibroblasts;
this necessitates another incision but may decrease the
bacterial overgrowth that is common in skin biopsies, even
with apparently adequate disinfection. The transport media
should contain antibiotics to suppress the growth of any
contaminating bacteria. Generally, the biopsy in the trans­
port media should be maintained at room temperature prior
(Q transfer into the growth media. Once fibroblast culture
has been established, the cultures can be frozen and stored
:n liquid nitrogen. With such appropriate storage, fibro­
blast culture can be re-established for many years.
Should culture facilities not be immediately available,
[he skin biopsy itself should be frozen at -70°C and stored
in that manner. It is notable that successful culture has
been established from such stored tissue. 162
Tissue for DNA Analysis
In any case in which there is a suspicion of an inherited dis­
ease, tissue should be obtained for potential DNA analysis.
If only limited facilities are available, blocks (approximately
5 g) of tissue (such as thymus, lymph nodes, spleen or liver)
should be frozen at -70°C. An alternative to the direct stor­
age of tissue is to extract the DNA at the time of the autopsy
and to store the DNA. This probably results in better quality
DNA for later analysis. There are numerous standard proto­
cols for the preparation of DNA from tissue samples.
The 'Acute' Metabolic Autopsy
:c is not uncommon for infants and children to die after a
')fief period in hospital during which the possibility of a
Technical considerations at the time of autopsy I 119
metabolic disease is raised by the clinical and associated
laboratory findings . In such circumstances, performing the
autopsy immedi a tely after death can circumvent the prob­
lems of post-mortem change in tissue and body fluids.
Autopsy consent should be obtained ante-mortem if at all
possible, or immediately at the time of death to reduce any
delay, and the body should be expeditiously transported to
the morgue. Most children 's hospitals offer autopsy serv­
ices around the clock to expedite the diagnosis of such
cases. Tissue and body fluids should be collected on open­
ing the chest and abdomen and immediately processed to
limit any of the post-mortem changes. Tissue should not be
allowed to sit at room temperature for any length of time.
Three 5-mm cubes of liver, skeletal muscle, healt, CNS and
kidney should be snap-frozen in liquid nitrogen and stored
at -70°C. Urine and blood should be collected and stored
as mentioned above. Very small samples (1 mm maximum)
of liver, skin, CNS, healt and other organ should be fixed
in glutaraldehyde or other suitable fixative for electron
microscopy. Thymus/lymph node or spleen can all be used
for DNA studies as indicated above. Once the tissues have
been colJected and stored the autopsy can then proceed in
a more considered and 'leisurely" fashion . ln many institu­
tions the formal autopsy may even be delayed to the next
working day.
However, if it is not feasible to perform an autopsy
immediately after death then all is not lost. The largest sac­
rifices are in enzyme activity and electron microscopy.
Although celtain assays, such as enzyme activity determi­
nations, may become impossible to interpret, considerable
information can still be obtained from blood, bile, urine
and vitreous humour.1 6J -1 65 As has been demonstrated for
many of the fatty acid oxidation defects, analysis of post­
mortem liver, even as long as J days after death, can still
yield helpful and diagnostic information. 14J As DNA tests
for more and more of these diseases become available, it is
likely tha t t he need for such expeditious sampling may
diminish . Tandem mass spectroscopy, using blood spots on
newborn screening cards or bile, have also expanded the
potential diagnostic repertoire, as these a re quite stable a nd
can be mailed to a distant laboratory for analysis.
The following list shows the recommended specimens in
the majority of cases in which death has occurred many
hours previously. Not all of these specimens will be needed.
The autopsy findings and any clinical history will s hed
light on the possible aetiologies and will guide their subse­
quent use.
• urine (freeze 1-mL aliquots at -80' C);
- for organic acid analysis by gas
chromatography/mass spectroscopy;
- for amino acid analysis;
- ketones;
- reducing substances;
• blood (peripheral site) ;
- anticoagulated with EDTA (for haemoglobins) ;
- centri fuged, and the sentm stored at -70°C;
120 I Biochemical investigations on post-mortem specimens
• vitreous humour;
- for electrolytes, osmolality, glucose, BUN,
in victims of sudden infant death syndrome. J Pediatr 1981;
creatinine;
• bile;
- for metabolic studies;
• skin or Achill es' tendon biopsy for fibroblast culture;
• fro zen sections of liver, muscle, heart and kidney
sta ined for fat with Oil Red 0;
• spleen, thymus or lymph nod e (5-g portions) snap­
frozen in liquid nitrogen and stored at -70'C for DNA
analysis;
• l.iver, heart, skeletal muscle and CNS (5- to lO-g
portions for poss ible enzyme or metabolite analysis,
frozen at -70'c. 142 ,143
Cerebrospinal Fluid
A number of investigators have attempted to use CSF for
the measurement of a variety of compounds in an effort
to mitigate some of the problems associated with blood
analysis, or for a more direct determination of the changes
occurring in the CNS around the time of death. The use of
this body fluid has not found widespread general applica­
tion in most forensic laboratories, although there are
certainly some clinical situations when this should be con­
sidered. The change in concentration of many compound s
in the CSF with increasing post- mortem interval is similar
to that described in the vitreous humour. Enzymes, such as
creatinine kinase and lactic dehydrogenase, increase in a
linear fashion over time, as do potassium and lactate,
whereas glucose, pyruvate, sodium and chloride all
decline,I 66
REFERENCES
Evans WED. The Chemistry oj Death. Springfi eld, IL: Charles
C Thomas Publishers, 1963 .
2 Druid H, Holmgren P. A co mpilation of fatal and control
concentrations of dru gs in postmortem femoral blood.
J Forensic Sci 1997 ; 42:79-87.
3 Coe J1. Postmortem chemistry update. Emphasis on foren sic
application. Am J Forensic Med Patho11993; 14 :91-117.
4 Reay DT, Insalaco SJ, Eisele JW. PostmOl1em methemogl ob in
concentrations and their significance. J Forensic Sci 1984 ;
29: 11 60- 3.
5 Puschel K, Lockemann U, Bartel J. Postmortem investigation
of se rum myoglobin leve ls with special reference to electrical
fat a liti es. Forensic Sci fnt 1995; 72:171-7.
6 Zhu Bi, Ishida K, Quan L, Tanaguchi M et al. Postmortem
urin ary myoglobin leve ls with referen ce to the causes of
death. Foren sic Sci Int 2001; 115: 183-8.
7 Wyatt DT, Erickson MM, Hillman RE, Hillman LS. Elevated
thiamine levels in SIDS, non-SIDS and adults: postmortem
artifact. J Pediatr 1904 ; 104:585-8.
8 Rachut E, Rynbrandt OJ, Doutt TW. Postmortem behavior of
serum thyroxine, triiodothyronine and parathormone.
J Forens ic Sci 1980; 25:67-71.
9 Chacon MA , Tildon JT. Elevated values of triiodothyronine
99:758- 60,
10 Benjamin DR , Beckwith JB . Elevated values of
triiodothyro nine in victims of sudden infant death syndrome
(Letter). J Pediatr 198 2; 100:841.
11 Ross IS, Moffat MA, Reid IW. Thyroid hormon es in sudden
infan t death syndrome. Ciin (him Acta 1983; 129:151-5.
J2 Peterson DR, Green Wi, van Belle G. Sudden infant death
syndrome and hyperthiodothy roninemia: comparison of
neonatal and postmortem measurements. J Pediatr 1983;
102:202 -9.
13 Schwarz EH, Chaslow FI, Erickson MM et al. Elevation of
postmortem triiodothyronine in sudden infant death
syndrome and in infants who died of other ca uses: a marker
of previous health. J Pediarr 1983 ; 102 :200- 5.
14 Edston E, Druid H, Holm gren P, Ostrom M. Postmortem
measureme nts of thyroid hormones in blood and vitreous
humor co mbined with histo logy. Am J Forensic Med Pathol
2001; 22:78-83.
15 Arroyo A, Valero J, Marron T et al. Pericardial fluid
postmortem: Comparative study of natura l and violent deaths.
Am J Forensic Med Parhol 1998; 19:266-8.
16 Madea B, Kreuser C, Banaschak S. Postmortem biochemica l
examination of synovial fluid: a preliminary study. Forensic
Sci Int 2001; 118: 29 -35.
17 Haider M, Haider SQ. Assess ment of protein-calorie
malnutrition. Ciin Chem 1984 ; 30: 1206-99.
18 Tuten MB, Wogt S, Dasse F, Leid er Z. Utilization of
prea lbumin as a nutritional parameter. J Parenteral Enteral
NutI' 1985; 9:709-11.
19 Benjamin DR. Laborato ry tests and nutritional assessm ent.
Protein-energy status. Pediarr Clin N Am 1989; 36:139-60.
20 Maeda H, Fukita K, Oritani S et al. Evaluation of postmortem
oxymetry with reference to the causes of death. Forensic Sci
inr 1997; 83:201.
21 Sturner WQ, Sullivan A, Suzuki K. Lactic acid concentrat ions
in vitreous humor: their use in asphyxial deaths in ch ildren.
J Forensic Sci 1983; 28:222-30.
22 Jetter WW, McLea n R. Biochemical changes in body fluids
after death. Am J Ciil1 Path 1943; 13:178-85.
23 Jaffe FA. Chemical postmortem changes in the intra-ocular
fluid. J Forensic Sci 1962; 7:231-7.
24 Rognum TO, Saugstad 00. Hypoxanthine levels in the vitreous
humor: evidence of hypoxia in most infants who died of
sudden infant death sy ndrome. Pediatrics 1991; 07 :306-10.
25 Poulsen JP, Rognum TO, Hauge S et al. Postmortem
co ncentrations of hypoxanthine in the vitreous humor - a
co mparison betwee n babies with severe respiratory failure,
conge nital abnormalities of the heart, and victims of su dden
infant death sy ndrome. J Perinar iVIed 1993 ; 21: 153-63.
26 Madea B, Kaferstein H, Hermann N, Sticht G. Hypoxa nthine
in vitreous humor and cerebrospinal fluid: a marker of
postmortem interva l and prolon ged (vital) hypoxia?
Forensic Sci illt 1994; 65:19 - 31.
27 Belonje PC , Wilson GR, Siraka SA. High postmortem
co ncentration of hypoxa nthin e and urate in th e vitreous
humor of infants are not confined to case of sudd en infant
death syndrome. S AJrican Med J 1996; 86:8 27-8.
28 Carpenter](H, Bonham JR, Worthy E, Variend S. Vitreous
humour and cerebrospinal fluid hypoxanthine concentration
as a marker ot' ante mortem hypoxia in SIDS. J Clin Parhol
1993; 46:650-3 .
29 Opdal SH, Rogum TO, Vege A, Sa ugsta d 00. Hypoxa nthine
levels in vitreous humor: a study of influencing factors
in sudden infant death syndrome. Peciiatr Res 1990;
44:192-6.

30 Vege A, Chen Y, Opdal SH et al. Vitreous humo"
hy poxanthine levels in SlDA and infectious death. Acta
Paediatr 1994; 83 :634-9.
31 Jones KL, Krous HF, Nadeau J et al. Vascular endoth elial
growth factor in the cerebrospinal fluid of infa nts who die of
sudden infant death syndrome: eviden ce for antecedent
hypoxia. Pediatr 2003; 111 :358-63.
32 Putto A, Rllllskanen 0, Meurman 0 et al. C-reactive protein
in the evaluation of a febrile illness. Arch Dis Child 1986;
61 :24-9.
33 Peltola H, Homberg D. Rapidity of C-reactive protein in
detecting potential septicem ia. Pediatr InJ Dis 1983; 2 :374-6.
34 Benjamin DR , Siebert JR. C-reactive protein and prealbumin
in su spected sudden infant death syndrome. Pediatr Parhol
1990; 10:503-7.
35 Uhlin-Hansen L. C-reactive protein (CRP), a comparison of pre­
a nd post-mortem blood levels. Forensic Sci hrt 2001; 14:32-5.
36 Tsokos M, Reichelt U, Jung R et al. Interleukin-6 and
C-reactive protein serum level s in sepsis-related fatalities
during the early postmortem period. Forensic Sci Int 200 I;
119:47-56.
37 Fujita MQ, Zhu Bl, Ishid a K et al. Serum C-reactive protein
level s in po stmortem blood-an analysis with special reference
to the cause of deat h and survival time. Forensic Sci Int 2002;
130:160-6.
38 Chiesa C, Panero A, Rossi N et at. Reliability of procalcitonin
concentrations for the diagnosi s of sepsis in critica lly ill
neonates. Clin I11J Dis 1998; 26:664-72.
39 Monneret G, Lab a une JM, Issac C et al. Procalcitonin and
C-reactive protein leve ls in neo natal infections. Acta Paediatr
1997; 86:209-12.
40 Tsokos M, Reichelt U, Nierhaus A, Puschel K. Serum
pro calc itonin (PCT): a valuable biochemical parameter for the
postmortem diagnosis of sepsis. Jnt J Legal Med 200 1;
114:237-43.
41 Mei sne r M, Tschaikowsky K, Schanbel Set al. Proca lcito nin­
influence of temperature, storage, anticoagulation and arterial
or venous asservation of blood samples on procalcitonin
concentrations. Eur J C/in Chem Clill Biochem 1997 ;
35:597-60J.
42 Yunginger JW, Nelson DR, Squillace Dl e t al. Laboratory
investigation of deaths due to anaphylaxis. J Forensic Sci
199 I; 36:857 -60.
43 Ani sa ri MQ, Zamora JL, Lipscomb MF. Po st mortem diagnosis
of acute anaphylaxis by serum tryptase levels. Am J c/ill Parh
1993; 99:101-3.
44 Randall B, Butts J. Halsy JF. Elevated pos tmo rtem tryptase
levels in th e absence of anaphylaxis. J Forensic Sci 199 5;
40:208- I I.
45 Edsto n E. van Hage-Ha msten M. Beta -tryptase measurements
postmortem in anaphylactic deaths and controls. Forens ic Sci
lnt 1998; 93:135-42.
46 Schwartz HJ, Yunginger JW. Schwartz LB. Is unrecognized
anaphylaxis a cause of sudden unexpected death ? Clin Exp
Allergy 1995; 25:866-70.
47 Holgate ST. Walters C, Walls AF et a l. Th e anaphylaxis
hypothesis of sudden infant death syndrome (SIDS): mast cell
degranulation in cot death revealed by eleva ted concentrations
of tryptase in serum. c/in Exp Allergy 1994; 24: 1I 15-2 2.
48 Edston E. Gidlund E, Wickman M et a!. Increased mast cell
tryptase in sudden infant death: anaphylaxis, hypoxia or
artifact. C/in Exp Allergy 1999; 29: 1648- 54.
49 BuckJey MG, Variend S. Walls AF. Elevated serum
concentrations of beta- tryptase, but not alpha-tryptase, in
Sudden Infant Death Syndrome (SIOS). A n investigation of
anaphylactic mechanisms. C/in Exp Allergy 2001;
31: 1696-704.
References I 121
50 Schonheyder HC, Henriqu es U. Postmortem blood cultures.
Eva luation of sepa rate sampling of blood from the right and
left ventricle. APMIS 1997; 105:76-8.
51 Bentley AJ. Zorgani AA, Blackwell CC et al. Bacterial toxins
and sudden unexpected death in a young child. Forensic Sci
lilt 1997; 88:141-6.
52 Platt MS. Elin RJ, Hosseini JM, Smialek JE. Endotoxemia in
sudden infant death sy ndrome. Am J Forensic Med Pathol
1994; 15:261-5.
53 Coe JI. Postmortem chemistries on blood: particular reference
to urea nitrogen. electrolytes and bilirubin. J Forensic Sci
1974; 19:33-42.
54 Naumann HN. Postmortem chemistry of the vitreous body in
man. Arch Ophthalmol 19 59 ; 62: 356-63.
55 Coe JI. Postmortem chemistries on human vitreous humor.
Am J C/in Patholl969; 51 :741-50.
56 Swift PGF. Worthy E, Em ery JL Biochemical state of the
vitreous humour of infants at necropsy. Arch Dis Child 1974;
49:680-5.
57 Sturner WQ. Dempsey JL. Sudden infant death: chemical
a nalysis of the vitreo us fluid. J Forensic Sci 1973; 18:12.
58 Blumenfeld TA, Blanc WA, Devlin J. Postmortem v itreou s
humor concentrations of Na, K, CI, Ca and Mg in children.
Pediatr Res 1974 ; 8:356.
59 Zumwalt RE. Hirsch CS. Subtle fatal child abuse. Hum Pathol
1980 ; 11:167 - 74.
60 Emery JL, Swift PGF. Worthy E. Hyp ernatremia and uraemia
in unex pected death in infancy. Arch Dis Child 1974; 49:686.
61 Coe JI. Use of chemical determinations on vitreous humor in
forensic pathology. J Forensic Sci 1972; 17: 54 1-6.
62 Flear CTG, Gill GV Hyp onatremia : mechanisms and
management. Lancet 1981; ii: 26 -31.
63 Coe JI. Further thoughts and observations on postmol1em
chemistry. Forensic Sci Gazeffe 1975; 5 :2-6.
64 Coe JI. Vitreous potassium as a measure of the postmortem
interval: an historical review and clinical evaluation. Forensic
Sci Int 1989 ; 42:201-13.
65 Coe JI. Apple FS. Variations in the vitreous humor chemical
values as a result of the instrumentation. J Forensic Sci 1985;
30:828-3 5.
66 Sturner WQ. Dowdey ABC, Putnam RS. Dempsey JL.
Osmol ality and other chemical determinati on s in postmortem
vitreous humor. J Forensic Sci 1972; 17:387-93.
67 Devgun MS, Dunbar JA. Biochemical investigation of
vitreous: application in forensic medic ine esp eci ally in
relation to alcohol. Forensic Sci Jnt 1986; 31:27-34.
68 Hen ssa ge C. Krompech er T, Madea B et al. The Estimation oj
the Time Since Death in the Early Postmortem Period. 2nd
edn. London: Hodder Arnold, 2002.
69 Schleyer F. Determinations of the time of death in the early
postmortem interval. {VIeth ods oj Forensic Science, vol. 2.
Chichester: Inte rscience Publ., John Wiley and Sons, 1963.
pp. 253-93.
70 Furuno J . Momura S. Inorga nic phosphorus content in human
aqueous humor in rel ation to hours postmortem. Tohoku
J Exp Med 1976; 119:293-5.
71 Henry JB. Smith FA. Estimation of the postmortem interval by
chemical means. Am J Forensic Med Patho11980; 1:341-7.
72 Stephens RJ, Richards RG. Vitreous humor chemistry: the use
of po tassiu m concentration for the prediction of the
postmortem interva l. J Forensic Sci 1987; 32:503-9.
73 Adjudanti s G. Co utselinis A. Estimation of the time of death
by potass ium levels in the v itreous humor. Forensic Sci 1972 ;
1 :55-60.
74 Madea B. Henssge C. Honig W, Gerbracht A. References for
determining the time of death by potassium in vitreous
humor. Forensic Sci Jnt 1989; 40:231-43.
122 I Biochemical investigations on post-mortem specimens
75 Saugsatd OD, Oliasen B. Postmortem hypoxanthine leve ls in
the vitreous humour. Foren sic Sci Int 1978; 12:33-6.
76 Rognum TO, Hauge S, Oyasaeter S, Saugstad OD. A new
hioch emical method for the determina tion of postmortem
time. Forensic Sci Int 1991; 51 :131-46.
77 Madea B, Herm a nn N, Hengbe C. Precision of estimating the
time since death by vitreous potassium- co mparison of t'.NO
different equations. Forensic Sci Int 1990 ; 46:277-84.
78 James RA, Hoad ley PA, Sampson BG. Determination of
postmoltem interval by sampling vitreous humour.
Am J Foren sic Med & Patl7 ol1 997; 18:158-62.
79 Mason JK, Harkness RA, Elton RA, Bartholomew S. Cot deaths
in Edin burgh: infant feeding and socioeconomic factors.
J Epidemiol Commlln Health 1980; 34:35-41.
80 Coe JI. Postm ortem periph era l blood glucos e and
cardiopulmonary resuscitation. Forensic Sci Gaz 19 75; 6:1-2.
81 Gormsen H, Lund A. diagnostic value of postmortem blood
glucose determinations in cases of diabetes mellitus . Forensic
Sci Int 1985 ; 28: 103-7.
82 Peele t C, Picotte P, Jobin F. The use o f vitreous humor levels
of glucose, lact ic acid and blood levels of acetone to establ ish
ante-mortem hyperglycemia in diab etics. Foren sic Sci Inr
1994; 65:1-6.
83 Coe JI. Comparative postmortem chemistries of vitreous humor
befo re and after embalming. J Forensic Sci 1976; 21 :583-6.
84 Bray M. The eye as a chemical indic ator of environme ntal
temperature at the tim e of death. J Forensic Sci 1984;
29:396-403.
85 Coe JI. Hypothermia: auto psy findin gs and vitreous gl ucose.
J Forensic Sci 1984; 29:395-8.
86 Chen C, Glagov S, Mako M et al. Postmortem glycosylated
hemoglobin (HbAlc): evide nce for a history of diabetes
mellitu s. Alln Ciin Lab Sci 1983; 13 :407-10.
87 Winecker RE, Hammet-Stabler CA, Chap man J F, Ropero-Miller
JD. HbA I c as a postmortem tool to identify glycemic contro l.
J Forensi c Sci 2002; 47: 1373-9.
88 Khuu HM , Robinson CA, Brissie RM , Konrad RJ. Postmortem
diagnosis of unsuspected diabetes mellitus established by
determination of dec ede nt's hemo globin Alc leve l.
J Forensi c Sci 1999; 44:643-6.
89 Goulle JP, Lacroix C, Bouige D. Glycated hemog lobi n: a useful
post-mortem referen ce marker in determining diabetes. J
Forensic Sci 2002; 128 :44-9.
90 Osuna E, Garcia-Villora A, Perez-Carceles!VI et al. Vitreous
humor fructosamine concent rations in the autopsy diagnosis
of diabetes mellitus. fnt J Legl Med 1999 ; 112:275-9.
91 Osuna E, Garcia-Villora A, Perez-C arcel es M et al. Glucose
and lactate in vitreous humor compared with the
determination of fructosamine for the post-mortem diagnosis
of diabetes mellitus. Am J Forensic Med Pathol 2001;
22:244-9.
92 DiMaio V JM, Sturn er WQ, Coe JI. Sudden and unexpected
deaths after the acute onset of diabetes mellitu s. J Foren sic
Sci 1977 ; 22: 147-51.
93 Sturner WQ, DiMaio VJM. Sudden and unexpected death
following the acute onset of diabetes mellitus (Abstract).
Foren sic Sci 1975; 5: 168.
94 Rozin L, Perper JA , J a ffe R, Drash A. sudden unex pected
death in childhood due to unexpected diabetes mellitus.
Am J Forensic Med Patho11994; 15:251-6.
95 Russel MA, Opitz JM, Visekul C et al. Sudd en infant dea th s
due to conge nital ad renal hypopla sia. Arch Pathol Lab Med
1977 ; 101:168-9.
96 O'Don ohoe l\rv, Holland PDJ. Fam i.lial con genita l adrenal
hypopl asia. Arch Dis Child 1968; 43:717- 23.
97 Jindrich EJ . Adrenal hypofullction and sudden death.
J Forellsic Sci 19134 ; 29:930-3.
98 Sperling !VIA, Wolfson AR, Fisher DA. Congenital adrenal
hypoplasia: an isolated defected of organogenesis. J Pedjatr
1973; 82:444-9.
99 Batch JA, Montalto J , Yong ABW et a l. Three cases of
congenita l adre nal hypopl asia; a cause of salt wasti ng and
mortality in the neonatal period. J Pediatr Child Hea lrh
1991; 27:108-12.
100 Molander N. Sudden natu ral death in later childhood and
adolescen ce. Arch Dis Child 1982; 57 :572-6.
101 Favara BE, Franciosi RA, Miles V. Idiopathic adrenal
hypoplasi a in children. Am J Clin Pathol 1972; 57:287-96.
102 Finlayson NB. Blood cortisol in infants and adults: A
postmortem study. J Pediatr 1965; 67 :248-52.
103 Sa nders FW. The preservatio n of blood for chemical analysis.
J Bioi Ch em 1923; 58:1-15.
104 Paul JR. Postmortem blood chemical determinations. Bull
Ayer Ciin Lab Pen Hosp 1925; 9:51-62.
105 Leahy MS, Farber ER. Postmortem chemistry of human
vitreous humor. J Forensic Sci 1967; 12:214-22.
106 Levonen E, Raekallio J, Saikkonen J. Postm orte m
determination of blood creatinine and urea . J Forensic Sci
1963; 10:22-9.
107 Jensen OM. Diagnosis of uremia postmortem. Dan Med Bull
1969; 8(Suppl.):1-97.
108 Coe J1. Postmortem chemistries on human vitreous humor.
Am J Clin Pathol 1969 ; 51 :741 - 50.
109 Pol ayes SH, Hersh ey E, Ledered M. Postmortem blood
chemistry in renal disease. Arch liltem Med 1930;
46:283-9.
110 Bennett MR, Rinald o P, Millington DS, Tanaka K, Yokota I,
Coales PM. Medium-chain acyl-CoA dehydro ge nase
deficiency; postmortem diagnosis in a case of sudden infa nt
death a nd neonata l diagnosis o f an affected sib ling. Pediatr
Patlzol 1991 ; 11 :8139-95.
111 Bennet l\IlJ, Allison F, Pollitt RJ, Variend S. Fatty acid
oxidation defects as a cause of unexpected death in infa ncy.
Prog Clin Bioi Res 1990; 321 :349-64.
112 Duran M, Hofkamp M, Rhead WJ et al. Sudd en child death
and 'healthy' affected family members with medium- cha in
acyl- coe nzyme A dehydrogenase deficiency. Pediatrics 1986;
78:105 2-7.
113 Chalmers RA, Stanley CA, English N, Wigglesworth JS.
Mitochondrial camitine-acylcarnitine transl ocase deficiency
presenting as sudd en neonatal death. J Pediatr 1997;
131:220-5.
114 Rinaldo P, Stanley CA, Hs u BY et al. Sudden neonatal death
in carnitine transp orter deficiency. J Pediatl" 1997 ;
131:181-2.
115 Di onis i- Vici C, Seneca S, Zeviani Met al. Fulminant Leigh
syndrome and sudde n unexpected death in a family with the
T9176C mutation of the mitochondrial ATPase 6 gene.
J Inherit Metab Dis 1998; 21:2-8.
116 Ha rpey J-P, Charpentier C, Coude M et al. Sudden infant
death sy ndrom e a nd multiple acyl-coenzyme A
dehydrogenase deficiency, ethylmaionic-adipic aciduria or
system ic carnitin e deficiency. J Pediatr 1987; 110:881-4.
117 Lundemose JB, Kolvraa S, Gregersen N et al. Fatty acid
oxidation disorders as primary ca use of sudden and
unexpected death in infants and yo un g children: an
investigation performed on cu ltured fibroblasts from
79 children who died aged between 0-4 years. J Ciin
Palhol Mol Patho11997; 50:212-17.
1113 Emery JL, Va riend S, Howat AJ, Vawter GF. In vestigation of
inborn errors of metabolism in unexpected infant death.
Lancet 1988; ji:29-31.
119 Hale DE, Benn ett MJ. Fatty acid oxidation disorders: a new
class of metabo.lic diseases. J Pediatr 1992; 121: 1-11.

120 Ho wa t AJ, Bennet MJ, Variend S, Shaw L. Medium cha in
fatty acyl -coenzyme A dehydrogenase pres en ting as the
sudd en infant death. BM] 1984; 288:976.
121 Howat AJ, Bennett MJ, Variend S et al. Defects in the
metabo lism of fatty acids in the su dden infant death
syndrome. BlVlJ 1985; 290:1771-3.
122 HollOn JB , Allen JT, Green CA et a l. Inherited metabolic
disease in sudden infant death sy ndrome. Arch Dis Child
1991; 66:13)5-17.
123 Miller ME, Brooks JG , Forbes N, Insel R. Frequency of
medium-cha in acyl-CoA dehydrogenase G-985 mutation in
sudden infan t death syndrome. Pediatr Res 1992; 31:305-7.
124 Chinsky J, Tolsm a T, Cowan T, Blitzer M. MCAD deficiency
in SIDS. Am] Hum Gen 1991 ; 49(Suppl.):AI83.
125 Arens R, Gozal D, Jain K et al. Prevalence of medium-chain
acyl-CoA dehydrogena se deficiency in sudd en infant death
syndrome.] Pediatr 1993; 122:71 5-18.
126 Lemieux B, Giguerre R, Cyr D e t al. Sc reenin g urine of 3­
week old newb orns: la ck of associ at ion betw ee n sudden
infant death synd rome and some metabolic disorders.
Pediatrics 1993 ; 85:986-8.
) 27 Bonham JR, Dow ning M. Metabolic deficiencies and SIDS.
] Clin Pathol 1992; 45(Suppl.):33-8.
128 Bennett MJ, Powell S. Metabolic disease and sudden,
unexpec ted death in infancy. Hum Pat/1011994; 25:742-6.
129 Roe CR, Millin gto n DS, Maltby KF. Recognition of med ium
chain acy l CoA dehydrogenase deficiency in asy mptomatic
siblings of children dy ing of s udden infant death or Reye­
like syn drome.] Pediatr 1986; 108:13-18.
130 Rebuffat E, Sottiaux M, Goyens D et al. Sudd en infant death
syndrome, as first expression of a metabolic disorder. In
Sch aub J, Van Hoof F, Vis HL (eds) Inborn Errors of
Metabolism. New York: Vevey/Raven Press, Nestle Nutri tion
Workshop Series 24, 1991, pp. 71-80.
13 1 Bennett MJ, Hale DE, Coates PM, Stan ley CA. Postmortem
reco gnition of fatty acid ox idation disorders. Ped iatr Pathol
1991 ; II :365-70.
132 Bennet MJ, Allison F, Pollitt RJ, Variend S. Fatty acid
oxid ation defects as causes of unexp ected death in infancy.
In Tanaka K, Co ates PM (eds) Fatty Acid Oxidation: Clinical,
Biochemical and Molecular Aspects. New York: AJan R Li ss,
1990, pp. 349-64.
J 33 Blau N, Duran M, Blaskovics ME (eds) . Physicia n 's Guide to
the Laboratory Diagnosis of Metabolic Diseases. London :
Chapman and Ha ll Medical. 1996.
134 Scriver CR, Beaudet AL, Sly WS, Valle D (eds). The Metab olic
and /VIolecular Bases of Tnh erited Disease. New York:
McGraw -Hili, 1995.
135 Applegarth DA, Dimmick JE, Hall JG (eds). Organelle
Diseases. London: Chapman and Hall Medical, 1997.
136 Kemp PM, Little BR , Bes t RO, Dawson DB. Whole blood
leve ls of dodecanoic acid, a routinely detectable fore nsic
marker for a gen etic disea se often misd iagnos ed as
Sudden Infant Death Syndrome (SIDS): MCAD deticiency.
Am] Forensic NIed Path 1966; 17:79-82.
137 Ross KF, Guileyardo JM, Bennet MJ, Ba rnard JJ. Comment
on wh ole blood levels of dod eca noic acid, a routin ely
detectabl e foren sic marker fo r a genetic disease ofte n
misdiagnosed as Sudden Infant Death Sy ndrom e (SlD S):
MCAD deficiency. (Letter). Am ] Forensic Med Path
1966;17:349-50.
138 Blakemore AJF, Singleton H, Politt R et at. The frequency of
the G985 MCAD mutation in the gen eral populati on. Lancet
1991; 337:298-9.
139 Coates PM. Histori cal perspecti ve of medium chain acyl-CoA
dehydrogena se deficiency. A decade of discovery. Prog Clin
BioI Res 1992; 375:40 9-23.
References I 123
140 !jIst L, Ruiter JP, Hoovers JM et al. Common missense
mut ation G 1528C in long-chain 3-hydroxyacyl-CoA
dehyd roge nas e deficiency. Characterization and expression
of the mutant protein, mu tation analysis on geno mic DNA
and chro mosomal localization of the mitochondrial
trifunct ional protein alph a subunit gene.] Clin Invest 1996;
98: 1028-33 .
141 Rashed MS, Ozan d PT, Bennett MJ et al. Inborn errors of
metabolism dia g nosed in sudden infant death case by
acylcarnt ine analysis of postmortem bile. Clill Chem 1995;
41:1J09-14.
142 Hale DE, Cornell JE, Bennett MJ. Stability of lon g- chain and
short-chain 3-hydroxyacyl -CoA dehydroge nase activity in
the postmortem li ver. Clin Chem 1997 ; 43:273 -8.
143 Boles RG, Martin SK , Blitzer MG, Rin a ldo P. Bio chemical
diagnos is of fatty acid oxid atio n disord ers by metabolite
analysis of postmortem li ver. Hum Pathol 1994;
25:733-4.
144 Gregersen N, Blakemore AIF, Winter V et al. sp eci fic
diagnosis of medium-chain acyl-CoA de hyd rogenase
(MCAD) deficiency in dried blood spots by a polymerase
chain reaction (PCR) assay detecting a point mutat ion (G9 85)
in the MCAD gen e. Clin Chim Acta 1991 ; 203 :23-4.
145 Matsubara Y, Na risawa K, Tada K. Medium-chain acy l-CoA
dehydrogenas e deficiency. Molecular aspects. Eur] Pedia tr
1992; 151:154-9.
146 Ding J-H , Roe CR , lafolla AK, Chen Y- T. Medium-chain
acy l-CoA dehydrogenase deficiency and sudden infant
death (Letter). N Engl] Med 1991; 325: 6 1-2.
147 Kelly DP, Hale DE, Rutledge SL et al. Molecular basis of
inherited med ium-cha in acyl-CoA dehydrogenase deficiency
ca using sudden child death.] Inherit Metob Dis 1992 ;
15:171-80.
148 Hallock J, Morrow G, Karp LA, Barness LA . Postmoliem
diagnosiS of metabolic disorders. The findings of maple
syrup urine disease in a cas e of sudden and unexp ected
dea th in infancy. Am] Dis Child 1969; 118:649- 5 1.
149 Byard RW, Cohle SD. Sudden Death ill Infancy, Childhoo d
and Adolescence. Camb ridge: Cambridge University Press,
1994.
150 Clayton PT, Hy land K, Brand M, Leonard N. Mitochondrial
phosphoenolpyruvate ca rboxykinase deficiency.
Eu r] Pedia tr 1986: 145:46- 50.
151 Norman MG , Taylor GP, Clark e LA. Sudden, unexpected,
natural death in childhood. Pediatr Patlwl 1990; 10:7 69 -84.
152 Burton BK, Roach ES, Wolf B, Weissbecker F. Sudd en dearh
associated w ith biotinid ase deficiency. Pediatrics 1987;
79:482-3.
153 Danks DM, Campbell PE, Stevens BJ et at. Menke's kinky
hair syndrome. An inherited defect in copper metab olism
with widespread effects. Ped iatrics 1972; 50: 188-20 t.
154 Forres t AR (ACP Broadsheet no. 137, April 1993). Obtaining
samples at postmoliem examin a tion for toxicological and
biochemical analyses. ] Clin Pathol 1993 ; 46: 292-6.
155 Pounder DJ , Carson DO, Johnston K, Orihara Y. Electro lyte
concentration differenc es between the left and right vitreous
humor sampl es. ] Forensic Sci 1998; 43:6 04 -7.
156 Bennett MJ, Ragni MC, Hood 1, Hale DE. Comparison of
postmortem vitreous and urinary organic acids. Ann Clill
Biochem 1992; 29:541-5.
157 Pallick W JA, Logan RW. Free amino acid content of the
vitreous hum or in cot deaths. Arch Dis Chil d 1988; 63:660-2.
158 Chong APY, Aw SE. Postmortem endocrine levels in the
vitreo us hum our. Ann Acad Med (Singapore) 1986; 15 :606-9.
159 Hill E. Significance of dex trose and nondextrose reducing
substa nces in postmortem blood. Arch Pat/IOI 1941;
32: 4 52-73.
124 I Biochemica l investigati ons on post-mortem spec ime ns
160 Lythgoe AS. The act ivity of lactic dehydrogenas e in ca daver
sera: a comparison of different sampl ing sites. Med Sci Law
1980; 20:48-53.
161 Zhang DJ, Elswic k RK, Mill er WG, Bailey JL. Effect of
serum-clot contact time on clinical chemistry laboratory
results. Clin Cliem 1998; 44:1 325-33.
162 Fowler KJ . Storage of skin biopsies at -70°C for futu re
fibroblast cu lture. J Ciin Patlioil984; 37:1191-3.
163 Bennet MJ , Rinaldo P. The metabo lic autopsy comes of age.
Ciin Chem 200 1; 47 :11 45-6.
164 Wilcox RL, Nelso n CC, Stengel P, Stei ner RD. Postmortem
screening for fat ty acid oxid atio n disorders by analysis of
Guthri e cards with tandem mass spectrometry in sud den
unexpected death in infancy. J Pediatr 2002; 141 :833 -6.
165 Chace DH, DiPerna JC, Mitchell BL et al. El ectrosp ray
tandem mass spectrometry for analysis of acylca rnitines in
dried postmortem blood spec imens collected at autopsy from
infants with unexplained cause of death. Clin (hem 2001;
47: 1166- 82.
166 Karke la JT. Critical evaluation of postmortem changes in
hum a n autopsy cistern al fluid . Enzymes, elect rolytes, acid­
base balance, glucose and glycolysis, fre e amino acids and
amm onia. Con-elation to total brain isch emia. J Forensic Sci
1993; 38 :603- 16.
Further Reading
Applegarth DA, Dimmick JE, Toone JR. Laboratory detectio n of
metabolic di sease. Pediatr C/in North Am 1989; 36 :49-66.
Applega11h DA, Dimmick JE , Hall JG reds). Organ elle Disms es.
Lo nd on: Chapma n and Hall Med ical, 1997.
Beny J, Alibon e E, McKeever P et al. The contribution of ancillary
pathology tests to the in vestigation of sudden infant death. In
Fleming P, Blair P, Bacon C et al reds} Sudden Unexpected Dea ths
ill InJancy. The CESDI SUDI Studies. Norwich: The Stationery
Office, 2000, pp. 97 -112.
Bla u N, Duran M, Blaskovics ME (eds). Physician's Guide to the
Laboratory Diagnosis oJ M etabo lic Diseases. London: Chapman
and Hall Med ical, 1996.
Brettell TA, Safers tein R. Forensic science. Anal Chem 1995;
67:27 3R- 94R.
Byard RW. Sudden Death in inJancy, Childhood and Adolescence, 2nd
edn. Cambridge : Cambridge University Press, 2004.
Coe J1. Postm ortem chemistries on human vitreous humor.
Am J Clin Pathol 1969; 51:741 - 50.
Coe J1. Postmorte m chem istlY of blood, cerebrospinal fluid and
vitreous humor. In Ted esc hi CG, Eckert WG, Tedesc hi LG (eds)
Foren sic Medicine, vol. 2. New York: WB Saunde rs, 1977.
Coe J1. Post mortem biochemistry of blood and vitreous hum our in
paediatric prac tice. In Mason JK [ed.} Paediatric Forensic
1\1edicine and Pathology. London: Chapman and Hall, 1989, pp.
191- 203.
Coe J1. Post mortem chem istJy upd ate. Emph asis on fore nsic
appli catio n. Am J Forensic Med Pathol 1993; 14:9 1-117.
Scriver CR, Beaudet AL, Sly WS et al (eds). The Metabolic and
Molecular Bases oJ Inh erited Disease. New York: McGraw-Hili,
1995.

OCULAR INVOLVEMENT IN
NON-ACCIDENTAL INJURY
Harry Willshaw
I t roduction 125
~i:ope of ocular and adnexal injury 125
INTRODUCTION
SCOPE OF OCULAR AND ADNEXAL INJURY
_\11 abused child is most often the victim of damage
'n flicted by a parent or carer. Although both self-mutila­
TIo n and Munchausen's syndrome by proxy do occur in
children, I they are rarities in the setting of child abuse
globally.
The practice of child abuse was first challenged legally
in New York in 1871 2 using legislation written to prevent
cruelty to animals. Since that time the incidence of child
abuse, or at least its identification, has risen to such an
ex te nt that in 1962 it was considered the major cause of
child death and maiming in the USA. 3
The frequency of ocular involvement in non-accidental
injury (NAI) has been variously reported at between 5 and
6 1 per cent. 4 ,5 Eye damage is usually seen in association
':;ith head and facial injuries. 6 In reported series in which
- e injuries resulted in death, the rate of ocular involve­
ment is high/- 9 suggesting that ocular involvement is a
sig nificant marker for severe neurological involvement.
Fatal neurological involvement is particularly common in
'ery young victims; all of 13 fatalities attributable to shak­
in g reported in 1987 were of children aged 24 months or
less . 10
A proportion of children will suffer life-long visual
andicap as a consequence of NAl. The likelihood of long­
lerm visual sequelae from a n injury is usually directly
re lated to the extent of ocular damage; however, it must be
remembered that some children are blinded by their neuro­
logical injuries with no evidence of persisting ocular
pathology. II
I CHAPTER 6 I
Fundus haemorrhages 128
References 134
Fundus changes are often considered the hallmark of ocu­
lar involvement in NAl, but in fact any ocular or adnexal
structure may be damaged.
Eyelids
Swelling, laceration and burning of the eyelids have all
been described. 12 Lid bruising and swelling are relatively
common events in a young life and should not be
overinterpreted, unless they are seen in combination with
other injuries that are suggestive of abuse. In children
under the age of 12 years, lid injuries are most commonly
caused by implements. The characteristic pattern of belt
buckles etc. may be obvious in the injured site. Typically,
after the age of 12, the injuries are inflicted with fists and
feet. 5
Burns to the eyelids are an occasional feature of ab use.
Unlike their accidental counterpart, non-accidental facial
burns tend to be deep and multiple. At 60°C, heat must be
applied to skin for at least 3 seconds to produce even a
first-degree burn,13 so that multiple deep cigarette
burns, as seen in Figure 6.1, are not compatible with the
child having inadveliently brushed against the hand of a
smoker.
Occasionally more exotic agents of abuse are applied to
the lids. Recently, a case has been reported in which super­
glue was deliberately applied to the eyelids, 14 resulting in a
short period of permanent lid closure.
126 l Ocular involvement in non-accidental injury
Figure 6_1 Cigarette burns on the eyelids - inflic ted injury.
External Eye
Subconjunctival haemorrhage is ano ther rela tively co mmon
finding (Fig. 6.2) associated with facial IDJUlY or
Valsalva manoeuvre, and does not in itself indicate
deliberately inflicted damage. On the other hand, persisting
corneal damage is a great rarity in childhood. In otherwise
healthy children with no cause fo r chronic corneal exposure
(e.g. previous ptosis surgery), corneal epithelial defects usu­
ally settle quickly. Persisting epithelial damage may well be
caused by chemical in stillat ion into the eye. This is particu­
larly likely to involve the epithelium of the lower hal f to
one-third of the cornea and conjunctival cui de sac. This is
because an intact Bell 's phenomenon causes t he eyes to roll
upwards as the lids are pulled apart to allow instillation.
We have treated a child with unexplained bilateral
corneal clouding accompanied by total epithelial loss. The
corn ea s healed well when the child was isolated from her
parents, and her father subsequently admitted to repeatedly
spitting methadone into the child's eyes. Any type of harmful
material may be instilled into the eye - no paliicular sub­
stance is more common than oth ers. In 1987 the app licatio n
of hot peppers (capsicum and capsaicin) to the conjunctival
fornices was reported as a type of child abuse. 15
Anterior Segment
The presence of anterior seg ment ocular injUlY, in the form
of iris or lens damage, tends to suggest severe impact
trauma and carries a relatively poor visual prognosis (Fig.
6.3). Pupil sphincter rupture (which manifests as slight
irregularities in th e pupil and tears at the pupil margin),
hyphaema , recession of the drainage angle, lens sublux­
ation and cataract leading to glaucom a have all been
described in a single child. 16 These events usu ally follow
direct injury to the gl ob e,17 but can be seen after the rapid
deceleration caused by an impact injury.
As with all of the ocular features of child abuse, the
changes seen are not pathognomonic of NAJ and appropriate
Figure 6.2 Bilateral subco njunctival haemorrhage and
periocular bruising following assault.
syste mic investigations are required. This is palii cularly so
for lens subluxation, in which investigations must exclude
Marfa n's syndrome, homocystinuria, sulphite oxi dase defi­
ciency and Weil-Marchesani syndrome.
When lens damage does occur, it may be possible with
the slit lam p to detect small tears in the anterior lens cap­
sule. These tears allow t he ingress of aqueous humour into
the protein of the lens cortex leading to delayed cataract
formation. The resultant cataract may not become clinic­
ally app arent for 4 or 5 days after the original injury.
Posterior Segment
Features of NAI in the posterior segment are dominated
by the finding of fundus haemorrhages. However, retinal
detachments can also occur. These may be caused by giant
retinal tears 18 or vitreous base avul sion 19 and may be
apparent immediately after the injury has occurred.
Because the vitreous gel is rel atively firmly adh erent to the
retina over the peripheral portion (the ora serrata), violent
movements of the vitreous gel exert strong tractional
forces on that area of the retina. As a consequence, the
retina may tear over more than 90 0 of its periphery, or the
vitreous base may tear away.
Alternatively, detachments may present later because of
traction from the condensation of traumatic vitreous haemor­
rhage. In one study, post-mortem examination revealed what
have been called haemorrhagic retinal detachments in up to
63 per cent of case. 20 In that particular report, they occurred
primarily in the retinal periphery and, though they pathologic­
ally appeared as retinal detachments, were seen ophthalmo­
scopically as accumulations of sub retinal blood. Peri macular
folding of the retina (Fig. 6.4), seen either at the time of firs t
assessment or at later fundus ex amination, is extremely sug­
gestive of NAJ. The folds can result from either direct head
trauma or violent shaking, but are almost unique to the eyes
of injured children. 21 .22
Late fundus manifestations of abuse include hole for­
mation in the retina, cysts, gliosis and scarring. 2J Features

Figure 6.3 Anterior segment damage. (a) Dislocated lens;
(b) tear of peripheral iris (iridodyalysis); and (c) traumatic
cataract.
Scope of ocular and adnexal injury I 127
Figure 6.4 Retinal haemorrhages and a peri macular fold.
Figure 6.5 Perineural haemorrhages wi thin the optic ne rve
(haematoxylin and eosin [HEtEj stain).
such as gliosis, pigmented dema rcation lines and fi xed ret­
inal folds are strongly suggestive of long-standing damage
and, if seen in the company of more acute features , such as
nerve fibre layer haemorrhages, indicate injuries of differ­
ent ages. 24 Many of these posterior segment features carry
a poor visual prognosis and, when combined with neuro­
logical features, can be used to produce a score that indi­
cates the likelihood of severe neurological or visual
morbidity2o - this is discussed in more detail later in this
chapter (see section on long-term visual sequelae, p. 134).
Visual Pathways
Neurological involvement in NAl is the major ca use of
morbidity and mortality and is often closely correlated
with the degree of ocular involvement (see section on asso­
ciated neurological injury). Direct involvement, particu­
larly of the optic nerve (Fig. 6.5) or visual cortex, may lead
to a poor visual outcome in children who show app arently
complete ocu lar recovery. In a group of three children with
unilateral retinal haemorrhages reported in 1998, two had
128 l Ocular involvement in non-accidental injury
profound long-term visual handicap because of optic nerve
or higher pathway damage,25 even though the fundus
involvement was strictly unilatera l and resolved.
Bleeding into the optic nerve sheath and peripapillary
sclera has been identified in autopsy specimens,26 and
complete avulsion of the optic nerve from the sclera may
occur, as may optic nerve transectionY
Optic nerve damage is often seen with accompanying
visual cortex injury. The visual cortex is vulnerable to
trauma via a number of mechanisms, including direct con­
tusion, raised intracranial pressure and the effects of poor
perfusion during periods of collapse.
Ocular Motor Control
Neurological disturbances of gaze have been reported but
are relatively uncommon. Abnormalities of the vergence
system, regularly seen after accidental head trauma,28 do
not feature significantly in the literature of NAJ. More
obvious cranial nerve palSi es, gaze palsies and nystagmus
can occur 29 .]O but are usuall y seen in the context of other
signi ficant neurological injuries.
Fin ally, in the most extreme circumstances, the clinician
may be faced with an auto-enucleation as a result of either
self-mutilation or NAI.'
FUNDUS HAEMORRHAGES
Though all ocular and adnexal structures can be affected
by child abuse, fundus bleeding is the most typical feature
and the one that has probably received most attention in
the world literature.
Types
The clinical appearance of fundus haemorrhages is depend­
ent on the site of ex travasation of blood from the retinal
blood vessels, and also upon the extent of bleeding. The
vessels involved are usually the capillaries or the post­
capillary venules and, because of the rich circulation at the
posterior pole of the eye, haemorrhages ten d to be most
plentiful there. However, characteristic bleeding in NAI
involves all the layers of the retina and can extend right
out to the ora serrata. Fundus examination of an NAI vic­
tim must therefore involve visualization of the whole
retina, usually necessitating pupil dilatation (see section on
examining the child with fundus haemorrhages, p. 136).
The physical characteristics of retinal haemorrhages are
dictated by the layer into which they occur.
NERVE FI BR E
Superficial haemorrhage into the nerve fibre layer of the
retina tends to be fl ame-shaped or splinter in type, and is
Figure 6.6 Intraocular haemorrhage: nerve fibre layer and 'blot'
haemorrha ges - some contain white centres (Roth spots).
derived from the superficial capillary bed. The sha pe of the
haemorrhage depends on the tracking of blood between the
nerve fibres, and for this reason the haemorrhages change
shape beyond the posterior pole. The nerve fibre bundles,
densely packed at the posterior pole, make a more open
network when one examines them more than 7 mm from
the optic disc.]' In these more peripheral locations, there­
fore, the superficial haemorrhages are rounded (Fig. 6.6).
Not infrequently they have white centres, and are known
clinically as Roth spots. These were first described in 1872 32
and thought at that time to be a pathognomonic feature of
subacute bacterial endocarditis. Subsequent clinical reports
have identified them in a variety of other disorders, includ­
ing leukaemia, anaemia, Behc;et's disease and hyperten­
sion. 3] They consist of a white centre with a round or oval
red surround, and it is our experience that they are a com­
mon form of superficial retinal haemorrh age in NAt
DEEP RETINA
The shape of deeper retinal haemorrhages is also dictated by
their anatomical relations. Bleeding from deep retinal capil­
laries is into the veliically orientated neural structures of
the outer retina. Physically, the smaller ones are described
as dot haemorrhages and may be mistaken for micro­
aneurysms. Larger areas of bleeding form blot haemorrhages,
which extend throughout the thickness of the retina .
SUBRETI NA
Bleeding into a subretinal site may take one of two forms.
Blood may extravasate from the deep retina and accumu­
late betwee n the photo receptors and the retinal pigment
epithelium layer. In this location the haemorrhages are
generally large and tend to be situated at the posterior pole.
When they clear, they usually leave a residual visual defect
because of damage to both the choriocapillaries and the

o "erlying macula photoreceptors. If, on the other hand, the
bleeding originates from choroidal vessels, then it lies deep
iO the retinal pigment ep ithelial layer, appears slate­
oloured, has rounded margins and is associated with ele­
,;ation of the retinal pigment epithelium (RPE).34 Once
bl ood in this site clears, visual recovery tends to be much
more complete.
PR E- RETINA
Pre-retinal haemorrhages result from breakthrough of blood
fro m the retina into either the space between the nerve fibre
laye r and the internal limiting membrane35 or the space
bervveen the retina and the vitreous gel. The haemorrhage
may be extensive, and sedimentation of the solid components
of the blood gives rise to a fluid level that can shift under the
in fluence of gravity. Such haemorrhages are typically associ­
ated with significant intracranial bleeding in the form of sub­
dura l or subarachnoid haemorrhage. 36
POSTERIOR HYALO ID
fin ally, blood may break through the posterior hyaloid face
and give rise to a localized or diffuse vitreous haemor­
:hage. Vitreous haemorrh age is thought to happen 2-3
days after the initial injury,37 a feature that may have con­
siderable forensic significance. The vitreous will usually
clear spontaneously, but the process can be slow, taking
several months. Since these injuries typi cally affect chil­
dren in the sensi tive period for visual development,38 the
visual outcome following such severe bleeding is often bad
not only because of structural damage to the eye, but also
because of stimulus deprivation amblyopia. 39 Studies on
children with other causes of stimulus deprivation have
shown t hat relativel y short periods of asym metric interfer­
ence with visual input, at the right time in life, can lead to
significant amblyopia. 40 Figure 6.7 illus tra tes the distribu­
tion of haemorrhages within different layers of the retina.
No particular pattern of fundus haemorrhage is exclu­
sive to NAI, but it is frequentl y observed that the presence
of bleedi ng at all depths of the retina, throughout 360 0 of
the retina and extending from the posterior pole to the ora
serrata, is strongly suggestive of bleeding caused by the
severe acceleration and deceleration forces seen in
NAI. 41 - 43 If such extensive haem orrhages are seen in com­
bination with peri macular folds 44 or haemorrhagic retinal
cysts,45 then anything but the most severe accidental injulY
can almost certainly be excluded as the cause.
Mechanisms
The mechanism of blood vessel damage during NAI remains
debatable, and it may well be that several different mecha­
nis ms all contribute. The acceleration/deceleration forces gen­
erated by the vigorous shaking of an unsupported infant head
Fundus haemorrhages I 129
Figure 6.7 Retinal haemorrhage stained using haematoxylin
and eosin: a photomi crog raph shows recent haemorrhage at
several sites. pr, pre-retinal; nf, nerve fibre layer; ir, intraretinal.
(whiplash shaken baby) are substantial, but they increase
greatly with head impact against a solid surface. These forces
then give rise to a number of contributolY events:
• Blood vessels within the retina may shear because of
the relatively firm adherence between the vitreous and
the retina in yo ung children. The violently moving
vitreous transmits its mom entum to the layers of
the retina, tending to tear the retina and give rise to
splits within it (schisis) and blood-filled cystS. 45
Post-mortem examination of 190 eyes 46 demonstrated
that bleeding, even when it extended in to all layers of
the retina, invariably arose from the vessels in the
inner layers . This tends to support the view that
tangential forces are responsible for retinal shearing
and bleeding.
• Raised intraocular venous pressure results from
either raised intrathoracic pressure or raised pressure
within the optic nerve sheath. In either case, venous
stasis results, and subsequently retinal bleeding. There
are several potential causes of raised intraocular
venous pressure in an abusive situation.
A Purtscher type of retinopathy secondary to chest com­
pression has been postulated as the cause of retinal haemor­
rhage in child sexual ab use 47 or when an excessively firm
grip is taken of the infant chest during shaking. Chest com­
pression may well be a contlibutory factor in the re tinal
haemorrhages commonly found in the neonate after normal
vaginal delivery48,49 (see later). Raised intratlloracic pressure
during cardiopulmonary resuscitation (CPR) also has been
causatively linked with retinal haemorrhage. A number of
individual case reports testify to the possibility of fundus
haemorrhage following CPR, but large prospective studies
have failed to validate that association (see later) .
130 lOcular involvem ent in non-accidental inj ury
Bleedin g into the optic nerve sheath a nd th e perip apil­
lary sclera is a commonly reported feature of the more
seve re cases of NAl. Eleve n of s ixtee n children dy ing from
the central nervous syste m (CNS) complications of NAl
showed perineural bleeding. The bleeding was into the dis­
ta l portion of the optic nerve, indicating that it was not
simply forward extension of th e accompanying subdural
haemorrhage. 20 It is entirely possible that perineural bl eed­
ing raises the pressure within the optic nelve sheath, thus
ten ding to impede ven ous return, wi th fundu s bleeding
seco nda ry to the ve nous co ngestion.
Simila rly, raised intrac ranial pressure with pap illoedema
is a recog ni zed cause of retinal h ae morrhage with a sim ilar
mechanism. There is a close correlation between severe
CNS injury and prominent fundus bleeding. 50
A third possible source of fundus h aemorrh age, particu­
larly in the presence of subarachnoid haemorrhage, is
Terson's synd rome, 51 in which subhyaloid blood is seen
shortly after the development of a subarachnoid haemor­
rhage. Paton 51 postulated that the subarachnoid blood, under
pressure, passes forward through the la mina cribrosa to e nter
the eye, but a mechanism rel ated to raised intrasheath
pressure is prob ab ly more likely. Examinati on of children
w ith intracrani al haemorrhage from a variety of causes other
than NAl showed that associated retinal h ae morrhage is
extremely rare.52
Forces Required to Cause Bleeding
The forces required to tear retinal, optic nerve and intra­
crani a l blood vessels have proved difficul t to quantify.
Biomechanical studies at the Children's hospita l of Phi lade 1­
phia were reported in 1987. 53 Using 'j ust bom doUs ', the
heads of which were fi lled with soaked cotton of a weight
th at simulated th e infa nt brain, the forces generated during
both vigo rou s shakin g and during impact with a solid sur­
face were measured. Mean angular accelerations achieved
during shakjng we re calculated at 9.29 G, but leapt to
428.18 G when impact was involved in the injury. On the
basis of this work, the authors concluded that severe
'shaken baby syndrome' was not usually caused by shaki ng
alon e, and was unlikely to be cau sed by shaking during
ordinary pl ay. When they th en went on to examine the
heads of 57 children ad mitted with su spected shaking
injury, they found evidence of impact, often subtle, in 75
per cent.
Othe r rep orts, however, have clearly documented fundus
haemorrhages in the absence of any impact inj uryS4
Fu rthe rmore, the forces asso ciated w ith both bungee
jumpi ng 55 an d aircraft ejection,56 which carry no impact
component, can alo g ive rise to retinal haemorrhages.
Despite t his divergence of views, it is agreed that the
forces required to gen erate retinal bleeding are cons ider­
able. These forces ca n almost never be ac hieved during
ordinary domestic ac cidents. Three recent case reports 57
have detailed bleeding following accidental trauma. In
one case, a 13-month-old boy developed subdural haemo r­
rhage and intra- and pre- retin al haemorrhage follow ing a
fall down 13 concrete steps. The fundus haemorrhages
were unilateral and clea red within 3 months. In the second
case, a 9-month-old boy fell 1-2 feet, striking his head on
the floor, and sustained bilate ral reti nal haemorrhage in
associ at ion with a subdural h aemo rrh age. In the third case,
a 7-month-old gi rl fell through a stair rail onto a concrete
basement flo or and again sustained both subdural and
(unilateral) retinal haemorrhage.
These repo rts serve to emphasize that no ocular feature
can be considered pathognom onic for NAl. None the less,
in la rge selies, accidental injury has almost a lways seemed
insuffic ient to cause retinal bleeding. For example, of 79
under-3-year-olds admitted to the Sick Children's Hospita l in
Toronto with head inj ury, none of the 75 who had experi­
enced accid ental trauma had fundus haemorrhages S8 Th e
three with NAl and the one whose injUlies were of un certain
cause all had fundus haemorrhages. Similar reports have
been provided from other instituti ons,59,50 indicating that,
excluding high-speed road traffi c accidents and falls from
extreme heights, accid ental injUly is most unlikely to provoke
retinal bleeding. The summary view ex pressed in a recent
Roya l College of Ophthalmologists Working Party ana lysis
was that 'No absolute values can be given for the a ngula r
acceleration forces required to produce retinal bleeding or
other injUly, but th ere is good evidence that they must be
considerabl e'.51
ASSOCIATED NEUROLOGICAL INJURY
The forces that give ris e t o fundus bl eed ing a lso will
commonly cause neurological damage. The correlation
betwee n retinal hae morrhages , cerebra l bleeding and
bo th morbidity and mortality is strong. The full impact of
an injury may not be immediately apparent, and it is
known that subdural haemorrhage may ap pear 2-3 days
afte r the retinal haemorrhages. 37 For this reason, detailed
exami natio n a nd documentation will be necessalY for sev­
eral days following the admission of a child with a suspi­
cious injury.
Attempts to correlate ocular and neurologica l fin din gs
have tended to involve pathologi cal exami nation of more
severely affected children. Subdural bl eed ing is t he most
common CNS finding in a child with traumatic retin al
hae morrhages. Of 13 fatally injured children examined in
Phil adelphi a , nine had suffe red blunt trauma w h ilst four
showed no evidence of blunt trau ma and were co nsidered
to be 'shaken bab ies '.62 All the children with subdural
h aemorrhage, re ga rdless of the type of injury, showed evi­
dence of retinal a nd optic nerve haemorrhage. A contro l
grou p of six chil dren dy in g from sudden infant death syn­
d rom e (SlDS) had no evidence of optic nerve bleeding.
In the sa me series, subarachnoid haemorrhage was less
commonly associ ated with retin al haemorrhages, and only

- f of the children with subarachn oid h2e morrhage
wed fundus bleeding as well. This study also showed
:::-.a- me babies who had been severely shaken were more
d_ - to show multilayered retinal haemorrhages than were
- _0 e who had suffered blunt trauma.
One child (subsequently considered to have suffered
- m SIDS) who underwent CPR had a few red cell s in
- e anrerior optic nerve - a findi ng in sharp contrast to
- e extensive bleeding associated with acceleration/
_C{'eleratio n injuries. Whatever the type of injury, cerebral
ema was found in all fatally injured children.
A similar study from Sheffield 20 examined the eyes and
brai ns of 23 fatally injured children; 16 of these children had
ead injuries, while the remainder died from suffocation,
abd ominal injUlies or other injuries. The children were all
Wlder 3 years of age. The authors constructed a 'total eye
ore' based on the presence of haemorrhagic retinal detach­
menr. retinal and optic nerve bleeding. They compared this
-',-ith an 'intracranial score' based on the extent of the
inn-acrani al bleeding, cerebral lacerations and evidence of
pri mary axonal injury. Though cerebral oedema was com­
mo nly reported from im aging, they found it difficult to iden­
. pathologically and, therefore, unlike the Philadelphia
dy, did not include it in the score.
The authors found a close relationship between the
-everity of the ocular and t he cerebral injuries, and felt able
to deduce a severity sequence for the damage suffered in
_-Al. They suggested that with increasing force of trauma
:0 the head a ch il d is likely to suffer:
• first, subdural haemorrhage, followed by subhyaloid,
intraretinal and then perineural optic nerve sheath
haemo rrh ages ;
• second, haemorrhagic retinal detachment (more
accurately subretin al bleeding);
• third, choro id al and vitreous haemorrhages, which
coincide with subarach noid haemorrhage, intracerebral
haemorrhage and cerebral laceration.
The authors concluded that the momentum of soft tissue
'."ithin a relatively rigid and stab le surround (e.g. the brain
:n the skull or the vitreous in the glob e) is responsible for
traction on vessels and the subseq uent bleeding. As with
other studies, they found that optic nerve bleedi ng tended
to be into the anterior portion of the nerve and did not
appear to extend forward from the site of intracranial
bleeding. They specu late that a sim ilar mechanism could be
:nvoked to explai n optic nerve bleeding, i.e. the mobile eye
generates axial and rotationa l forces aro und the relatively
fixed point that is the origin of the optic nerve from the
eye. As noted in other studies, milder eye changes may
initially seem to cause no associated CNS damage on imag­
ing. However, evidence mi ght appea r within a few days,
further emphasizing the need for careful review of affected
children.
Clinically based (rather than patholo gically) scoring
systems have also shown a significant correlation between
Fundus haemorrhages I 131
the severity of retinal bleeding and the associated neuro­
logical injuries. 50 However, t hough CNS and ocular haem­
orrhage commonly coexist, it is quite possible for either to
occur independently. A 1991 study 63 fou nd that 39 per cent
of subdural haemorrhages had no associated ocular haem­
orrhage, and that seven per cent of children with retin al
bleeding had no evidence of CNS bleeding.
Differential Diagnosis of Fundus Haemorrhages
Since no ocular pathology is unique to NAl, it is essential that
the clinician excludes other possible causes of the observed
clinical signs. This needs to be done as quickly as possible so
that appropriate steps can be taken, whatever the outcome of
investigations. A comprehensive list of diagnostic differentials
has been produced by a Working Party ofthe Royal College of
Ophthalmologists G1 ,64 (Table 6.1). Some of these will be obvious
following initial assessment whilst others demand particular
diligence on the part of the supervising clinici an.
Alternative explana tions that need to be consid ered
includ e accidental inju ry, CPR, epileptic seizures, arousal
shak ing, bleeding disorders, bleeding secondalY to infection,
genetic metabolic disorders and birth-associated retina l
haemorrhage.
ACCIDENTAL INJURY
A hi stO lY of accidenta l injury or injury associated with
attempts to aro use or resuscitate is commonly cited to
account for neurological and ocular damage. As discussed
above, accid ental injury, except the most severe, is almos t
never the cause of significant cerebral or ocular bleeding.
Table 6.1 Differential diagnosis of retinal haemorrhages in
children 61 •64
Leukaemia (particularly ac ute lymp hatic)
Haemorrha gic disease of the newbo rn
Retinopathy of prematurity
Sickle cell retinopathy
Extraco rporeal membrane oxyge nation
Metabolic (galactosaemia, glutaricacid uria)
Henoch-Schonle in pu rpura
Maternal cocaine ingestion
Meningitis
Intracranial vascular malformation
Optic nerve dru sen, tuberous sc lerosis, X-linked retinoschisis
Chronic severe papilloedema
Intraocular surgery
Severe hypertension
Protein C deficiency
Von Willebrand's disease
Cerebral malaria
Vaginal delivery (sponta neous and assisted)
1 32 l Ocular involvement in non-accidental injury
CARDIO PULM ON ARY RESUSCITATIO N
As in the case of accidental injury, isolated case reports
exist to suggest that vigorous CPR, particularly if per­
formed by inexperienced practitioners on very young
children, may be responsible for retinal haemorrhage. 65
The infant in Kramer's case report was fully documented,
with fundus examination prior to the resuscitation
attempt showing there had been no bleeding. This, there­
fore, represents one of the few cases in which the bleeding
unequivocally followed the CPR rather than being present
prior to instigating CPR. This particular child received
60 minutes of vigorous CPR, after which she was found to
have intraretinal and pre-retinal haemorrhage out to the
mid-periphery of the retina.
Set against such individual reports, prospective series
of paediatric patients and experimental studies on piglets
have shown retinal haemorrhage secondary to CPR to be
extremely unlikely.
In a large clinical series, Kanter found only a single small
haemorrhage in 40 children requiring CPR. 66 He concluded
'When retinal haemorrhage is detected in the paediatric
patient after CPR, prior trauma should be assumed'. More
recently, 169 post-mortem ocular examinations were
performed in children after failed CPR. 67 Sixty-one of
the children had retinal haemorrhage after CPR, of whom
56 had suffered head injuries and four had CNS disease or
sepsis; in only one case was cause of death undetermined.
No case was found with retinal haemorrhage in which CPR
was the only explanation for the bleeding. In a prospective
study of 43 children admitted to hospital with non­
traumatic problems and requiring resuscitation, small punc­
tuate retinal haemorrhages were seen in a single case. 68
Six newborn piglets receiving CPR for up to 50 minutes
with monitored intrathoracic and intracranial venous pres­
sure showed no evidence of retinal bleeding.69
EPI LE PTIC SEIZURES
In our hospital, we examined a series of children admitted
with convulsions. 70 The aetiology of the seizure disorder
varied but none was traumatic. Even though 40 of the chil­
dren were less the 2 years of age and, therefore, in the most
vulnerable age group, none showed any evidence of fundus
bleeding. The children examined were all seen within 24
hours of admission, and examined in detail with an indir­
ect ophthalmoscope using pupil dilatation in most cases.
In a more recent prospective study of 143 children aged
between 2 months and 2 years who were seen in an acci­
dent and emergency department following seizures and
examined by an ophthalmologist, unilateral retinal haem­
orrhages was documented in only one child. 71
As with resuscitation attempts, despite individual case
reports72 it should be assumed that, if retinal haemorrhage
is found in a child admitted with fits, the likelihood is that
injury caused both the fits and the fundus bleeding.
AROUSA L SHAK ING
Arousal shaking is always a difficult area, with many practi­
tioners considering it inconceivable that the forces required
to cause intracranial and ocular bleeding could be generated
in a reasonable attempt to arouse an unconscious child.
Nevertheless, it is a defence used in court with sufficient fre­
quency for the British Medical Journal to have published an
article warning carers of the possible dangers of baby shak­
ing and an admonishment to exercise caution. 73
BLEEDING DISOR DE RS
Various types of bleeding disorder have been shown to give
rise to fundus haemorrhages. Lymphoblastic leukaemia, in
particular, is recognized to cause widespread retinal bleed­
ing, involving all retinal layers. This occurs particularly
when the children are thrombocytopenic. 74 Sickle cell dis­
ease, aplastic anaemias, Henoch-Sch6nlein purpura and
vitamin deficiencies are all capable of producing retinal
haemorrhage and must be excluded before pursuing a
diagnosis of NAl.
BLEEDING SECONDARY TO INFECTION
Meningococcal meningitis was reported in 1995 as a cause
of fundus bleeding?5 In our hospital, we have seen a child
with massive, unilateral retinal haemorrhage (associated
with fatal meningococcal disease) in whom the provisional
diagnosis was NAI and the true diagnosis was revealed
only by post-mortem studies.
GENETIC METABOLIC DISOR DERS
Although individually rare, a number of genetic metabolic
disorders (GMDs) have been shown to cause retinal bleed­
ing and must be actively excluded in the evaluation of NAl
victims. Galactosaemia may cause widespread bleeding
into both retina and vitreous humour, particularly when
associated with significant liver dysfunction?6 Less com­
mon GMDs have attracted attention in recent years, since it
became clear that glutaricaciduria could give rise to both
retinal and subdural bleeding along with cerebral atrophy.
Children presenting with both subdural and intraretinal
bleeding associated with glutaricaciduria were reported as
long ago as 1987. 77 These reports further emphasize the
need to exclude metabolic causes in all suspicious cases.
Autosomally recessive protein C deficiency can also be
responsible for intracranial (usually subarachnoid) and
intraocular bleeding. Vitreous bleeding with severely
reduced protein C levels has been reported at birth in a
child of 37 weeks' gestation. 78
Other entities, such as hypertension, prematurity and
AlDS-related cytomegalovirus (CMY) infection, should be
apparent from the history and examination of the child on
admission to hospital.

The most common type of retinal haemorrhage in
infancy is birth associated. It merits separate consideration.
BIRTH -ASSOCIATED RETINAL HAEMORRHAGE
Retinal haemorrhages are found in the first few days after
delivery in up to 59 per cent of children. 4B The haemorrhages
clear quickly and, of 1238 children reported in 1970,49 18.9
per cent examined in the first 24 hours showed signs of
having had a retinal haemorrhage. When examination was
delayed until 3-5 days after birth, only 2.6 per cent showed
evidence of fundus bleeding, presumably indicating that
most of the haemorrhages clear in the first few days of life.
Recent studies found retinal haemorrhages in about one­
iliird of babies following spontaneous vaginal delivery and
in 75 per cent or more following vacuum extraction.79.BG
The incidence of ocular bleeding following Caesarean
de livery is lower, at around 7 per cent/ 9 ,BO suggesting that
the raised intrathoracic and intracrania l pressures associ ­
ated with compression of the baby in the birth canal are
not the only aetiologica l factors . It has been specu lated Bl
chat hypoxia and hypercapnia are also contributory.
The vast majority of neonatal retinal haemorrhages are
superficial and clear within 5 days?4 a less frequent but more
1V0rrying injury is a pre-retinal haemorrhage, which clears
more slowly. Nevertheless, even pre-retinal haemorrhages
associated with delivery should disappear within 6 weeks;
fundus bleeding after this stage needs to be fully investigated .
Examining the Child With Fundus
Haemorrhages
- - - - -. - - ­
Though a paediatrician may be the first doctor to suspect
~etina l haemorrhages in a child, it is essential tha t an oph­
j]almologist is involved at an early stage (Fig. 6.8). An eye
assessment shou ld be requested on any child suspected of
being a victim of NAI, particularly if the head (including
CNS) and neck area are involved. The responsibility of the
ophthalmologist is to examine the whole of the retina
(where possible), and accurately document the distribution
and depth of the retinal haemorrhages along with associ­
a ted optic nerve and retinal features. The tool of greatest
':alue is an indirect ophtha lmoscope, w hich allows a view
of the retina right out to its peripheral extent. Such a view
can be achieved only through dilated pupils. Thus, if neuro­
:ogical monitoring precludes pupillary dilatation at first
co ntact, the supervising team must be prepared to notify
rbe ophthalmologist when pupil dilatation is permissible.
In some instances a short-acting mydriatic, such as tropi­
(a mide, may be appropriate. However, in other instances,
!Ja n icularly in a child with dark-brown irides, a stronger
and longer-acting agent, such as cyclopentolate or
~h enylepherine, is likely to be necessary. There is no reason
i...n these circumstances to use agents that cause an effect
~-or several days, such as atropine or homatropine.
Fundus haemorrhages I 133
See at earliest opportunity after referral
Document all fin dings (including negati ve fin dings) in notes. In clude time
and date of examination. Sign examination record
Do not dilate pupils without discussion with supervising
pa edia tri cia n/i nte nsivis t
If dilatation not possible, record this limitation in notes
Notify consultant of findings - immediately if positive
Consultant review as soon as possible if positive findings
Photography of fundus pathology if dilatation possible
Print off copies of fundus photographs
Place in sealed envelope with note - signed by yourself and another
witness - indicating the time and date of the photographs
Place sealed envelope in patient's main hospital notes
A second copy - similarly validated - retained in eye department
Specify date of review and ensure review undertaken
Figure 6.8 A guide to the examination of children with
suspected non-accidental injury.
Once the appearances have been documented in detail,
the child shou ld be reviewed on a regu lar basis, both to
observe the changing appearance of the haemorrhages and
to determine whether any therapeutic intervention is likely
to be needed. It is not possible to date an injury on the
basis of retinal haemorrhages (in the way that skin bruising
can be used).61 If, however, the appeara nces change with
time, knowing the ti mescale of the absorption of the
observed haemorrhages may give a guide as to the timing
of their original appearance. Dating the time of haemor­
rhage with post-mortem specimens is more practical and
uses the appearance of haemosiderin. The presence of
haemosiderin in the eyes of abused children has been con­
sidered to indicate that an injury is at least 3 days 01d. 82
To further aid the documentation of the retinal haemor­
rhages, fundus photography may be useful. Traditional
fundus photography is unlikely to be possible, but photog­
raphy with a variety of hand-held cameras certainly will
134 lOcular involvement in non-accidental injury
be. The limitation on such documentation is related to the
relatively small angle of view of most cameras and can be
overcome using a RetCam device (Massie Research Labora­
tories, Inc., Dublin CAl. This recently developed digital
camera allows a uniquely wide, 1200 angle of view and is
combined with a telemedicine facility. 83 It is still being
evaluated, but tools of this sort are sure to remove some of
the uncertainty surrounding clinical descriptive recording.
When the view of the fundus is compromised by vitreous
haemorrhage, imaging with ultrasonography can be coupled
with electroretinography (ERG) to define the location of the
retina and its functional integrity. Fishman and coJJeagues 84
found ERG to be unhelpful in the initial assessment of six
children with retinal haemorrhages. Their ERG parameters
were not significantly different from those in six age­
matched control subjects. However, with time, one parameter
improved, suggesting that some subtle neurophysiological
damage was caused by the injury. They also stressed the
importance of ERG in distinguishing between neurological
and retinal visual loss at an early stage. In the presence of
vitreous haemorrhage, ultrasonographic evidence of retinal
detachment and ERG evidence of impaired retinal function
should provoke early referral to a vitreo-retinal surgeon.
Clearing the vitreous blood and surgical relocation of a
detached retina will help long-term visual prognosis. Some
authorities have also succeeded in clearing pre-retinal haem­
orrhage using intravitreal injections.85 All such procedures
need to be considered when the child 's long-term visual
development is likely to be compromised .
LONG-TERM VISUAL SEQUELAE
The majority of victims of NAI who escape major n euro­
logical injury will have a fuJI visual recovery. A study of 30
child victims of shaking showed that non-reactive pupils at
presentation coupled with midline shift on imaging correl­
ated strongly with a fatal outcome. 85 Of the 22 children
who sUlvived, 18 had good vision in at least one eye, with
two children having their poor vision on the basis of cor­
tical visual impairment.
A much larger study of ophthalmic morbidity following
head injury from any source and at any age 27 revealed that
48 per cent had some long-term morbidity. However, the
majority of these patients had ocular motility disturbances,
and only a minority had vitreo-retinal, optic nerve or cor­
tical damage sufficient to cause visual loss. A study of 99
children from Nigeria 87 revealed similar results, with 28 per
cent having visual loss in the long term, but most com­
monly only one eye was involved . Significant visual
impairment is difficult to divine from this series, but
may have been limited to one child with cortica l visual
impairment.
It is clear from these studies that, despite the potential
for visual loss secondary to macular folds, retinal detach­
ment, persistent vitreous haemorrhage and other structura l
injuries to the eye, such visual loss is in fact relatively
uncommon, except in the child who has also sustained
serious neurological injury.
ADDITIONAL OBSERVATIONS
It is a surprising observation that, despite the nature of the
injury, strictly unilateral retinal haemorrhages may result
from severe shaking. In 1997 we reported three consecutive
cases in which the bleeding was unilateral,25 and high­
lighted the fact that such a presentation should not cast
doubt on the diagnosis of NAl. Two of those three children
suffered severe long-term visual loss, despite the unilateral
nature of their ocular injury.
A review of the literature revealed strictly unilateral
bleeding referred to in other series, and in those cases, too
the associated neurological injury was often severe. 52 .G3
The presence of unilateral eye signs does not diminish the
likelihood that the injuries are a consequence of abuse; nor
do they predict a more minor level of eye and neurological
damage.
Non-accidental injury is a harrowing topic for all clin­
icians. We owe it to the children in our care to consider it as
a diagnosis to avoid the child being exposed to further
injury. However, we also owe it to the family to ensure that
the vista of NAI is only raised in appropriate circumstances.
Detailed examination of the eye and modern documen­
tation wiJJ contribute to the diagnosis, but none of the ocu­
lar features can, in isolation, establish a diagnosis of NAI .
Despite recent reports, domestic accidents rarely, if ever,
cause the level of ocular damage that is commonly associ­
ated with severe shaking or impact injuries.
REFERENCES
1 Voutilain en R, Tuppurain en K. Ocular Munchausen syndrome
induced by incest. Acta Ophthalmol (Copen) 19 89;
67:3 J9-21.
2 RadbiJi SX. A his tory of child abuse and infanticid e. In Helfer
RE, Kempe CH (eds) The Battered Child. Chicago, IL:
University of Chicago Press, 1968.
3 Kempe CH , Silverman WA, Steele BF et al. Th e battered child
syndrom e. JAMA 1962; 181:17-24.
4 Cam eron JI\1 , Johnson HR, Camps FE. The battered child
syndrome. Med Sci Law 1966 ; 6:2-7.
5 Johnson CF, Showers 1. Injury variabl es in child abuse. Child
Ab1lse Negl 1985; 9:207-11.
6 Hendeles S, Barber K, Willshaw HE. The ri sk of ocula r
involvement in non- accidental il\iuly. Child Care Health De))
1985; 11 :345-8.
7 Riffenburgh RS, Sathyavagiswaran L. The eyes of child abuse
victims: autopsy findings. J Forensic Sci 1991 ; 36:741-7.
8 Rao N, Smith RE , Choi JH et al. Autopsy findings in fourt ee n
fatally abused children. Forensic Sci 1I1t 1988; 39:293 - 9.
9 Hadley MN, Sonntag VKH, Rekate HZ, Murphy A. The infant
whiplash-sh ake injury syndrome: a clinical and pathological
study. NeurosLiTgelY 1989; 24:536-40.
10 Duhaime A- C, Gennarelli TA, Thibault LE et al. The sh aken
baby sy ndrome. A clinica l, pathological and biomechanical
study. J Neurosurg 1987; 66:409-15.

11 Jacobi G. Damage patterns in severe child ab use with and
without fatal seq uelae. Monatssch r Kinderl1eilkd 1986;
13 4: 307-15.
12 Harcourt B, Hopkins D. Ophthalmic man ifestations of the
batte red-baby sy ndrome. BM} 1971; 3 :390 .
13 Feldman KW, Schaller RT, Feldma n JA. Tap water scald burns
in children. Pediatrics 1978 ; 62: 1-9.
14 Landers A, Belfer KF, Jenki ns AD. Supergill e Jids: possibly non­
acc idental and a medicol ega l problem. Eye 1996; 10:402 -3.
15 Tom inack RL, Spy ker DA. Capsicum and caps aicin - a review:
case rep ort of the lise o f hot peppers in child abuse. } Toxicol
Ciin Toxicol1987; 25:591-60 1.
16 Tse ng SS, Keys MP. Battered child syndrome simul ating
congenital glaucoma. Arch Ophthalmol 1976; 94:8 39-40.
17 Fri en dl y DS. Ocu lar manifestations of phys ical child ahuse.
Tran s Am Acad Ophthalmol OtolarYllgol 19 71 ; 7 5:318 - 3 1.
18 Weidenthal DT, Levin DB. Retinal detachment in a battered
infant. Am J Ophthalmol 1976; 81:725-7.
19 Gonzales CA . Bilateral rhegmatogenous detachments with
unilateral vitreo us base avu lsion as the presenting sig ns of
child abuse. Am} Op hthalmo/ 1999 ; 127:475-7.
20 Green MA, Lieberman G, Milroy CM, Parso ns MA. Ocu lar and
ce rebral trauma in non- accidental injury in infan cy :
underlying mecha nisms and implications for paediatric
practice. Br} Oph tl1OImoi 1996 ; 80:282-7.
21 Massic ote SJ, Folberg R, Torc zynski E et al. Vitreoretinal
traction and peri macular retinal fold s in the eyes of
deliberately tra umatized children. Ophthalmology 199 1;
98:1124-7.
22 Munger CE, Peiffer RL, Bouldin TW et al. Ocular and
associated neuropathological observa tio ns in suspected
whiplash shaken infan t syndrome. A retrospective stu dy of
12 cases. Am} Forensic Med Pathol 1993; 14:193-200.
23 Giangiacomo J, Barkett KJ. Ophth alm oscopic findin gs in
occ ult child ab use.} Pediatr Ophth almol Stra bismus 1985;
22:234-7 .
24 Levin AV. Oc ular mani festa tions of ch ild abus e. Ophthalmol
Clin No rth Am 1990 ; 3 :249 -64.
25 Tyagi AK, Willsha w HE, Ai nswo rth JR. Unilateral retinal
haemorrh ages in non- acc idental injury. Lancet 1997; 349:1 224.
26 Lin KC, Glasgow BJ. Bilateral peri op tico intrascleral
hemorrhages associated with traumatic child abuse.
Am} Oph tlw lmol 1999; 12 7 :473- 5.
27 Ktitzinger EE. Pers ona l com mun ication.
28 Kowal L. Ophthalmic manifestati ons of head injury. Aust NZ
} Oph thalmol 199 2; 20:35-40.
29 Frank y, Zimmerman R, Leeds NM . Neurologic manifestation
in abused children who have bee n shaken. Dev IVIed Child
Neurol 1985 ; 27:31 2-1 6.
30 Jensen AD, Smith RE, Olson Ml. Ocular clues to child abuse.
} Pediao' Ophthalmol Strabismus 1971; 8:270- 3.
31 Ball anty ne AJ. Th e nerve fibre pattern of th e huma n retin a.
Trail S Ophtha lm ol Soc UK 1946; 66: 179-91.
32 Roth M. Uber Netzhau ta ffection en bein Wund feibe rn. Deutsch
Z ChiI' 1872 ; 1:471-00.
33 Falcone PM, Larrison WI. Roth spots see n on
ophthalmoscopy: diseases with whi ch they may be associated.
Conn Med 1995; 59:271-3.
34 Kaur B, Taylor D. Fundus haem orrhages in in fancy. SurIJey
Ophthalmo/ 1992 ; 37 :1- 17.
35 Fred HL, Chandler FW. Traumatic asphyxia. Am} IVIed 1960;
29:508-17.
36 Holl en horst RW, Stein HA. Ocular signs and prognosis in
subd ural and sub arach noid bl eeding in young children.
Arch Ophthalmol1958; 60:187-92.
37 Mushin AS. Ocular damage in the battered baby sy ndro me.
BM} 1971; 3:40 2.
References I 135
38 Va ega n, Taylor D. Critical peri od for deprivation amblyopi a in
ch ild ren. Trall s Ophthalmology Soc UK 1979; 99 :432-9.
39 Mathews GP, Das A. Dense vitreo us hem orrh ages predict poor
visual and neurologica l prognosis in infants with shaken baby
syndrome.} Pedia tr Ophthalmol Strabismus 1996; 33:2 60 - 5.
40 von Noorden GK. New clini ca l aspe cts of stimulus deprivat ion
a mblyo pia. Am} Opllthalmol1981; 92:416-21.
41 Ober RR. Hem orrhagic retin opat hy in infancy: A clinico­
patho log ical report.} Pedia tr Ophthalmol Strab ismus 1980;
17:17-20.
42 Lambert SR , Johnson TE, Hoyt CS. Optic nerve sheath
hemorrh ages assoc iated with shaken baby sy ndrome. Arch
Ophthalmol 1986; 104: 1509.
43 Mun ger CE, Peiffer RL, Bouldin TW et al. Ocular and
associated neuropat hologic observ ations in suspected
w hipl as h baby sy ndrome: a retros pective study of 12 cases.
Am} Forellsic IVIed Pathol 1993; 14: 193- 200.
44 Mass icote SJ, Folberg R, Torczynski E et al. Vitreoretinal
traction and perimacular retin al folds in the eye of deliberately
traumatized children. Oph thilimology 1991 ; 98: 1124-7.
45 Greenwald MJ, Weiss A, Oesterle (S, Friendly DS. Traumatic
retinoschisis in battered babies. Ophthalmology 198 6;
93:618-25.
46 Riffe nbllrgh RS, Laks hmanan S. Ocular findings at autopsy of
child abuse victi ms. Ophthalmology 199/ ; 98: 1519-24.
47 Tomasi LG , Rosman P. Purtscher retinopat hy in the battered
child syndrome. A m) Dis Child 1986; 93:1435-9.
48 Ba um JD, Bulpi tt CJ. Retina l and conjun ct ival haem orrhage in
the newborn. Arch Dis Child 1970; 45:344-9.
49 Sezen F. Retinal haemorrhages in newborn infants.
Br} Opllthalmol 1970; 55: 248-53 .
50 Wilkinson WS, Han DP, Rappley MD, Owings CL. Retin al
hemorrhage pred icts neurolog ic injuty in the shaken baby
sy ndrome. Arch Ophtlwll1lol 1989; 107:147 2-4.
51 Paton L. V 11 diseases of the nervo us system: ocular sy mptoms
in subarachn oid haemorrhage. Trans Ophthalmol Soc UK
192 4; 44: 110-26.
52 Schlo ff S, Mullaney PB, Armstrong DC et al. Retinal findings
in children with intra crania l haemorrhage. OphtllGlmol 2002;
109: 1472-6.
53 Duhaime A-C, Gennarelli TA, Thibaul t LE et al. Th e shake n
ba by syndrome - a clini cal, pathological a nd biomecha nical
study.} Neurosu rg 1987 ; 66:409-1 5.
54 Gilliland MG, Fo lberg R. Shaken babies: so me ha ve no impact
injuries.} Fore nsic Sci 1996; 41:/14-1 6.
55 David DB, Mears T, Quinlan MP. Ocul ar co mplications of
bllngeejumping. Br} Op/lthalmol 1994; 70:234-5.
56 Lyle DJ, Sta pp JP, Button RR. Ophthalmologic hydrostatic
press ure syndrom e. Am J Opllthalmo/1 95 7; 44: 652 -7.
57 Christian CW, Taylor AA, Hertle RW, Duhaime A- C. Retinal
haemorrhages caused by acc idental household traum a.
} Pediatrics 1999; 135:1 25 -7.
58 Buys YM, Levin AV, Enzenauer RW et al. Retinal findin gs
after head tra uma in in fants and young children.
Ophthalmology 1992; 99: 1718-23.
59 Billmire ME, Myers PA. Serio us head injury in infants:
accident or ab use. Pediatrics 1985; 75 :340- 2.
60 Elder JE, Taylor RG, Klug GL. Retina l haemorrhage in
accidental hea d traum a in childhood. } Pediatr Child ffealth
1991 ; 27:286.
6 1 The Ophth almology Chi ld Abuse Wo rkin g Patty. Child abuse
and th e eye. Eye 1999; 13 :3-10.
62 Budenz DL, Fa rber MG, Mirchandani HG et al. Ocular and
optic nerve hemorrhages in abused in fa nts with intracranial
injuries. Ophthalm ology 1994; 101 :559-64.
63 Riffenburgh RS, Lakshm anan S. Ocular findin gs at autopsy of
ch ild abuse victims. Ophthalmology 1991; 98:15/9-24.
136 l Ocular involvement in non-accidental injury
64 Adams G, Ainsworth J, Butler Let al. Update from the child
abuse working party: Royal Coll ege of Ophthalmologists.
Eye 2004; 18:795-8.
65 Kramer K. Retinal hemorrhages following cardiopulmonary
resuscitation. (lill Perliatr 1993 ; 13 :366-8.
66 Kanter RK. Retinal hemorrhages after cardiopulmonary
resu scitation or child abuse. ) Pediatr 1986 ; 108:430-2.
67 Gilliland MG, Luckenbach MW. Are retinal hemorrhages
found after resusc itation attempts? A study of the eyes of J 69
children. Am) Forensic Med Pathol 1993; 14:187-92.
68 Odom A, Christ E, Kerr N et al. Prevalence of retinal
haemorrhages in pediatric patients after in-hospital
cardiopulmonary resusc itation: a prospective study.
Pediatrics 1997; 99:E3.
69 Fackler JC, Berkowitz !D, Green R. Retinal hemorrhages in
newborn piglets foll ow ing cardiopulmonalY resusc itation.
Am) Dis Child 199 2; 146:1294-6.
70 Tyagi AK, Scotcher S, Kozeis N, Willshaw HE. Can
convulsions alone cause retinal haemorrh ages in infants?
Br) Ophtha/mol 1998; 82:659-60.
71 Mei-Zahav M, Uziel y, Ra z J et al. Convulsions and retinal
haemorrhage: s hould we look fUJther. BM) 2002; 86:334-5.
72 Feyi-Waboso AC, Bec k L. Minerva. BM) J997; 314:688.
73 Markovitch H. Opening eyes to child abuse. BM) 1999 ;
318:950.
74 Guyer DR, Sch achat AP, Vitale S et a1. Relationship between
fundus lesion s and haematological parameters at diagnosis.
Ophthalmology 1989; 96:860-4.
75 Frase r SG. Retinal haemorrhage in meningitis. Eye 1995;
9:6 59 -60.
76 Levy HL, Brown AE, Williams SE, de Juan E. Vitreous
hemorrhage as an ophthalmic complication of galactosemia.
) Perliatr 1996 ; 129:922 -5.
77 Hoffman GF, Nau ghten ER. Abuse or metabolic disorder. Arch
Dis Child 1998; 78:399.
78 Pulido JS, Lingua RW, Cristol 5, Byrne SF. Protein C
deficiency associated with vitreous hemorrh age in a neonate.
Am) Ophthalmol J987; 104:546-7.
79 Emerson MY, Pieramici OJ, Stoessi K]\il et al. Incid ence and
rate of disa ppearance of retinal haemorrhages in newborns.
Ophthalmology 200J; 108:36-9.
80 Hughes LA, May K, Talbot JF, Parsons MA. Incidence,
distribution and duration of birth-related retinal
haemorrh ages : a prospective study. ) AAPOS 2006 ; 10: 102-6.
8l Von Barsewisch B. Perilla tal Retinal Haemorrhages:
iVlorplw/ogy, Aetiology and Significance. Berlin: Springer­
Verlag, 1979.
82 Gilliland MG, Luckenbach MW, Massicote SJ, Folberg R. The
medicolegal implications of detecting hemosiderin in the eyes
of children who are suspected of being abused. Arch
Opi1thaimoI1991; 109:321- 2.
83 Schwartz S. Telemedicine and ophthalmology Jules Stei n Eye
IJlst: Clin Update 1998; 7:3-4.
84 Fishman CD, Dasher WB, La mbert SR. Electroretinographic
findings in infants with shaken baby syn drome. ) Pediatr
Ophthalmol Strabismus 1998; 35:22-6.
85 Conway MO , Peyman GA , Recasens M. Intravitrea l tPA and
SF6 promote clearing of premacular subhyaloid hemorrhages
in s haken and battered baby syndrome. Ophthalmic Surg
Lasers 1999; 30:435-41.
86 McCabe CF, Donahue SP. Prognostic indicators for vision and
mortality in sh aken baby syndrome. Arch Ophthalmol 2000;
118 :3 73-7.
87 Shokunbi T, Agbeja A. Ocular co mplications of head injuly in
children. Childs NeTlJ System 1991; 7:147-9.

THE DEATH SCENE FOLLOWING THE
SUDDEN DEATH OF A CHILD
Anthony Busuttil
Introduction 137
Scene management 137
The crime scene manager 138
Sequence of events at the death scene 139
Unclothing the body 139
A good look around 139
Sudden infant death syndrome or non-sudden infant
death syndrome 140
External petechiae 140
INTRODUCTION
they feel have led to the injuries an d the demise of the child.
The proper, co mprehensive, planned and professional
investigation of a suspicious death of a child requires a
carefully managed, orde rly and thorough inspection of the
scene where the death has occurred, wit h its full documen­
ta tion. J This should take place whether the bo dy of the
child is still in situ or whether it has already been removed.
Th e scene investigation forms as much an essential part of
~he death investigation as does the autopsy and an access
~ o all the medical records of the deceden t.
The patho logist investigating the death should always be
given app ropriate access to the scene, as this will assist fur­
iher in the reconstruction of the events that led to the death
of the ch ild. 2 ,3 Access to the scene may be required again
a fter the au topsy has been completed to attempt to identify
-le vario Lls sites at which blunt trauma may have been SLlS­
(ained. If a cou nter- argument of an accidental infliction of
Jl e injuries is being pursued, for example rolling over from
a table or couch or a fall down stairs, then it is essential that
" is hypothesis be tested at the scene in the exact location
-,';here it is suggested to have occurred. This reconstruction
should a lso involve the forensic pathologist.
On occasions, carers of a baby or child who has died
- nexpectedly give a complex narrative of the events that
I CHAPTER 7 I
Bruising 140
Abandoned neonates 140
Deaths from trauma 141
Dyadic and multiple deaths 141
Sudden deaths of older children 142
Sensitivity and stress of the investigation 142
Inquests and inquiries 143
References 143
The police have often found it useful to encourage these wit­
nesses to re-enact quite carefull y and meticulously at the actual
scene what t hey are suggesting has taken place, using dolls
as substitutes for the child or infant. These re-enactments,
which ha ve been resorted to more frequen tly given the con­
troversies that may arise about the manner of whi ch head
and neck injuries have been sustained, are videoed by the
police. Forensic pathologists should view these and comment
thoughtfully on the plausibility or otherw ise of what is being
suggested. It has also been known for legal defence personnel
to do this; if it is known that this tactic will be employed, it
is essential that the pathologist sees such reconstructions in
good time before the case comes to court.
SCENE MANAGEMENT
It has to be decided at a very early phase, after tile deatll has
been brought to the attention of the police, whetller it is
thought that the death under investigation is suspicious or not.
It is always the counsel of perfection to be safe rather than
sorry, and to upgrade - at least initially - the investigation of
a scene to suspicious status when one is unsure. This decision
may require an early input from a forenSically trained medical
138 I The death scene following the sudden death of a child
practitioner, though not necessarily a pathologist, and the
viewing and careful inspection of the body. In situations that
speak for themselves, in particular when evidence of violence
is present, this, of course, can be dispensed with.
The fact that the body of the child may already have been
taken to hospital for attempted resuscitation does not dispense
with the scene examination, and a close retrospective inspec­
tion of the scene where the death has occurred is called for at
the earliest possible opportunity. Once the attention of the
police has been drawn to the occurrence of a suspicious, vio­
lent or unexpected death, every effoli should be made to return
to the scene where the death took place. In these circumstances,
it is essential that hospital staff and emergency personnel are
interviewed soon after the event to obtain from them informa­
tion regarding: the findings on the child when brought into
hospital , dead or moribund; the results of any emergency bio­
chemical , haematological and radiological investigations car­
ried out; and the various aspects of the resuscitative process
that have taken place, including drugs administered to the
child. It is also useful to obtain access at an early stage to - and
indeed to secure seizure of, against appropliate receipting ­
any samples collected from the now-deceased child prior to
any medication or intravenous infusions having been admin­
istered. These so-called 'pre-transfusion specimens' may prove
extremely useful for further analysis.
Similarly, if clinical photographs have been taken of the
injuries of the child then police should obtain access to
these. Any clothing removed from the now-deceased child,
including nappies, should be handed over to the police.
Slightly more controversial decisions may have to be
taken, as to when and whether information should go to the
police, in cases in which a child is admitted with severe
injuries and is likely to die as a consequence. If injuries sus­
tained are serious and, perhaps, 'are likely to prove' fatal, a
decision to inform the police or not has to be taken at the
earliest possible time by the clinical staff. There should be
no dubiety in the minds of the doctors in accepting that this
would not constitute a breach of medical confidentiality
due to the patient giving their oveniding obligations as cit­
izens of a country to ensure that crime is investigated and
its perpetrators detected. A senior member of the medical
team should make this disclosure to the police, speaking to
a senior police officer. Hospitals should produce written
protocols about the procedures that should be followed in
such circumstances. The child's carers must be told that the
police have been informed.
THE CRIIVIE SCENE MANAGER
The scene will often be placed under the control of a spe­
cially trained scientist or police officer designated as the
crime scene manager (CSM). The CSM's g'uiding principle is
that enunciated by Edmond Locard over a hundred years
ago - namely that every contact between two persons or a
person and a scene will leave a trace of this contact, which
if looked for and found, thereby proves that such a co ntact
had indeed occurred. 4
The agenda set before the CSM comprises the following:
1. From the time the police first attend the crime scene,
an accurate and detailed record must be maintained of
the comings and goings and of who did what and
when. At an early phase, thought should be given to a
retrospective timetabled account of events occurring
prior to the arrival of the police.
2. That the scene itself be documented fully, both
photographically and by video recording, and, if
required, by sketches to sca le also. It is an essential good
practice for individuals attending the scene to make their
own notes and to make their own rough sketches as
aide-memoirs. Such documentation will enable 'best
evidence' to be produced in front of the courts. All this
documentary material will form part of documentary
disclosure when a criminal procedure is in force. The
photographs and videos of the scene will also be
available later to refresh the minds of those working on
the case, which allows better briefings of further
additional personnel joining the investigative team.
J. All trace evidence has to be collected painstakingly from
the scene. A systematic, planned and thorough search
has to be made of the entire scene and, initially,
everything that may have some relevance has to be
co llected. To be exclusive and selective in evidential
collection at an early phase of an enquiry may be shown
later to have been counterproductive and inefficient. The
approach to the collection of evidence should be that
each item collected is logged and a 'continuity of the
chain of evidence' is established for every item. This
refers to the progression of each item from the scene to
the forensic laboratory and eventually to couli; each step
of this procedure has to be accounted for fully and not
found to be wanting. Every person handling the item of
evidence signs the original label attached to it at the
scene. Strict attention has to be given to ensure that
appropriate containers, methods of preservation,
labelling and packaging are used. The collection,
labelling and logging of the items collected is delegated
to a police officer designated as the exhibits' (or
productions', in Scotland) officer.
4. All relevant investigative personnel should be given
access, in turn, to the scene, ensuring that they are all
wearing appropriate protective clothing in the course
of their attendance to decrease scene contaminati on .
They will be briefed to work in unison and in mutual
collaboration; altho ugh, in order to avoid congestion
and overcrowding at the scene, each team may have to
work separately, with the crime scene manager present
at all times. The pathologist's role at the scene should
principally concern the body. However, even if the
body has left the scene, there is often still scope for the
pathologist to inspect the scene of death, particularly if
it is being treated as a suspicious death.
 A good look around I 139

5. It is essential that there is adequate lighting at the 9. Specially trained 'scenes of crime' officers inspect the
scene. It is foolhardy to attempt to carry out decent scene and collect trace evidence.
and proper investigations without appropriate lighting. 10. Forensic scientists inspect the scene and collect
The crime scene manager should ensure that portable evidence.
generators or other sources of light are made available. 11. Pathologist is given access to the body.
The commencement of the investigation may have to 12. Trace evidence from the body, in terms of tapings
await the arrival of such equipment and if possible from the exposed surfaces of the body (for fibre
daylight. evidence), is collected, and swabs are taken from
6. Information is sought from the scene that will assist accessible orifices.
with an estimation of the time of death; this requires a 13. The body is transferred on to a clean cotton or
recording of the ambient temperature. If death occurred polythene sheet and tra nsferred to the mortuary for
indoors and sequential changes to room temperature more detailed inspection.
over time are known to occur (e.g. central heating 14. If required, finger print experts cover the scene.
cycle), these should be appropriately logged. If the death 15. The scene is secured.
occurred out-of-doors, meteorological reports should be 16. Further examinations as required - all fully logged.
sought for the area to assist with such estimates. These
will have to include details about humidity, ground
temperature, wind-chill factor and rainfall. UNCLOTHING THE BODY
7. Exposure to carbon monoxide should be considered in
all domestic childhood and infant non-violent deaths, This should be avoided at the scene as it may lead to loss
and any potential sources of this gas, for example a of evidence, to potential contamination and perhaps also
blocked chimney, faulty kerosene heaters, blocked gas cross-transference of evidence to it. It is a better policy to
flue and coke burners, should be carefully looked for postpone the unclothing till later, and for this to take place
and expert advice sought about any possible under controlled conditions, in a good light, in a mortuary.
malfunction. This would enable each item of clothing to be removed
8. If possible, the scene should be secured after all the sequentiaJJy and to be carefully preserved. Contents of
searches have been carried out, as it may become pockets can be removed, identified and considered, and
necessary to return to the scene at a later time to check areas of damage to items of clothing can be identified and
over certain facts or to look for other evidence; new considered, as each layer of clothing is removed in
additional members of an investigative team also may sequence.
wish to see for themselves the scene of the fatal In babies, the clothing is of particular importance in
incident. attempting to assess whether the baby has been subjected
to excessive heat;l headgear, in particular, retains a signif­
icant amount of heat. If the clothing is soaked in sweat it is
SEQUENCE OF EVENTS AT THE DEATH SCENE a very good indication of excessive heat exposure (or of a
recent convulsion or hypoglycaemia). Inadequate, flimsy or
Once the crime scene manager has been designated, the wet clothing worn by the child in a cold bedroom may be
sequence of events at the death scene often follows the fol­ similarly of importance if a death from hypothermia is
lowing pattern of activity:5,6 being considered.
1. If the body is still in situ, establish a path to the body
that will be adopted by all those wanting to inspect
the body, this is referred to as th e Co 111 1110 17 Approach A GOOD LOOK AROUND
Pathway (CAP).
2. The CAP may have to be demarcated and preserved
From the room in which the body is found, the search has
by elevated platforms or stepping plates that have
to proceed to every other room in the house in a systematic
bee n thoroughly cleaned previously.
fashion. In addition to looking for other evidence, a good
3. If in the open, construct protection (e.g. tent) for the look around the house will give a better appreciation of the
body and its immediate surroundings. lifestyle of the parents and the other inhabitants of the
4. Set up a written log of named comings and goings to house, and also the prese nce of pets. The police will be alert
the scene. to signs of chronic controlled substance misuse and alcohol
5. (A medical practitione r or a suitably trained
misuse or overuse. This would also mean some degree of
paramedic formally pronounces life extinct).
attention to sites of waste disposal and garbage containers,
6. Provide adequate lighting or wait till daylight. to gardens and outhouses.
7. Photographs a nd video team document the scene
Special care must be taken in examining any bathrooms
fully.
and kitchens. In the latter, any partially consumed food or
8. Appoint an exh ibits' officer. made-up bottles of formula milk should be retained. These
140 I The death scene following the sudden death of a child
should be refrigerated to prevent fermentation, which leads
to spurious results of ethanol estimations. They can also be
tested for medical and illicit drugs as well as alcohol.
Any medication found within the household, in particu­
lar antipsychotics, antidepressants, benzodiazepines, anti­
convulsants and controlled drugs, should all be noted and
removed.
SUDDEN INFANT DEATH SYNDROME OR
NON-SUDDEN INFANT DEATH SYNDROME
The most common cause of post-perinatal motility in the
western hemisphere is still sudden infant death syndrome
(SIDS). However, as the incidence of SIDS has declined
worldwide,S the number of deaths related to abuse and
neglect has increased proportionally, and the task of distin­
guishing a natural fr om a suspicious dea th is becoming
increasingly frequent. 9,1O,1I In such instances, a thorough
scene investigation assists greatly in reconstruction of the
events that preceded the death. 12. IJ ,14
The contribution that death-scene examinations make in
relation to the definition of such infant deaths has been a
source of some controversy. In the 1969 defini tion of SIDS
produced by Beckwith,15 there is no mention of the scene
inspection. However, in 1976 there was a recommendation
by an American Working Party in their Special Report that
scenes of death should be visited. 16 In his second attempt at
more elaborately defining SIDS, during the second SIDS
International Conference held in Sydney in 1992, Beckwith
suggested the introduction of a scene investigation into a
three-tiered definition; this also failed to meet with general
approva l. During the second SIDS Glob al Strategy Meeting
held in Stavanger, a further attempt at making a satisfac­
tory scene investigation an essential criterion for the diag­
nosis of SIDS also failed.17 However the National Institute
of Child Health and Development's (NICHD's) definition of
SIDS has, since 1989, required that an examination of the
death scene be performed before a diagnosis of SIDS can be
made. IS
The definition of the scene has been widened 3 to incor­
porate not only the location where the infant was discov­
ered dead or unresponsive, but also the no rmal home
environment of the child and any other locations that may
have been visited by the child immediately prior to the
death. Protocols have been published from the United
States outlining the scientific and medical input in such
scene investigations. 19
If the baby has been found dead in a cot, crib, Moses bas­
ket, or something similar, then it is important to discover
the position in which the baby was found prior to the atten­
da nce of ambu lance crews and paramedics. Life-sized flex­
ible dolls have been made use of, in this respect, to enable
the carers of the child to recollect the initial discovery of the
child. Well over 50 per cent of babies dying of SIDS are
found in the prone sleeping position. s This reconstruction is
even more important, if the baby has been co-sleeping, in
order to assess the possibility of overlaying. 2I ,22
It is important to retrieve the bedding in order to dis­
cover whether this was stained wi th blood or saliva or was
covered in vomit, and also to assess whether it could have
made the baby's environment too warm or, ind eed, inap­
propriately cold.
When various theories related to the possible production
of inespirable and toxic gases from cot mattresses were being
investigated, it was considered important to retain the mat­
tress for appropriate testing.2J,24 This is no longer necessary.
Great caution must be exercised, if on investigation it is
discovered that the 'cot death' under investigation is the
second, or even more concerning, the third sudden infant
death in the same family. Greater suspicion and , therefore,
vigilance should be entertained in these circumstances. 25 ,26
EXTERNAL PETECHIAE
Although there is some dispute as to the percentage inci­
dence of the presence of petechiae on the serous membranes
of children dying of SIDS, there is no doubt that th e presence
of extern al petechiae in the face, behind the ears, in con­
junct ival membranes and on t he mucous membranes of the
lips and gums of a baby dying of SIDS is exception al. 27 ,2B
It is a simple enough test to carry out in a good light
with the baby in situ. If petechiae are found localized to
the upper t runk and face, in the absence of exposure to
meningitis or known coagul ation problems, then that par­
ticular baby's death has to be treated as suspicious; it
should be considered a result of mechanica l asphyxiation,
an accident or a criminal act, until proven otherwise by the
investigation.
Accidental mechanical asphyxial deaths do occur in
babies and young children and only an examination of the
scene of death would enable the mechanisms of death in
such instances to be elucidated. 29
BRUISING
The presence of bruises on children, and particularly
on babies, is always a cause for concern and suspicion,
particularly in children who have not started to toddle and
if bruises are of different ages . Bruising in unusual sites,
such as the back, arms, neck, chest, thighs and genitals, is
also particularly worrying in infants.
ABANDONED NEONATES
The abandoned body of a neonate is still a matter that
greatly exercises the sensitivity of the public and the interest
of the media. On few occasions, it is the consequence of
homicide. Nowadays, the emphasis is clearly biased towards

attempting to trace and then assist the mother of the aban­
doned baby, addressing any medical, social welfare or finan­
cial problems being experienced by the mother, rather than
in seeking punishment. The potential for a criminal tlial,
however, still exists, particularly if the body shows features
of injury and the mother is eventually traced. Therefore, the
police have to investigate the matter and carefully search the
scene where the body is discovered.
The incidence of neonate abandonment varies between
countries but even in Western countries such instances are
not exceptionaI. 2o ,21,JO,JI It is often the case that the baby
has been concealed prior to its disposal. Although putre­
factive changes are delayed in newborn infants, owing to
the sterile environment from which they have been deliv­
ered and due to the absence of bacteria in their intestines ,
some decomposition, albeit incomplete, may be present.
Over time, there also may be partial mummification of the
body in dlY and well-ventilated surroundings. Often the
body of the child has a lready been attacked by animals and
bears the marks of this. These have to be distinguished
fioman~-mortemtrauma.
In all these instances it is important to ensure that a
sea rch of the scene is made for the placenta, and for any
other items that may have been discarded together with the
baby, for example distinctive baby clothes, newspapers and
wrappings; this may give a clue to the locus at which the
baby was delivered. Occasionally, notes may have been
pinned to the body and sometimes items of jewellelY left
with it; forensic examination of these may ass ist with trac­
ing the mother. This may be indispensable in the identifi­
cation of the abandoned body, which is often a major
problem and often rema ins unsolved.J2.J 3
Infanticide has been committed in a number of ways,
such as intentional neglect, exposure, killing with a
weapon (e.g. stabbing or by the head being struck against
a solid surface). suffocation (by the insertion of foreign
material into the mouth or with a soft pillow) and by
drowning (see Chapter 18). In all instances, it may be use­
ful to attempt to distinguish whether trauma at the time of
birth was brought about by inexpeJiise in delivery or by
precipitate labour from intentional injury infliction. 34
Occasionally, abortuses and products of conception
are found and the police may have to investigate. It is
essential, in the first instance, to ensure that one is indeed
dealing with human remains. Products of conception of
non-human origin may closely reassemble human material,
for example rabbits and other small mammals. Some years
ago, in many co untries, toy manufactures produced android
toys, which had the same size and general appearance of a
fetus, and many a n investigation had to be abruptly termi­
nated with a ceJiain amount of 'red face' all around when it
was discovered that it was not a human fetus which had
been discarded but one of these toys, sometimes as a prac­
tical joke, in poor taste, or simply accidentally.
Another common finding is the presence of bones of
babies that appear to have been lying undisturbed for
Dyadic and multiple deaths I 141
decades and which come to light with renovations and dem­
olition of old buildings (see Fig. 10.2, p. 183), a remnant of
the days where the birth of an illegitimate child was a mat­
ter of public shame and dishonour and contraception was
primitive. These cases are velY difficult to investigate, not
least because with the passage of time the cause of death and
indeed the potential perpetrator cannot be identified. J5 An
attempt will usually be made to assess the age of the skele­
ton to determine whether it is pre- or post-viable.J 6,37
DEATHS FROM TRAUMA
A careful review of the scene is a sine qua non in traumatic
deaths, particularly when different scenarios have been
alluded to in statements given by those who were there
when the injUlY took place. The potential veracity of one or
other scenario has to be investigated by a careful inspec­
tion of the scene, with measurements being taken as appro­
priate22.J8,39
A debate still rages over whether a head impact against
a solid unyielding surface, causing acute deceleration, is
essential to produce the pathological changes that a re
ascribed to the 'shaken baby' syndrome.4o ,41 Evidence for
such an impact may be scanty, such as small indents in fur­
niture or the deposit of a sma ll amount of hair and skin
from the baby, but shou ld be carefully sought.
Deaths in Fires
In Britain, particularly in Scotland,42 there are a number of
deaths every year in which children are killed by household
fires . This incidence has hardly changed over the years
despite major educational attempts by fire brigades and cen­
tral governments to ensure that houses are fitted with work­
ing smoke alarms, and discouraging the cultures of smoking
in bed or cooking while under the influence of alcoholY,44
The deaths may be accompanied by adult deaths,45,46
and it is frequently the case that smoke inhalation is the
cause of the death. The children may be too young to be able
to find their way to safety or may become disorientated
and terrified at an early phase and succumb to smoke
inhalation before reaching safety. The possibility of drug or
alcohol admin istration to children in this circumstance
should not be forgotten.
The dynamics of the commencement and spread of the fire
and the use of accelerants will be reported on by the special­
ized fire officers and the forensic team that attends the fire. In
all such cases, the possibility of arson should be kept in mind.
DYADIC AND MULTIPLE DEATHS
A dyadic death is one in which the perpetrator of a homi­
cide later commits suicide,47 often at the same site where
142 I The death scene following the sudden death of a child
the homicide occurred, with multiple bodies found in the
same scene. Many of these instances of homicid e cum sui­
cide are th e consequence of severe, occasionally previously
undiagnosed, mental disorde rs, such as severe dep ression
or paranoid schizophrenia. In many instances, there is an
acute, emotional trigger for such occurrences, for example
bankruptcy, prosecution of the fath er for some alleged
offence, the discovelY of a n extramarita l affair, an impend­
ing threatened sepa ration or divorce. Men and women can
both be involved in such deaths. Often, an entire family
can be eliminated in the same event. Mechanical asphyxial
methods, fire arms and hefty sharp weapons, suc h as axes
or machetes, are made use of.
The wea r and tear of modern living appears to have
taken its toll, resulting in an increasing incidence of such
deaths. These scenes are particularly harrowi ng for the
investigators, but in all these instances it is essential to
ensure that there need not be a search for another person.
Even more harrowing and soul-destroying are scenes in
which many children have met their death together in a
violent manner. In major incidents of natural or unnatu ra l
causation, children are often among the fat a lities. It is,
however, the exception al and unusua l situation in which
all the victims in a majo r incident are children. 48 The noto­
rious exception in this co unt ry was the incident in Dun­
blane primary school; othe r countries, such as the USA
have had to deal with several simil ar incid ents . These inci­
dents have to be dealt with as if one is dealing wi th multi­
ple single homicid es, each decedent being treated
individually as a vic tim of homicide. The scene in its
entirety, no matter how large, becomes a scene of crime.
SUDDEN DEATHS OF OLDER CHILDREN
PsychiatIists are beginning to build up a better profile of
child sex abuse rs and killers. Most child murders are carried
out by fami ly members - often mentally ill. In other cases
the perpetrators have been abused as a child themselves.
However, there are also a small number of paedophilic crim ­
inals responsible for child homicides. In such cases, the chil­
dren involved are older children and the motive for their
murder is sex ual g rati fication of the perpetrator. These clim­
inals will do their best to conceal their crime and it is often
the case that the body of the deceased child will not be fo und
until an exhausti ve search has been carned out; when the
body is eventua lly found, it is often in a state of decomposi­
tion. The exa mination of such scenes is even more fraught
and requires a meticulous attention to detai l.
In order to glean as much information as is possible
fro m the sites where such bodies are deposited, in an effort
to solve the crime, it is often the case th at the expeltise of
a number of other experts may have to be called into play.
These include, for example, arc haeologists a nd anthropol­
ogists, soil experts, experts in pollen and entomologists.
The final resting place of the body may well differ from the
site where the murder has been perpetrated or whe re the
body has been stored prior to its disposal. In such
instances, more than one scene will have to be carefully
examined once these have been identified, each being
organized and supervised by a different CSM.
To enable a ll the various experts to congrega te at the
scene, it may be necess ary to retain the body in situ for a
lengthy period. occasionally of several hours' duration. The
public media may suggest that this is a callous and insen­
sitive way in which the police de al with such gri m discov­
eries ; from personal experience, there is no substitute to
ensuring that such investigations are thorough, compre­
hensive and as all-encompassing as is possible. Delay in
carrying out the autopsy is a sm all penalty to pay for this.
Id entification of the victi m becomes paramount in such
situations, and the assistance of forensic odontologists and
deoxyribonucleic acid (DNA) exp erts are invaluable in suc h
instances.
SENSITIVITY AND STRESS OF
THE INVESTIGATION
Deaths of babies and children are becoming increasingly
uncom mon and thus when such a death occurs, particu­
larly if it is sudden or violent, it wi ll cause major emotional
upheaval and grief in the immediate family. If there is an
index of suspicion that has to be investigated and excluded
then great care must be taken to ensure that the fami li es
are treated with resp ect and sensitively at all times, even
when there is high index of suspicion re lating to their
direct involvement in the death .
The police have, over recent years, trained officers as
'family li aison officers' who can speak to and communicate
with the families, keeping them informed of the progress of
the investigation. There wilJ be a certain amo unt of antag­
onism and animosity shown towards the police and what
they are perceived to represent. It may be n ecessa ry to
liaise with primary medical care rs and the social wel far e
department to ensure that the family gets best treatment.
If it is thought that other children in the same family may
be at risk of physical injury or neglect, the appropriate social
welfare legislation must be utilized to safeguard other siblings.
The family may wish to view the body of the deceased
baby: they have a right to do so, a right enshrined in the
European Convention of Hum an Rights. Unless there are
very good reasons to the contrary, their wishes should be
accommodated as soon as possible. The dead baby should be
presented in surroundings that make this harrowing experi­
ence less stark. The infant ca n be held and touched and
mementos, such as a hair lock or an imprint of the hands or
feet in ink, wax or plaster of Paris be provided.
Matters relating to post-traumatic stress cannot be
ignored or side-stepped for any reason , not least the possi­
bility of issues of compensation. The pathologist at the scene
may feel that the officer in charge of the scene, the senior

investigating officer (SIO), should set in motion appropriate
provisions, which his police force will p ut in place. In cases
involving mutilation of the body, whatever the cause, this
attains an even more acute importance.
INQUESTS AND INQUIRIES
All that transpires at the scene will be scrutinized in court.
This may be a coroner's court where a n inquisitori al
scheme of thin gs is in operation, or in subseq uen t criminal
cases in which an adversarial and confro ntatio nal system is
at wo rk, an d all aspects of the investigatio ns may be a nd
a re likely to be scrutinized through the fine sieve of cross
examinatio n.
Certa in cases may also come to the attention of Tribunals,
Public Inquiries or other judicial investigations in which dif­
ferent nuances of procedure may hold, but evidence will stil l
have to be printed under oath or affirmation.
REFERENCES
Wagner GN. Crime scene investigation in child abus e cases.
Am J Forensic Med Pathol 1986; 7:94-9.
2 Usher A. Th e role of the pathologist at the scene of crim e.
J Fo rensic Sci Soc 1970; 10:213-8.
3 Bass M, Kravath RE, Class l. Death scene investigation in
sudden infant death. N Engl J Med 1986; 315:100- 5.
4 Locard E. The analysis of dust traces, Parts I-Ill . Am J Police
Sci 1930; 1: 276- 298,405-406, 496-514.
5 Vanezis P, Busuttil A. Suspicious Death Scene investigation.
London: Hodder Arnold, 1996.
6 Wh ite P. Crime Scene to Court. The Essentia ls oj Forensic
Science. Lond on: The Roya l Soc iety of Chemistry, 1998.
7 Stanton AN. Sudden infa nt death. Ove rheat in g and cot death.
Lancet. 1984; 2: 11 99-201.
o Byard RW, Krous HF. Sudden Injant Death Syndrome. Lond on :
Arnold, 200 I.
9 American Academy of Ped iatrics. Distinguishing sudden
in fant death syndrome from child abuse fata lities. Pediatrics
1994; 94( I): I 24 - 6.
10 Christoffel KK, Zieserl EJ, Chiaramonte J. Should child abuse
and neglect be considered when a child dies unexpectedly?
Am J Dis Child 1985; 139:876-80.
II EmelY JL. Chi ld abuse, sudden in fant death syndrome, and
Kong jVled J2006; 12: 6 1-4.
unexp ected infant death. AnI J Dis Child 1993;
33 Herman -Gidd ens ME, Smit h JB , Mittal fvJ et al. New born
1476( 10) : 1097 - 100.
killed or left to die by a pa rent: a population-based study.
12 Bya rd RW, Carm ichae l E, Bea l S. How useful is the post­
JAMA 2003; 209:1425-9.
mortem examination in sudden infant death syndrome?
34 Mitchell LB, Dav is JH. Spontaneous births ill to ilets.
Pediatr Patho1 1994; 14:817- 22.
IJ Bya rd RW, Becker LE, Berry 1. Formal retrospective case
35 Kerley ER. Forensic Anth ropology and crime involving
review and sudden infant death. Acta Paediatr 1997;
ch il dren. J Forensic Sci 1976; 21[2):33 3- 9.
86:10 11-1 2.
36 Hargrave DR. A study of chil d hom icide over two decades.
14 Newlands J, Emery JS. Child abuse and cot deaths. Child
Med Sci Law 1992; 32[3): 196- 245.
Abuse Negl 1991; 15(3):275- 8.
37 Kerley ER. The ide ntificatio n of battered- infant skeleto ns.
15 Beckwith JB. Discussion of terminology and defin iti on of th e
sudde n in fant death syndrom e. In Bergma n AB, Beck with JB ,
Ray CD (eds) Sudden Infant Death Syndrome. Proceedings oj
the Second international Conjerence on Causes oj Sudden
Death in Injants. Seattle: Uni versity of Washi ngto n Press,
1970, pp. 14- 22.
Referen ces I 143
16 Jones AM, Weston JT. Examination of the sudden infant
death syndrome: investiga tive and autopsy protocols.
J Forensic Sci 1976; 21 :833-41.
17 Hilton J, Berry JP. Pathol ogy. In Fitzgerald K (ed.) Second
S/DS Global Strategy Meeting. Norway: Stavenger, August
5-6, 1994, p. 334. In Rognum TO (ed.) Sudden Infant Death
Syndrome. New Trends in the Nineties. Oslo: Scandinavian
University Press, 1995.
18 Willinger M, James S, Catz C. Definin g the SIDS: deliberat ions
of an expert panel conceived by the National Inst itu te of
Child Health and Human Development. Pedilltr Pathol J 994 ;
II :677-84.
19 Center for Disease Control (CDC). Guidelines for dea th scene
in vest igatio n of sudden , unexplained infant dea th s:
reco mmendations of the Inter Age ncy Panel on Sudden Infant
Death Sy nd ro me. MMWR Rec0111111 Rep 1996; 45:RR-JO.
20 Iyasu S, Ha nzlick RE, Rowley D, Wi llin ger M. Pro ceedings of
Workshop on Guidelines for Scene Investigation of Sudden
Unexp lained Infant Deaths. J Forens ic Sci 1994; 39:
1126- 39.
21 Dragon DA, Lan denberg Ai. Infant mechanical suffocation
deaths in the United States, 1980-1977. Paediatrics. 199 9;
103 :59.
22 Nakamura AS, Wind M, Danello MD. A review of hazards
associated with children placed in adult beds. Arch Pediatr
Adolesc Med 1999;153:1018-23.
23 Fact Sheet on the Tox ic Gas Hypothesis of Cot Death,
www.sids.org.uk/ed itpics/364-I.pdf
24 Coghlan A. .Mattress bugs may link to cot deaths. New
Scieutist 2002; 10:41.
25 Meadow R. Recurrent cot death and suffocation. Arch Dis
Ch ild 1989; 64(1): 179-80.
26 Emery JL, Gi lbeli EF, Zugu ibe F. Three crib deaths.
A babyminder a nd probab le infa nti cide. Med Sci La w
1988 ; 28(3) :205 -11.
27 Byard R, Petechia l hemorrhages and unexpected infa nt deat h.
Legal Medicine 2003; 1(4): 193-7.
28 Betz P, Hausmann R, Eisen menger W. A contribution to a
possible differe ntiation betwee n SIDS a nd asp hyxiatio n.
Forensic Sci Int 1998; 9 1: 147 -52.
29 Shepherd RT. Accidenta l se lf-s trangula tion in a yo un g chi ld.
Med Sci Law 1990; 30: 11 9-23.
30 Mend lowicz MV, Jean -Lo ui s G, Gekker M, Rapaport MH.
Neonaticide in the city of Ri o ele Janeiro: forensic and
psycholegal perspectives. J Forensic Sci 1999; 44:741 - 5.
31 Cheung PI. Materna l filicide in Hong Kong, 1971-85. jVled Sci
Law 1986; 26: 185 - 92 .
32 Lee ACW, Li CH, Kwong NS, So KI. Neonat icide, newborn
abandonm ent, and denial of pregnancy - newborn
victimisation associated with unwanted moth erhood. Hong
J Forensic Sci 1984; 29(2):591-6.
J Forensic Sci 1978; 23(1): 133- 8.
38 Kasim MS, Cheah I, Shafie HM. Childhood deaths from
physical ab use. Child Abuse Negl 1995 ; 19(7):84 7- 54.
39 Kellet RJ. Infanticide and ch ild des U'llCtion - the historica l,
legal and pa thological aspects. Forensic Sci Int 1992:
53 (10):1-28.
144 I The death scene following the sudden death of a child
40 Alexander R, Sato Y, Smith W, Bennett T. Incidence of
impact trauma with intracranial injuries ascribed to shaking.
Am] Dis Child J990; 144 :557-63 .
4J Duhaime AC, Christian CW, Rorke LB, Zimmerman RA.
Non-accidental head injury in infants - the 'shaken-baby'
syndrome. N Eng] Med 1998 ; 330:1822-7.
42 Squires T, Busuttil A. Child fatalities in Scottisb house fires:
a case of child neglect? Child Abuse Neg l J995; 19:6 5-73.
43 Roberts I. Deaths of children in house fires. EM] 199 5;
311: 1381-2.
44 Squires T, Busuttil A. Alcohol and house fire fat aliti es in
Scotland. Med Sci Law 1997; 37:325-40.
45 Runyan CW, Bangdiwala IS, Lin zer MA et a1. Fatal house
tires: who dies and who su rvive? ]AMA 1998; 279:
1633- 7.
46 Istre GR, McCoy BS, Barnard JJ, Bolton A. Deaths and injuries
from house fires. N Eng] Med 2001; 344(25): 1911-16.
47 Buteau J, Thompson B, Sexton D. Homicide followed by
suicide: a Quebec case-series, 1988 -90. Cal1 ] Psychiatry
1993; 38(8):552-6.
48 Busuttil A Lockerb ie Ct Dunblane. Disasters and dilemmas.
Med Leg] 1998 ; 66(4):34-42.
 I CHAPTER 8 I
POST-MORTEM EXAMINATION IN
BABIES AND CHILDREN
Jean W Keeling

Introduction 145 Weights and measurements 150

Death scene investigation 145 External examination 150
Rectal temperature 146 Estimating blood loss 152
Medical and family history 146 Dissection (infants and older children) 152
Other important information 146 Examination of the brain, spinal cord and eye 156
Radiological examination 146 Exam ination of the newly born 158
Photography 147 Histological samples 162
Microbiological samples 148 Retention of organs 163
Toxicological investigations 149 Exchange of information and multidisciplinary review 163
Biochemical and metabolic investigations 149 References 164

INTRODUCTION DEATH SCENE INVESTIGATION

This chapter considers the investigation of deaths in the
perinatal period, infancy and early childhood. The investi­
gation of deat hs in all th ree groups follows a similar gen­
eral line, but there are some diffe rences in techn iqu e,
emphasis and even in interpretation of the same patho­
logical findings, for example the presence of facial
and upper trunk petechiae, among the gro ups. Such differ­
ences wi ll be emphasized at appropriate points within the
chapter.
Sudden death in early life may be the outcome of a var­
iety of natural diseases (see Chapters )0-12) as well as
acc idents and homicide. Sometimes, a pre-existing con­
genital or acquired disorder may predispose a child to acci­
dents and, possibly, to deliberate assault. In an y sudden
death , whether or not a cause of death is precisely
defined an d particularly when some necropsy findings ,
such as petechial haemorrhages, can be the res ul t of natu­
ral disease, it is most important that a full range of inves­
tigations is carried out and that their results are
clearly presented. 1,2 Failure to do so weakens any assert ion
of non-natural death and enhances the scope of cross­
examination.
Thorough appraisal of the locus is as relevant to deaths
occurring in the home as it is to motor vehicle collisio ns and
other incidents away from the domestic environment. Bass
et al3 and Sturner4 stress the importance of detailed eval ua­
tion of the death scene in unexpected infant deaths. This
includes the physical state of the infant's place of sleep and
the precise position of the in fa nt when found. The latter is
particularly important when positio nal asphyxia is a possi­
bility and when co-sleeping is practised. The use of a life­
sized doll may assist the recall of witnesses and improve the
precision of information about position when fou nd. 4
Evaluation of an inappropriate sleeping environment in
respect of temperat ure, humidity and the presence of
noxious gases can only be made by prompt evaluation of
the scene. 3 ,5
Seizure of feeding bottles and domestic items, such as
jugs and spoons used in preparation of feeds, for toxico­
logical analysis is important. The contribution of drugs and
alcohol to sudden deaths in babies and children is probably
un de resti mated.
A visual record of the death scene can aid the interpret­
ations of necropsy findings and high-quality photographs
146 I Post-mortem examination in babies and children
or video recording are often extremely useful. Death scene
investigation is discussed in detail in Chapter 7.
RECTAL TEMPERATURE
Measurement of rectal temperature at the death scene or on
arrival at hospital may be a useful corroborative investiga­
tion. If elevated, it raises the possibility of pyrexial illness
or an inappropriately warm sleeping environment.
There are no tables or formulae relating rectal temperature
and likely time of death that are appropriate for use follow­
ing the death of babies or children. Surface area in children is
relatively greater and body mass less than that of adults, so
that cooling is likely to be more rapid and probably more
influenced by the amount of clothing and other coverings. 6
MEDICAL AND FAMILY HISTORY
Information about the individual's medical history and
family background should be available to the pathologist
prior to necropsy examination.
Details of pregnancy, birth and progress in the neonatal
period are essential in infant deaths, together with a
detailed account of recent illness and of events in the 48
hours prior to death.
In older children, a brief history of birth and early life
may be sufficient; however, information about any ill­
nesses and accidents, particularly those resulting in hos­
pital admission, are important.
Information about members of the immediate family
should include major problems of a social, legal or finan­
cial nature, recent illnesses and documentation of any
familial disorders. Information about sudden death or col­
lapse in the wider family, particularly in early life, should
be specifically sought. A history of many consultations
with the general practitioner or hospital, either by the
deceased or siblings, is important information for the
pathologist. Multiple hospital attendances can be difficult
to ascertain, as children may have been taken to a variety
of different institutions.
The 'At Risk' register, or its equivalent, should be
checked carefully, not only for the deceased individual, but
also for siblings and half-siblings, ta king cognisance of the
multiplicity of surnames used in some of today's compli­
cated family units.
Information about prescribed medication for the decedent
and any other family member is required, togetl1er willi
information about the availability of recreational drugs and
alcohol in the decedent's immediate environment.
OTHER IMPORTANT INFORMATION
Information about the deceased's mobility and physical
capabilities are important to the interpretation of any
injuries present, particularly in the infant period when
there is much individual variation. Non-mobile infants
should have few, if any, bruises. 7 ,B Awareness of supervi­
SOlY arrangements around the time of death is also impor­
tan t. An inexperienced carer may unwittingly permit
hazardous activities or inappropriate freedom and be less
aware than parents of recent recommendations about best
childcare practice. 9 Carers who are distracted by illness,
depression or other adverse events, such as marriage break­
down and bereavement, are likely to be Jess vigilant in
respect of both symptoms of illness and environmental
hazard.
RADIOLOGICAL EXAMINATION
A fuJJ radiological examination is an essential part of the
evaluation of any sudden death in infancy and childhood,
and this is further discussed in Chapter 3. Necropsy exam­
ination should not be contemplated in establishments
where radiography is not available. It is better to accom­
modate a small delay to necropsy so that radiological evalu­
ation can be carried out rather than having a major
disruption to dissection or an incomplete examination after
the post-mortem examination has been completed.
A single-film 'Babygram' is not sufficient; it is esse11tial
to have a complete set of appropriately penetrated regional
views. The radiographs should be taken by an appropri­
ately qualified and experienced radiographer. This individ­
ual is important for his/her expertise in producing the
appropriate views to encompass classical injUly sites and
ensuring technical excellence. An experienced radiog­
rapher may suggest additional films. Take the advice!
Recent rib and skull fractures cause most problems for
the radiologist; however, a pathologist is unlikely to miss
them as they can be observed directly and should always
be carefully sought. Small fractures in long bones in the
vicinity of joints (see Chapter 3) are best demonstrated
radiographically but require histological confirmation.
Histological examination of any area when there is dubiety
about the nature of a radiological abnormality is essential.
Sections through identified fractures can contribute to
assessment of the age of the fracture.
In any suspicious death, it is important that the radio­
graphs are reviewed prior to necropsy examination by an
experienced radiologist, preferably one with paediatric
experience. A verbal report will suffice.
It is often useful to X -ray individual bones or regions
(i.e. rib cage) following removal during necropsy. In partic­
ular, it ensures better views of posterior rib fractures (Fig.
8.1), which, being close to the costovertebral articulations,
are often obscured by thoracic viscera. Localized abnor­
malities of uncertain aetiology may be positively identified
using this procedure. Re-X-ray of a skull vault with
wormian bones ensures an optimal record and avoids later
argument.

Figure 8.1 Radiographic examination of an infant admitted
with subdural haemorrhage. (a) Portable radiograph taken on
the intensive care unit; rib fractures are difficult to see.
(b) Anteroposterior radiograph following evisceration; posterior
rib fractures are readily visible. (c) Portion of rib cage X- ra yed
after removal from the body; fine detail of fracture and callus aids
dating of injury.
Photography I 147
Radiological examination adequately documents gas or
air within chest, abdomen and pericardium and within
large vessel s (e.g. pulmonary arteries. portal venous system
and cerebral sinuses).
Cranial axial X-ray, following instill atio n of contrast
medium into the superior sagittal sinus. has been used suc­
cessfully to demonstrate torn bridging veins in infants. 10, II
Other modalities. such as magnetic resonance imaging.
may be useful in particular cases but are usually more dif­
ficult to arran ge (see Chapter J).
PHOTOGRAPHY
High-quality photographs are an important part of the
record of any necropsy on infants or children and are par­
ticularly important when investigating sudden deaths.
Whole-body views give a useful visual reminder of body
proportions and sta te of nutrition (Fig. 8.2). and assist in
the local ization and quantification of cutaneous injuries
such as burns or sca lds.
Close-up photographs of injuri es, with and without a
scale, and sometimes from different angles. permit assess­
ment of injuries much more effectively than pages of
Figure 8.2 Posterior view of whole body prior to necropsy.
There is little subcutaneous fat, muscle bulk is reduced, and
napkin dermatitis and hypothermic injury are present - asa
resu lt of inadequate care.
148 I Post·mortem examination in babies and children

(a)

(b)

... f ~ ,. ~ '\ > . ~'

i" ) .L./~""t)'% ),.\ ..:,: -)..., ~' , 1

Figure 8.3 (a) Bruises photographed on admission to an .

"
intensive care unit. (b) Comparable photograph taken 36 hours -,
later, prior to necropsy.

detail ed written description and measurements ever can.

Good photographs of injuries mean that referral to experts .r
is more likely to produce a useful opinion. They are most
helpful for a defence pathologi st and are particularly use­
ful when cases are re-opened or charges brought some
years after the death occurred. Additionally they are a
helpful aide-memoire before court appearance.
It is important that th e pathologist has an 0ppOltunity
to rev iew all of the photographs taken following a violent
death, and not merely those sel ected for possible court use.
Comp arison of photographs taken during life, for ex ample
on the intensive care unit, with post-mortem photographs
may enable a better estimate of the age of bruises than either
Figure 8.4 (a) Bruises photographed prior to necropsy.
set ex amined in isolation (Fig. 8.3). Sturner4 emphasizes the
(b) Image (a) re-photographed 24 hours later.
importance of a photographic record of petechiae and injuries
on admission to hospital as changes take place quickly.
A further external examination and re-photography of
It is good practice to obtain a core set of samples as early in
cutaneous injuries on the day following necropsy often
the examination as possible (Table 8.1) to minimize contam­
yields useful information (Fig. 8.4), as may photographs
ination. In some institutions, these samples are taken in the
taken using ultraviolet light.
acciden t and emergency department (emergency room)
according to a protocol agreed between pathologist an d clin­
IVIICROBIOLOGICAL SAMPLES ical coll eagues, with prior authorization of the legally
responsible department. IJ Additional sa mples are sometimes
Microbiological investigations are an important palt of the indicated by necropsy findin gs. The expansion of molecular
investigation of sudden death in both infants and children. 12 methods in microbiology has resulted in quicker results and

Table 8.1 Microbiological samples usefully token in sudden
death investigation in early life
Bacteriological samples
Nose/throat swabs
Cerebrospinal fluid
Blood culture
Bronchial swab
Lung
Large bowel or rectal swab
Virological samples
Nose swabs/nasopharyngeal aspirate
Tracheal ring
Lung
Heart
Small intestine
has extended the period over which samples may be use­
ful, 14, 15 but the significance of detection of viral RNA must
be considered in association with relevant histopathological
changes in appropriate organs.
Although microbiological investigations are positive in
only a small proportion of cases, negative results are partic­
ularly important in the sudden unexpected deaths in infancy
(SUDl)/sudden infant death syndrome (SlOS) context l6 and it
is important that these investigations are carried out, even
when death seems likely to be the result of trauma; failure to
do so can make for vexatious cross-examination.
It is worth remembering that some microbiological causes
of sudden death in early life, such as group B streptococcal
and some viral infections (such as echovirus), elicit no spe­
cific findings at either gross or microscopic level and so that
positive cultures are essential for diagnosis.
TOXICOLOGICAL INVESTIGATIONS
Toxicological investigations are not undertaken routinely in
many jurisdictions as part of the investigation of SUD! or
sudden death in older children. There may be cost con­
straints or perceptions of a low return for effort. This is par­
ticularly so in fetal deaths and amongst those infants found
dead in their cots on household awakening.17 Sims and
Collins 18 found evidence of drugs, particularly recreational
drugs, in 21 per cent of fetal deaths that were investigated
in a forensic pathology depa11ment. Such investigations are
more likely to be done when concerns are expressed in a
police rep0l1. Pathologists should be ready to recommend
that, as a minimum, screening procedures for some types of
medication and recreational drugs are carried out and
should ensure that appropriate samples are retained for
more detailed investigation if indicated. Drugs with hyp­
notic or sedative effects are sometimes given inappropli­
ately to keep an infant quiet. It should not be forgotten that
infants are inquisitive and casual ingestion is always a
Biochemical and metabolic investigations I 149
Table 8.2 Useful toxicological samples in the investigation of
sudden death in early life
Sample Toxological agent
Blood Alcohol
Com mon sedatives
Urine/bladder washout Common analgesics
Recreational drugs
Carers' prescribed medication
Recreational drugs
Lung (whole) Volatiles
Hair - cut
Earlier/long-term drug ingestion
Hair - plucked
Stomach contents Common sedative/analgesics
possibility in mobile infants and small children when pre­
scribed medication or recreational drugs are present in their
environment. Alcohol ingestion can produce rapid and
marked metabolic dis turbance in the young.
Rapid onset of drowsiness/incoordination and metabolic
disturbance, such as hypoglycaemia, are indications for full
toxicological investigation. Paracetamol (acetaminopohen) is
recommended only on presCliption for infants < 3 months of
age. 19 A single dose may be given following immunization.
Detection of non-prescribed paracetamol in young infants
requires a full explanation.
Suitable samples for toxicological investigation are tabu­
lated (Table 8.2).
BIOCHEMICAL AND METABOLIC
INVESTIGATIONS
Post-mortem biochemical investigations on blood samples
are unsatisfactory on the whole, with the exception of
toxicological investigations. Vitreous humour, urine and
cerebrospinal fluid (CSF) are much more useful, although
normal ranges for CSF values for the paediatric age group
are often not available in individual laboratories. Some
groups regard biochemical investigations as part of the
basic SUDI protocol, 13,20 whereas others only perform them
when there are historical features or abnormal pathological
findings. 17 Sampling the vitreous must be omitted if
it is likely that histological inv estigation of the eye is
important.
Whilst a range of investigations is possible, the size of
the sample is usu ally the limiting factor. This is a particu­
lar problem with urine, as the bladder is empty in very
many SUD!. A bl adder washout with - 10 mL of normal
saline can be used to investigate the possibility of organic
acid abnormalities. It can also detect low levels of com­
monly used analgesics when the blood levels are within the
therapeutic range (personal observation) .
150 I Post-mortem examination in babies and children
Prioritization of investigations is usually necessary and
is based on a combination of history and pathological
findings.
Sodium, urea and osmolality are useful baseline investi­
gations JJ and levels are generally stable in vitreous
humour for about 72 hours. Sodium and osmolality pro­
vide corroborative evidence of hypernatraemia or dehydra­
tion. An elevated level of urea is not usually found until
2-3 days into an illness and provides a general indicator of
subacute illness.
Glucose levels in vitreous humour are not so stable but
are useful for up to 12 hours after death,13 which makes
their usefulness at time of necropsy questionable.
The reliability and usefulness of post-mortem investiga­
tions is discussed further in Chapter 5.
In recent years, molecular methods have replaced
lengthy biochemical investigations, with somewhat
unceliain results, in the identification or exclusion of a
number of genetic metabolic disorders (GMDs) that might
underlie some SUD!. That these investigations can be
carried out on frozen tissue or even thick sections from
paraffin embedded tissue (especially spleen) is a
particular advantage, as fluid samples can then be used for
other tests.
Molecular methods on tissue samples can also detect
the molecular rearrangements of some forms of long QT
syndrome. 21
As the potential for molecular investigation of more and
diverse genetic disorders continues it becomes essential to
reserve both frozen and formalin-fixed wax-embedded
material indefinitely when sudden death in infants and
children is initially unexplained or incompletely explained .
Another sample that has proved useful in the detection
or exclusion of some GMD is the Guthrie card blood
spot. Wilcox et al 22 undertook tandem mass spectrometry
for acyl-carnitines. They identified three GMDs amongst
247 SUm.
WEIGHTS AND MEASUREMENTS
The value of recording body weight and external measure­
ments is much greater in the post-moliem evaluation of the
fetus, infant and child than it is in adults.
For full assessment of the fetus and infant, body weight,
crown-rump (sitting height), crown-heel (standing height)
and foot length, together with the occipitofrontal head cir­
cumference and biparietal diameter, are impoltant. These
are compared with normal values 23 - 25 (see Appendix B,
p. 471) and provide information about gestational age in
the fetus and the appropriateness of postnatal development
and nutritional status in babies and infants.
In older children, body weight, standing height and head
circumference give information about nutritional status and
appropliateness of development. Normal values by sex are
readily available in chart form from paediatric wards or
outpatient departments. The most widely used charts in the
UK are based on the work of Freeman et al 26 (Appendix B).
Weight and length are important when estimates of
blood volume and calculation of administered drug doses
are required.
Caution should be exercised when comparisons are
made between body weights recorded during life and post­
mortem weight. Weight in life may include clothing, a
number of different weighing scales may be used and any
may be inaccurate. (When were your mortuary scales last
inspected and calibrated?)
Weights may not have been recorded accurately - check
them yourself if in doubt, especially when using an
unfamiliar mortuary or when examination of babies is
carried out infrequently. Has there been a transcribing
error? Go back to the original record and compare with
that in the report.
Accurate organ weights in babies and children can pro­
vide pointers towards natural disorders and may be useful
for corroboration when investigating postoperative deaths.
An electric balance calibrated in O.l-g increments is appro­
priate and inexpensive. Gestation-related normal organ
weights for comparison from 12 to 42 weeks' gestation are
those of Singer et al,27 Hansen et al 28 and Maroun and
Graem 25 (Appendix B) . Age-related organ weights in
infants have been collated by Thompson and Cohle 29
(Appendix B). Organ weights in older children are more
difficult to access. The most recent papers, 30, 31 and Haddad
et ai, are statistical analyses of older data. Age-related
weights of major organs are found in Altman and Dillmer32
(Appendix B). The brain should be weighed before and after
fixation, being aware of a potential confounding factor
when different balances are used.
EXTERNAL EXAMINATION
A detailed external examination by region is essential, par­
ticularly when there are injuries or external changes of
uncertain aetiology present. Outline figures with infant
body proportions 33 are invaluable in the recording of focal
lesions (Fig. 8.5). Tbe general description includes racial
origin, clothing, jewellery, skin colour, regions affected by
livor mortis (including areas of sparing). extent of rigor
mortis, cleanliness and state of nutrition .
Signs of medical intervention are listed separately from
other external features. Kaplan and Fossum 34 describe
resuscitation injuries of the fa ce and neck, and relate them
to different types of apparatus used in resuscitation .
A careful search is made for petechial haemorrhages,
including eversion of both eyelids to examine tarsal plates,
examination of the oral cavity and behind the ears.
Petechiae are unusual in healthy infants, 35 and are unusual
in SUD! that remain unexplained. J6
Careful note should be taken of any secretions emanating
from the mouth or nose. Blood staining or streaking should
 External examination I 151

Name ____________________________ Age __________ Race _ ________ Sex ________

Autopsy number ________________________________________ _ _________ Date ________

Figure 8.5 Body surfaces with infant proportions aid accurate recording of injuries (after ref. 33).

be described carefully and photographed, with the appear­ Any tears or bruising of the frenula of the tongue and
ance (e.g. fro thy, fluid, dried) noted. When secretions are lips are recorded (Fig. 8.6), as well as scarring suggestive of
bloody, Krous et ae 7 urge careful examination of the mucous old injUly. The impression of teeth (even in an edentulous
membra nes of the nose and mouth with an a uroscope. infant) and patterned bruises of the inner aspects of the lips
152 I Post-mortem examination in babies and children
(b)
Figure 8.6 Injury to the upper frenulum (a), accompanied by
excoriation of the lower lip, with bruising of the adjacent alveolar
margin (b).
suggest peJioral pressure. Examination of the pharynx for
injury, secretion or foreign bodies can be carried out easily
using an infant laryngoscope. Remember that a foreign
body may have been removed in accident and emergency
but associated abrasions may be present. The epiglottis and
vocal folds can be visualized as well, giving a better
impression of upper airways' obstruction as a result of tis­
sue swelling from oedema or inflammation rather than an
examination after removal when swelling can diminish
rapidly as tissue fluid escapes.
Each injury is described and measured with reference to
a fixed bony point. Direct dictation of this part of the
examination while it is carried out is most accurate and
can easily be done while photography is under way. Care
should be taken that natural phenomena, such as Mongo­
lian blue spot (Fig. 8.7), haemangiomata and pigmented
naevi 38 are not inappropriately interpreted as injuries.
The genitalia are examined carefully. There is a wide
variation of appearance of the genitalia (see Table 2.1,
p. 30, and Table 2.2, p. 31), which should not be construed
as injury. Signs of abuse are listed in Table 2.4 (p. 39). Post­
mortem anal dilatation, a natural phenomenon ,39 should
not be confused with sexual assault (see Table 2.5, p. 40);
other perianal findings are also tabulated - see Table 2.3,
Figure 8.7 Mongolian blue spot above the gluteal fold in a baby
of Indian parentage.
p. 37. If assault is possible, swabs from the vagina and rectum
are obtained for DNA analysis of secretions and microscopy
for spermatozoa, foreign material and microbiology.
ESTIMATING BLOOD LOSS
Body cavities should be opened with care and any blood or
fluid measured. In infants, this is probably achieved most
easily using a large syringe without a needle.
The extent of blood loss should be related to the esti­
mated circulating blood volume, calculated with reference
to body weight (Table 8.3). Blood loss may be directly rele­
vant to cause of death, but its significance can be missed
unless calculated as a propoliion of normal blood volume.
The extent of intracranial blood loss is difficult to calcu­
late. There may be a haematoma, which can be measured by
weight or volume, but if it is attached to the dura or arach­
noid membranes it should not be disturbed for volumetric
measurement as histological assessment for injury to death
interval takes prioJity (see Table 8.4) A thin film of haemor­
rhage over one or both cerebral convexities is particularly
difficult to measure and one is left with only residual blood
in the cranial fossae after removal of the brain.
Blood loss into pericranial tissues and into muscles
attached to the skull can be considerable and is also diffi­
cult to measure, but an assessment should be attempted
and its contribution to overall blood loss stressed. See sec­
tion on 'Birth injury' later in this chapter.
DISSECTION (INFANTS AND
OLDER CHILDREN)
This section relates to the examination of infants beyond
the neonatal period and older children. Methodology
appropriate to the examination of the newly born is found
later in this chapter (p. 158). Protocols for the examination
 Dissection (infants and older children) I 153

Table 8.3 Blood volume/oge for infants and children

Age Volume (mL/kg)* Boys Girls

Average weight Average blood Average weight Average blood

(kg)+ volume (mL) (kg)''' volume (mL)

Newborn 80-85 3.6 288-306 3.4 272-289

1 month 80-85 4.5 364-383 4.2 336-357
6 month 75-80 8.2 615-656 7.5 563-600
12 month 75-80 10.2 765-816 9.4 705-752
2 years 75-80 12.5 938-1000 12.0 900-960
3 years 70-75 14.7 1029-1103 14.1 987-1058
4 years 70-75 16.5 1155-1238 16.2 1134-1215
5 years 70-75 18.6 1302-1395 18.2 1274-1365
10 years 70-75 31.5 2205-2363 32.2 2254-241 5
15 years 70-75 55.5 3885-4163 53.3 3731-3998
'Ref. 86.

+Boys' growth chart (birth-18 years) 1996/1 Child Growth Foundation.

'Girls' growth chart (birth- 18 years) 1996/1 Child Growth Founda tion.

Table 8.4 Histological appearance of subdural haematomata of different durations (after ref. 46)

Time after injury Haematoma Dural aspect Arachnoidal aspect

Histological features of subdural haematamata

To 24 hours Fresh red blood cells Fibrin Fibrin
24-48 hours Fresh red blood cells, Fibrin Fibrin
polymorphon uclear
leukocytes and fibrin
2-5 days Macrophages replace Fibroblasts present Fibrin
polymorphonuclear leukocytes at dural junction
4-5 days Mixture of intact and lysed Fibroblast layer 2-5 cells Fibrin
erythrocytes, siderophages layers thick
present
1 weeks Red blood cells lysed; early Fibroblast layer up to Single layer of fibroblast cells present
angiofibroblastic proliferation 12 cell la yers thick
2 weeks Haematoma liquefies, sinusoids, Fibrobla st layer is about half Fibroblast layer is a few cells thick, with an
'giant capillaries' appear of the thickness of dura occasional capillary
3 weeks Vascular sinusoids are well
developed
4 weeks Haematoma is liquefied Fibroblast layer is as thick as Fibrous membrane and a few capillaries see n
the dura, siderophages present
1-3 months Fibroblast layers are hyalinized and form membranes on both surfaces, large (giant) capillaries appear early,
secondary haemorrhages often develop
3-12 months The neomembranes fuse and contain mature fibrous tissue and scattered siderophages, after 3 months it is no
longer possible to accurately date (age) the haematoma
>1 year The neomembrane is a distinct fibrous connective tissue layer that closely resembles the adjacent dura mater,
occasional ca lcification and/or ossification

of sum cases are detailed by Valdes Oapena et al40 and
KroUS. 41 Comment will be directed paJiicularly at those
organs or regions for which a different approach or empha­
sis is appropriate in the older paediatric age group . The task
is easier when the following equipment is available: shOji
(10- to 12-cml, non-toothed forceps to mJlllmlZe tissue
damage; a range of tapered, round-ended scissors (Mayos);
and a variety of scalpels, rather than an autopsy knife.
Reference to the need for a digital balance was made in an
earlier section .
154 I Post-mortem examination in babies and children
Forma l neck dissection to detect injUly is carried out in a
sim il ar fas hion to that in adults and invo lves layer-by-Iayer
examination of the anterior neck muscles. In the infant, it
can be achieved following a vertical submental-symphysis
pubis incision as tissues stretc h readily but is probably eas­
ier with a shoulder-to-shoulder appro ach. Posterior neck
muscles are conveniently inspected by ensuring that th e
posterior scalp refection extends to the cervica l sp ines.
The thorax is best approached by serial division of costal
cartilages, avoiding damage to the osteochondral junctions,
some of which will be submitted for histological examina­
tion, a nd has the advantage of leaving the ribs intact. Fresh
rib fractures may be apparent following reOection of skin
and soft tissue (Fig. 8.8). Resuscitation Jib fractures are
un com mon, and are usually situated at the anterior ends of
lower ribs; accompanying haemorrhage is minor and there
is no vital reaction on histological examination. Rib frac­
tures are discussed further in Chapter 11, pp. 213- 215.
Relationships of orga ns are observed before disse ction
commences, some abnormal relati onsh ips may suggest a
syndrome diagnosis.
The thymus is a prominent organ in early life, weigh ing
10 ::!: 4 g at term. It is of normal size (often described as
large) in sudden death. A reduction in thymic weight below
th e normal ran ge for age is a non-specific indicator of ante­
mortem stress, probably in excess of 12 hours' duration 42
Petechial haemorrhages v isible through the thymic capsule,
particularly on its posterior aspect, a re present in about 80
per cent of SIDS 43 (see Chapter 11, p. 212). Larger, blotchy
haemorrhages may be present in asphyxial death.
The appearance of the lungs, noting the degree of expan ­
sion or collapse, is recorded. The lu ngs usua lly fill the chest in
SJDS but are often collapsed following attempted resuscitation
Figure 8.8 Fresh rib fractures accompanied by bruising, visible
after reflection of thoracic skin fl aps.
in hospital with a high concentration of oxygen. Fluid in the
pleural and pericardial cavities is observed, measured and
described. A sample of pericardium can be obtained at this
stage using sterile inst ruments and placed in tissue culture
Ouid for immed iate culture, if appropriate, or stored at
- 280°C for future availability for investigation of putative
genetic abnormality. This is followed by sampling of heart
blood, lung and myocardium for microbiological purposes.
External examination of the heart is conveniently canied
out at this point by fully opening the pericardial sac. Cardiac
abnormalities, both congenital and acqu ired, are common
causes of explained SUD! (see Chapter 11, pp. 208-210) and
older children (see Chapter 12, pp. 226-235). Cardiac situs is
dependent on atrial morphology. The comparative size an d
relationsh ip of the ascending aorta and pulmonary trunk are
noted - they should be equal in size. The course ofthe an te­
rior descending branch of the left coronalY artery should be
observed and the connection of t he vena cavae a nd pul­
monary veins noted. If all are normal then congenital heart
disease is excluded, with the exception of septal defects and
anomalous pu.lmonary venous return to the coronary sinus.
The liver is relatively large in infants and protrudes
below the costal margin. Colour and consistency are noted .
Fatty infiltration and evidence of trauma are sought. The
former requires frozen section and histochemica l staining
for confirmation. Preservation of samples to investigate the
possibility of GMD and full toxicological exami nation
should be considered.
The stomach is often distended, either because of attempted
'bag and mask' ventilation or microbial fermentat ion. The
mesentery is inspected for tears and contusio ns (Fig. 8.9).
Completeness of intestinal rotat ion is observed, along with
the presence of volvulus, intussusception and herniae - all of
wh ich are potential causes of intestinal obstruction, shock
and sudden death - are sought.
Figure 8.9 Contusion in the mesentery close to the
duodenojejunal junction.
 Dissection (infants and older children) I 155

Organs should be removed in blocks and not piecemeal.
In particular, the heart should remain attached to the lungs
until completely dissected. Th e Rokitansky technique is
suitab le but the following proced ure is easier: removal of
(he intestines from the duodenojejunal fl ex ure to the rec­
tum (having checked the mesentery for contusions) (Fig.
8.9), then rem oval of the thoracic and upper abdominal
viscera en bloc thus opening the inferior vena cava and
abdominal aorta, followed by removal of genitourinary
tract. When sexual assault is suspected, a cuff of perineal
skin is removed in continuity with the lower genital tract
and anus, usually including the coccyx, the pubic area and
the med ial aspect of the thighs. Swabs for forensic exam ­
ination should be taken before dissection is begun .
Following removal of the viscera, the thoracic and
abdominal cavities should be cleared of blood and fluid
and their walls examined. The ribs should be carefully
inspected for fractures. Rib fractures are often undisplaced
in babies and in fants. Attention will be drawn to fresh frac­
tures by subperiosteal an d intramuscular haemorrhage.
Older fracture sites will be supported by callus and are usu­
ally stable (Fig. 8. 10), although re-fracture can occur. Cal­
lus is particularly marked on the internal aspect of the rib
and is readily palp able. Fractures at the posterior ends of
ribs are easily missed and should be specifically sought,
with the remova l of pali of the rib cage fo r radiographic
and histological examin ation wh en there is suspicion of
injury. Previous avulsion of the osteochondral junction of
the lower ribs (6-10) is common when there are fractures
elsewhere and is seen as expansion of the internal aspect of
the bone adj acent to the cartilage (see Fig. 8.20, p. 160).
Stripping of the pleura to look for rib fractures in
infants is not recommended. The stripping process results
in tissue dam age, including any subperiosteal reaction and
may render histological evaluation probl ematic.
Th e upper airway is examined carefully and may be
fixed before opening when trauma to the neck has
occurred. Laryngeal fracture is unlikely but small haemor­
rhages may be present in the intrinsic lary ngeal muscles.
FUliher microbiological samples can be taken from the
lower trachea/main bronchi.
A convenient meth od of opening the heart is shown in
Figure 8.11. It is opened before separation from the lungs.
Should a cardiomyo pathy be suspected, a horizontal slice
through both ventricles, midway between base and apex,
demonstrates it well (Fig. 8.12) without interfering with
dissection; chamber connections and morphology, appea r­
ance of valves, septal defects and the origin and course of
coronary arteries are examined. The appearance and thick­
ness of ve ntricular myoca rdium is noted. Any subendocar­
dial fibroelastic thickening or haemorrhage is noted. Heart
and lungs are separated and weighed. Pulmonary arteries
are examined for evidence of thromboembolus and
hypeliension - both causes of sudden death in early life.
The former is most likely to be found in the poorly mobile,
postoperatively and follo wing the use of intravenous devices.

... ...
...
\
\
\
\
\
\4
\
\. \
- ...
"­ ...
\
\
\
... ... \
... ... \
\
\
"­
"-,,- 2 ... ... \

"­ .... \
.... ...
....... ...
......... ...

Figure 8.10 Thorax after evisceration; old posterior rib fractures
with ca llus are readily visible.
"
Figure 8.11 Opening the heart, cutting lines are numbered
sequentially (from Fetal and Neonatal Pathology, 4th edn, 2007,
p. 39, figure 2.20, with kind permission of Springer Science and
Business Media).
1 56 I Post-mortem examination in babies and children
Figure 8.12 Transverse slice through the ventricular
myocardium half way between base and apex. There is marked left
ventricular hypertrophy in this ll-month-old infant with
hypertrophic cardiomyopathy.
Subpleural petechiae are a common post-mortem finding
in infants. Larger, blotchy or confluent haemorrhage is less
common and may indicate an asphyxial episode (see Chapter
11, p. 218). Lung scarring may be present in babies who were
born preterm and who have received intensive care.
The stomach is opened and its contents noted. A sample
may be retained for toxicological examination. Foreign
material, such as paper or non-food items, should evoke a
high suspicion of non-natural death. The nature of food
residues should be related to the description of ante­
mortem events. The amount of food in the stomach is a
most unreliable guide to length of survival after the last
meal. One should remember that gastric aspiration may
have been carried out in the accident and emergency
department. When both stomach and intestines are found
to be empty, careful note of the history and nutritional
state is required. Microbiological samples from the intes­
tine are mandatory in these circumstances and detailed
questioning of carers may be appropriate.
Haemorrhagic infarction of the intestines, myocardium
and sometimes brain is commonly seen in infants who have
been subjected to intensive care procedures for periods of
> 12 hours prior to death. Care should be taken not to over­
interpret these changes (see Chapter 11, p. 215). Pulmonary
haemorrhage is common after attempted resuscitation
involving ventilation but siderophages, seen on microscopic
examination, require at least 2 days to develop.44
EXAMINATION OF THE BRAIN, SPINAL
CORD AND EYE
Before the cranial cavity is opened, the scalp and calvarium
are examined for evidence of injury; if injuries are found
then these are recorded as described previously. In the
Figure 8.13 Cranium opened by the modified Beneke technique.
Bone flaps are reflected following incision of suture lines. There is a
brown membrane attached to the parietal dura, due to organization
of a subdural haemorrhage that occurred several weeks before death.
newborn and in early infancy, the calvarium is opened
using the modified Beneke technique. 45 It allows visualiza­
tion of bridging veins before the bony flaps are completely
reflected and of ' membranous' adhesions that may be pre­
sent between dural and arachnoid membranes during organ­
ization of old subdural haematoma (Fig. 8.13). Any tearing
of bridging areas is recorded and a search made for a related
focal subarachnoid haemorrhage (Fig. 8.14) and thrombosis
of torn veins (Fig. 8.15).
The presence of extradural and subdural haematoma is
noted. The former are unusual in early life because of the
firm adherence of the dural membranes to the inner table
of the calvarium, particularly along the suture lines, a phe­
nomenon that may introduce difficulty in removal of the
calvarium following a circumferential saw-cut.
The infant brain swells rapidly following hypoxia or
trauma, largely obliterating the subdural space so that sub­
dural haemorrhage may be reduced to a confluent film
over the convexities, running in the falcine fissure and
with small collections in the cranial fossae.
Any abnormality of the convexities is recorded, the
brain is removed, weighed and fixed in 10 per cent buffered
formalin before examination. It is not advisable to attempt
photography of the brain prior to fixation, other than pic­
tures taken before removal of the brain from the skull base.
The infant brain is particularly soft following oedema and

Figure 8.14 Superior convexity surface; focal subarachnoid
haemorrhage may indicate the site of torn bridging veins.
Fi gure 8.15 Surface of the brain after fixation from a baby with
oJbdural haematoma and haemorrhage. Thrombosis of a torn
ridging vein is present.
. :"Poxic insult. Artefactual tearing of vulnerable structures,
ch as the corpus callosum, can result from manipulation
of the unsupported brai n (Fig. 8.16). When the brain is very
- ft, weighing in the fresh state can be omitted.
Following the removaJ of the brain, standard samples for
rristological examination are obtained from the parietal dura
w d the dural folds. Any dura with attached clot, organizing
~em brane or brown staining should be removed, additionally,
- r histological examination. Hardman 46 describes the histo­
.ogical appearances of subdural haematoma with approximate
_uration prior to death (Table 8.4, p. 153).
Examination of the brain, spina l cord and eye I 157
Figure 8.16 Fresh infant brain with flattening and posterior
separation of the hemispheres. There is stretching and tearing of
the corpus callosum.
Figure 8.17 Slice of the cerebral hemispheres after fixation;
death followed a motor vehicle collision. There is haemorrhage in
the right putamen/internal capsule and focal haemorrhages in the
corpus callosum.
In babies, infants and young children, the spinal cord is
most conveniently removed from the front. An intervertebral
disc is incised in the lower lumbar region and the proximate
vertebral body grasped and elevated; the pedicles can then be
divided with bone forceps on either side up to the upper cer­
vical region. The cord is examined through its intact dural
covering and any extraduraJ haemorrhage is noted. The nerve
roots are incised on either side as far laterally as possible to
preserve dorsal root ganglia and the cord is removed after
incising the dura at Cllevel. The cord is fixed flat for optimal
examination. This is easily achieved by carefully laying it on
a strip of card and permitting adherence for 2 or 3 minutes
before immersion in fixative.
Examination of the fixed brain and cord (preferably by
a neuropathologist) is carried out systematically. The brain
is photographed prior to slicing and any abnormalities
identified subsequently are also photographed (Fig. 8.17).
Samples for microscopy should follow appropriate fetal
and infant blocking schedule (Fig. 8.18). additionally, sam­
ples of any abnormalities are obtained.
158 I Post-mortem examination in babies and children
Rand L frontal parasagittal
Rand L parietal parasagittal
Rand L parietal convexity
Rand L basal ganglia at
level of mamillary bodies
Rand L thalamus
Rand L hippocampus at
level of lateral geniculate
body
Rand L occipital
Midbrain
Pons
Medulla
Rand L cerebellar
hemispheres, inCluding
dentate nucleus vermis
Figure 8.18 Diagram for sampling of the infant brain for
histological examination. Any lesions are sampled in addition
(from Fetal and Neonatal Pathology, 4th edn, 2007, p. 43, figure
2.24, with kind permission of Springer Science and Business
Media, courtesy of Professor JE Bell, Edinburgh).
The spinal dura is incised anteriorly and posteriorly, and
the cord surfaces photographed. The presence of blood or
clot is noted.
The cord is sectioned transversely from the upper cervi­
cal region. This can be done while it remains attached to the
dura. Any lesions within the cord are photographed. Sam­
ples for microscopic examination are taken at identified
levels, and the cord and dura are preserved in continuity so
that other identifiable blocks can be obtained if required.
The eye can be removed from the front (anterior
approach) following incision of the conjunctival reflections
and division of the intrinsic ocular muscles in turn. The eye
is prolapsed and the optic nerve divided. In the young, it is
easier to approach the eye from the floor of the anterior
cranial fossa (posterior approach), which is thin. This
affords a view of the total orbital contents so that any
extraglobal lesion can be sampled. 47 It is important that
vitreous sampling is omitted when detailed examination is
considered appropriate.
The eyes are fixed separately in labelled containers. Exam­
ination of the fixed eye is described by Lee.48 Photographs of
any external lesions should be made together with a photo­
graphic record of the interior after opening before sampling
for microscopy. Processing of the eye requires care to min­
imize artefactual distortion. Perls' Prussian blue reaction
(PBR)-stained sections are mandatory when assault is likely
and immunohistochemical examination against beta amyloid
precursor protein of the optic nerve is more effective than
haematoxylin and eosin staining for the demonstration of
axonal swelling when optic nerve trauma is suspected. 49 In
some of these cases, referral of the whole eye for specialist
examination may be the safest option.
EXAMINATION OF THE NEWLY BORN
Detailed techniques of necropsy examination appropriate
for the fetus and neonate are described by Wiggleswolth 50
and Keeling.4 5 The following observations are relevant
to the examination of the newly born for medicolegal
purposes. This may be required because the body was
concealed, abandoned , born in the absence of appropri­
ately qualified attendants or was unexpectedly still born.
Other situations when a medicolegal examination is
required are when attending clinicians cannot complete a
death certificate or if there are complaints by parents or
others and following maternal trauma or death (see Chap­
ter 10). Important considerations during the examination
of perinatal deaths are those of live birth and separate exis­
tence. These are discussed in Chapter 10, pp. 187-192.
Radiography
A single whole-body radiograph is not sufficient in susp i­
cious fetal or neonatal deaths; imaging should follow the
procedures outlined on p. 147 earlier in this chapter. As
well as providing a record of any fractures, radiographs in
the perinatal period provide useful information about ges­
tation (and, therefore, viability) ;51. 52 they also provide good
proof of skeletal malformations, which can contribute to a
syndrome diagnosis. A radiograph may demonstrate gas in
the stomach and intestines in live birth.
External Findings
Weights and measurements mLlst be carefully recorded.
A careful external examination is most important. Any
changes of maceration are carefully recorded as they
provide useful information about the fetal death to delivery
interval. Serial maceration changes are tabulated (Table 8.5);

Table 8.5 External changes of maceration by death to delivery
interval (after ref. BB)
Observed feature Death to delivery
interval
No maceration < 6 hours
No discolouration of cord insertion < 6 hours
Desquamation ?! 1cm ?! 6hours
Cord discolouration, brown/red ?! 6 hours
Desquamation face/back/abdomen ?! 12 hours
Desquamation ?! 5 per cent of body surface ?! 18 hours
Desquamation in two or more zones' ?! 18 hours
Skin brown/tan in colour ?! 24 hours
Moderate/severe desquamation ?! 24 hours
Mummification (dehydration :t fetal ?! 2 weeks
compression)
'Sody zones: scalp, face, neck, chest, abdomen. back. arms, hand, leg, foot,
scrotum.
p. 188) is a useful observation. Some moulding of the cranium
Ch anges of maceration are accelerated by fetal hydrops and retarded in
is nOlmal but excessive moulding of the cranium is an indica­
the presence of fetal growth restriction.
tion of long labour. In vertex presentation there is often an
early changes of maceration are ill ustrated in Figs 10.4- 10.6
p. 184. The presence of excessive meconium staining
indicates hypoxic fetal stress in the mature baby. If there is
much meconium, blood or vernix caseosa on the skin it is
advisable to remove it after initial photography so that the
skin can be examined in more detail.
Cutaneous pallor may be evident following significant
fetomaterna l haemorrhage. 53
The presence and nature of any cutaneous injury is
recorded and photographed.
It is not unusual to find cutaneous petechial haemorrhages
over the presenting part in fresh still birth, particularly when
there is placental abruption. They are commonly found over
the face, head and neck and, occasionally, on the chest wall
in a mature fetus when vertex presentation is usual. They are
often particularly prominent following placental abruption
(see Fig. 10.10, p. 188) and should not be interpreted as evi­
dence of stra ngu lation or deliberate airways obstruction
without corroborative evidence.
Cutaneous petechiae, and sometimes more extensive
bru ising, are seen on the legs following breech delivery or
the arm if one has prolapsed through a partly dilated celvix.
These findings are more usual in the premature fetus, for
which breech presentation is more common. 54 At 32 weeks'
gestation approximately 16 per cent of babies present by the
breech, compared with around 5 per cent at term.
The presence and nature of any dysmOlvhic features are
noted. These may contlibute to a syndrome diagnosis, directly
related to death. The presence oftvvo or more dysmorphic fea ­
tures is an indication for chromosome examination.
The umbilical cord and its insertion are examined and
any discolouration or dehiscence is noted. The free end of
the cord may indicate the method of division. Differences
in appearance of the cord ends are described 55 - traction,
Examination of the newly born I 159
picking apart, and cutting with scissors or bl un t/sharp
knives - and assessment is aided by dissecting microscopic
examination. It may correspond with the free end of the
cord attached to a placenta discovered elsewhere. The pres­
ence or absence of a clip or tie is noted and described.
Examination of The Scalp and Cranium
EXTERNAL
Examination of the head and cranial contents is undertaken
systematically. The head circumference should be approxi­
mately equal to the crown-rump length in the second half of
pregnancy. A larger measurement raises the possibility of
hydrocephaly, whereas a smaller one may result from crania l
distortion or, in its absence, chronic in trauterine brain pathol­
ogy. An increase in the occipitomenta l diameter (see Fig. 10.9,
area of localized oedema, usually with marked congestion and
sometimes with frank haemorrhage, over the presenting part.
INTERNAL
The scalp is incised from behind the ear over the posterior
fontanelle to the opposite ear and reflected forwards and
backwards, sufficiently far to completely expose the bones
of the cranial vau lt. The presence of haemorrhage within
the scalp is noted and a check is made for any corres­
ponding external injury. Marked congestion of the scalp is
a very frequent finding in antepartum still births after
vertex presentation. It should not be confused with tra uma.
Histological examination shows no vital reaction. The con­
tour of the occipital bone is examined. Occipital osteodias­
tasis , in which the inferior pali of the OCCipital bone is
displaced inwards resulting in damage to the brain or
venous sinuses, may occur after breech delivery or, occa­
siona lly, following forceps delivery;56 it is eas il y missed if
the sca lp is not ful ly reflected posteriorly.
The posterior neck muscles are conven iently ex amined
at this point and any contusion is photographed and sam­
pled for histological examination.
The atlanto-occipital membrane sho uld then be incised
under direct vision. Gentle pressure on the cranium wiiJ result
in a flow of CSF, which is likely to be heavily blood stained
when there is selious intracranial injury. The fontanelles and
suture lines are obselved. Congestion or haemon'hage is
sometimes seen at the margins of sku ll bones and probably
indicates the stress of forcible displacement.
Cranial Cavity
The cranial cavity is opened using the modifi ed Beneke
technique (Fig. 8.19), which allows examination of the
160 I Post-mortem examination in babies and children
Figure 8.19 Neonatal cranium opened using the Beneke
technique; there is confluent subarachnoid haemorrhage, of
hypoxic origin, over the right parietal region; this is not
accompanied by significant subdural bleeding.
superior surface of the brain while preserving the falx and
tentorium for subsequent examination.
Extradural haemorrhage is uncommon amongst peri-natal
deaths because of strong adherence of the dural membrane to
skull bones. Localized, non-space-occupying haemorrhage
can accompany skull fracture. The presence and position of
haemorrhage in the subdural and subarachnoid space is
noted. The form er is the result of trauma; the latter may be
traumatic when observed in the vicinity of torn bridging
veins or when, as a result of hypoxia, the presence of multi­
ple petechial haemorrhages is likely, although confluent sub­
arachnoid haemorrhage of hypoxic oligin is recognized in
the perinatal period, often over the temporal poles.
Observation of the cortical gyral pattern provides useful
information about fetal maturity. The gestation-related
appearance is remarkably constant and is illustrated by
Dorovini-Zis and Dolman 57 and Feess-Higgins and Lar­
roche 58 (see ref. 45).
The falx and tentorium should be examined with care.
This is best done by tipping the head forward then gently
eleva ti ng each occipital pole in turn so that the posterior part
of the falx and the tentorium can be viewed. The usual site
of significant injury is from the free edge of the tentorium
running into straight sinus (Fig. 8.20). Sheari ng tears in the
free edge of the tentorium , not involving venous sinuses, are
an indication of rapid moulding of the head and are not
infrequent following spontaneous delivery, particularly in
still birth. They do not, in themselves, contribute to death.
Tears in the falx are less usual and it is imp ortant to be aware
that lacunae, smooth-edged holes, can occur naturally
within the falx during development.
Figure 8.20 Birth injury. Tear at the junction of the falx and
tentorium follo wing instrumental delivery, displayed by removal of
the cerebral hemisphere on the affected side.
In view of the incompleteness of myelination of the
fetal brain, fixation of the whole brain before examination
is essential otherwise important information may be lost.
Spinal Cord
When intrapartum death occurs following in stru mental
delivery and no intracranial patho logy is app arent, the cer­
vica l spinal cord should be removed within its bony cover­
ing using Yates' method. 59
Body Cavities
Examination of thorax and abdomen follows a method
appropriate to this age group.45.50 Petechial haemorrhages are
numerous in the thoracic viscera follo wing acute hypoxic
stress, and may be particularly prominent around the ductus
arteriosus. They are particularly promin ent and numerous
after placental abruption. Potentially fatal congenital anom­
alies seen in externally normal infants are desCIibed in
Chapter 10, p. 193 . In the macrosomic fetus (>4000g) it is
important to look for evidence of birth injury.
Placenta
EvelY effort should be made to examine the placenta. It may
contribute important information to case evalu atio n in most
circumstances when there is a potential legal interest in a
perinatal death .6o Examination of the placenta is detailed by
Wigglesworth 5o and Keeling 45 . Cord length, weight and
measurements should be recorded as they contribute to the
overall assessment of the case. Gestation-related placenta
weights for twins and singletons are given by Pinar et al 61
and cord lengths by Naeye 62 (see Appendix BJ. Histological
examination of the placenta may contribute to the assess­
ment of the death to delivelY interval in stillbirth. 63
Help with interpretation of placental abnormality can be
found in Benirshke et al,64 Fox and Sebire 65 and Khong 66 .
 Examination of the newly born I 161

.-\ltshuler67 and Macpherson 68 discuss the relevance and

limitations of placental data and pathological abnormalities
in relation to alleged obstetric malpractice.

Recognized Birth Injuries

Birth injury is more common fo llowing instrumental, and

sometimes operative, delivery. However, there are occasional
case reports of birth inju ries following spontaneous delivery.
Birth injuries are considered in more detail by Wig­
glesworth 69 and Keeling.70
After delivery in hospital it is important to look for
signs of birth trauma such as a chinon following Ventouse Figure 8 .21 Birth injury. Subgaleal haemorrhage, massive blood
delivery, bruising and abrasions in the shape of a forceps loss into soft tissues. Instrumental delivery.
bl ade, and perhaps a linear skin incision following Cae­
sarean section. Scalp injury is a recognized complication of
vacuum extraction. 7] Bruising of the legs is common fol­
lowing breech delivery, especially in preterms. In term
breech delivery, it is important to look for predisposing fac­
tors, such as neuromuscular pathology or renal abnormal­
ity (because of oligohydramnios), which might be the
underlying cause of death.
Subaponeurotic (subgaleal) haemorrhage is an infre­
quent but serious complication of forceps or Ventouse
delivery, or a combination of both methods (Fig. 8.21). 72
A large volume of blood can be lost into the su baponeu­
rotic space. Robinson and Rossiter 7J calculated blood losses
exceeding 100 mL using a formula based on cranial diam­
eter and scalp thickness.
Subperiosteal haemorrhage, particularly over the par­
ietal bones, is common (Fig. 8.22) and of little consequence.
It may be found in live borns following spontaneous, easy Figure 8 .22 Birth injury. Subperiosteal haemorrhage following
vaginal delivery with vertex presentation, as well as fol­ spontaneous delivery; blood loss is trivial.
lowin g instrum ental delivery. The amount of bleeding is
limited by periosteal attachment at the margins of the bone the brain is unlikely to be traumatic and a local cause should
and by the limited distensibility of the membrane. be carefully sought.
Although it may be found following difficult delivery, with Fra ctures of the skull vault should be sought with care;
other pericranial lesions, as an isolated finding, it makes th ey are not always apparent on radiographs. A single
little contribution to the overall assessment of the case. fracture in the parietal bone from the sagittal suture line
Subdural74 and cerebral haemorrhage7 5 and very occa­ running down towa rds the ear is seen occasionally follow­
sionally, extradural haemorrhage 76 or posterior fossa haem­ ing forceps delivery or vacuum extraction. An undisplaced
orrhage 77 can complicate vacuum extraction as well as Iinear fracture is not of great significance in respect of
forceps delivery. Subdural haemorrhage is almost always cause of death, although it is an indication of difficult
due to trauma. When present over the cerebral convexities it delivery. Of more significance is an accompanying sub­
is the result of tearing of bridging veins; trauma here aponeurotic or intracranial haemorrhage, or cerebral
includes excessive moulding of the skull from obstructed hypoxic injury.
labour or rapid moulding of the cranium in precipitate deliv­ Clavicular fracture is the commonest birth injury. There
ery. Minor, focal subdural haemorrhage is not uncommon is usually a history of difficulty in delivering the shoulders
following normal vaginal delivery.78 Evidence of these in a large baby. Bruisin g of the neck muscles (Fig. 8.23) and
minor haemorrhages in the form of focal brown staining of fracture of the humerus can occur in similar circumstances.
the dura is not unusual in sudden infant death . Any sugges­ It is essential to correlate any injuries found with events in
tion that major subdural haematoma is th e result of unas­ labour. Babies in the extended breech position may suffer
sisted delivery should be addressed very critically.79 Tears of fracture of a femur (or occasionally both) when the legs are
the dural folds that extend into cerebral sinuses are rare in brought down. Rib fractures are exceedingly rare birth
non-instrum ental delivery but cause catastrophic fatal injuries. Careful attention should be given to delivery
haemorrhage. Subdural haemorrhage solely at the base of events, birth weight and events in the neonatal period
162 I Post-mortem examination in babies and children

Table 8.6 Samples for histological examination in perinatal and

infant forensic deaths20,45

la rynx
Trachea
Lung - five lobes, PBR reaction on all
Heart - right and left free ventri cu lar wall, interventricular
septum
Kidney - two
Liver - right and left lobes
Small and large intestine
Pituitary
Thyroid
Pan crea s
Adrenal - two
Figure 8.23 Birth injury. Bruising of anterior neck mu sc les. Gonad
Instrumental delivery. Diaphragm
Costochondra l junction
befo re accepting birth injury as a cause for rib fract ures in Any macroscopic lesions
an infant. Rib fractures are discussed further in Chapter 11. Dural fold
The commonest visceral injury is to the liver, when rup­ Dura - parietal
ture of a subcaps ular haematoma can result in catastrophic
In SUOI (including sudden neonatal death)
haemo rrhage. The baby is usually mac rosomic. A simil ar
Frozen section for oil red 0 staining from:
injulY is seen in preterm infants, particularly following
He art
breech delivelY, as the liver is poorly protected by the rib
Liver
cage at this time. Splenic rupture is a rare birth injury.
Ki dney
Adrenal
HISTOLOGICAL SAMPLES ~L u ng

Thorough sampling for histological examination is an
essential part of perinatal, infant an d childhood post­
mortem examination. In ea rly life, much diagnostic infor­
mation is derived from microscopi c examination when
macroscopi c ap pearan ces may be normal, non-specific or
not understood by the inexperienced. Sampling should fol­
Iow a standard protocol (Table 8.6). It is the in vestigation
most likely to make a positive contribution in SUDl. 2o
Norma l his to logical appearances of organs in the infant
period are illustrated in Valdes-Dapena et al. 40
In addition to convent ional sectio ns of fo rmalin-fixed
para ffin-emb edded samples, it is important to reserve sm all
fresh samples of heart, liver, kidney and adrenal gland for
frozen sections in order to look for lipid infiltration ; when
small in amount, lipid infiltration is probably st ress-related,
but massive lipid infiltration is su ggestive of genetic meta­
bolic disease (see Chapters 11 and 12). These sa mples can be
kept frozen, perhaps wit h a sample of spleen, for molecular
studies should the need arise. If inhalation is suspected, a
froze n sample of lung for lipid staining is also useful.
In SUDI, it is advisable to examine lung sections stained
fo r iron (PBR) to assess the possibility of previous pulmonalY
haemorrhage. It may be useful in other circum sta nces, Wh en
death occurs suddenly in the neonata l period, staining for
reticulin is very usefu l in the evalu-ation of microscopic
lung anatomy: it permits exclusion of congenital alveolar
dysplasia and better assessment of the appropriateness of
In perinatal death
Pla centa X 2
Umbilica l cord
Extra placental membranes
lung maturity. An elastic stain, (Elastic van Gieson is my
preferred technique) allows assessment of postnatal
adaptation of pulmon ary vasculature, ventilation-induced
damage and other residua of neonata l intensive care.
Exa mination of the ca rdiac conduction system is desir­
ab le following instantaneo us death or observed rapid col­
lapse that is unexplained after conventional investigation.
It should be considered when sudd en death follows cardiac
surgelY and when a second sudden death occurs within the
family. In these circumstances the heart should be retained
complete with this in mind. Conventional sampling is
desc ribed and illustrated by Davies et al. BO Michaud et alBI
and Ashworth B2 describe a simplified method for analysis
of the conduction system.
It is imp ortant no t to over interpret minor differences
in microanatomy, either in distributio n of conduction tis­
sue B3 or of the process of so -called 'resorptive degeneration'
of the atrioventricular nodal tissue and the His bundl e. 83 ,84
The so-call ed marke rs of SIDS, pulmonary haemosiderin
and the significance of minor degrees of infla mmation in
SUD! are discussed in Chapter 11, pp, 209-212, 215-2 16
an d p. 210, respectively.

Figure 8.24 Posterior end of rib trimmed for histological
examination after fi xa tion and decalcification. A fracture with
callus is present.
Demonstration of Bony Injury
Histological exa mination of bones following radiographic
:dentification of fracture or unidentified localized abnor­
mality is essential. The bone(s) in question can be excised
completely or a segment excised with a clear margin on all
aspects of the area to be investigated.
It is useful to re-X-ray the excised bone, either before or
after fixation. These films are useful when sampling for
microscopy. Sampling should be done following decalcifi­
cation, otherwise further damage to the bone is likely and
histological assessment rendered more difficult. Blocks for
microscopic examination are taken at right angles to the
fracture line (Fig. 8.24) or along the long axis of the lesion
(Fig. 13.25). Blocks should include cortical bone, periosteum
and adjacent soft tissues. Histological examinatio n of frac­
ture contributes to assess ment of timing of injury, particu­
larly those that have occurred within 1 week of death,
when radiological abnormality is minimal. The significance
of fractures in perinatal and infant deaths are discussed
further in Chapter 3 (pp. 51-581, Chapter 10 (p. J92) and
Chapter 11 (p. 218).
RETENTION OF ORGANS
In the course of the medicolegal post-mortem examination
in babies and children, careful thought should be given to
the need to retain organs. Many coroners and procurators
fiscal are unwilling to sanction organ retention, particu­
larly the brain. It is importa nt that pathologists ensure that
legal personnel are appropriately informed of the benefits
of so doing in any particular case. It is difficult, and in
some cases impossible, to examine the brain of a still born
or young infant in the fresh state. The information
obtained from an appropriately thorough examin ation of
Exchange of information and multidisciplinary review I 163
Figure 8.25 Whole section of the anterior end of a lower rib.
There is expansion of the interna l aspect of the rib and irregularity
of the osteochondral junction as a result of prior trauma.
the central nervous system is of utmost importan ce, in the
context of both assessing injury when inflicted injury is
suspected and in respect of cause of death in SUDI and
pelinatal deaths. Local expelience of the contribution of
formal neuropathological examination in deaths investi­
gated by a procurator fiscal is greater than that cited by the
CESDI SUDI studies. 8s
Brief (overnight) fixation of other organs, particul arly
the heart, makes sampling for microscopic examination
easie r and more accurate.
In cases of inflicted injury, organs may require fixation
so that they are in an appropriate condition for examina­
tion by another pathologist. In these cases, if there is doubt
concerning the identity of lesions then the organs or tissues
should be retained.
EXCHANGE OF INFORMATION AND
MULTIDISCIPLINARY REVIEW
In complex cases, it is important that there is regular
exchange of information between all the professionals
involved so that appropriate interpretation of findings
ensues. Th e pathologist should ensure that the results of any
ancillary investigations are available and that conclusions
are based on the appropriate interpretation of all findings.
Multidisciplinary reviews are conducted on all sudden and
unexpected death in a number of areas within the UK, con­
vened either through hospital paediatric units, community
paediatricians or via child protection agencies. Following the
publication of an intercollegiate report, such reviews are likely
to become standard practice throughout the UK. 86 The empha­
sis of such reviews is likely to vary depending on the lead
agency and the composition of the review gro ups. Although
these activities are time-consuming, they are productive both
164 I Post-mortem examination in babies and children
in respect of child protection a nd in our understanding of
causes and contributing factors in sum, putting in place
app ropriate support for famili es and in the appropriate plan­
ning of selvices for chiJdren.o 1
REFERENCES
elective preterm delivery on birth- weight and head
Vujanic GM Cartlidge PHT, Stewart JH. Improving the quality
of pe rinatal and in fant necropsy examinations: a follo w up
study.] Clin Patho/1998; 51:850-3.
2 Sheehan KM, McDonnell M, Doyle EM et a!. The qua li ty and
val ue of sud den in fan t death necropsy reporting in Irelan d.
] Ciin Path 01 2003; 56:753 - 7.
3 Bass M, Kravath RE, Glass L. Death-scene investiga tion in
sudden infant death . N Engl] Med 198 6; 315:100-5.
4 St urner WQ. Commo n errors in forensic pediatric pathology.
Am] Forel1 sic Med Pat110/1998; 19:317 -20.
5 Bass M. The falla cy of the sim ultaneous sudden infant
maturation: with stand ards for body and organ development.
dea th syndrome in rwi ns. Am] Forensic Med Pathol 1989;
In Wigglesworth JS , Singe r D (eds) Textbook of Fetal and
10:200-5.
Perinatal Pathology, 2nd edn. Ma lden, MA: Blackwell Sc ience,
6 Sa ukko P, Knight B. The pathophysiology of death. In
Kn ight's Forel1s ic Pathology, 4th edn. London: Arnold , 2004,
pp. 52-97.
7 Ca rpenter RF. The prevalence and distribution 0 f bruising in
babies. Arch Dis Cilild 1999 ; 80:363 - 6.
8 Sugar NF, Taylor JA, Feldman KIN et a1. Bruises in infa nts and
toddlers. A rch Pediatr Adoles c Med 1999; 153:3 99- 403 .
9 Moon RY, Patel KM, Schaefer SJ. Sudden Infant Death
Syndrome in child care settings. Pediatrics 2000; 106:295-300.
10 Maxeiner H. Demo nstrati on and interpreta tion of bridg ing
vein rup tures in cases of infantile subdura l bleedings.
] Foren sic Sci 2001; 46:85-93.
11 Erlich E, Maxeiner H, Lan ge J. Postm ortem radiological
in vestigation of torn bridging vein ruptures. Legal Med
(Tokyo) 2003; 5lSupp1. 1) : 225-7 .
12 Sadler DW. The value of a thorough pro tocol in the in vestigation
of sudden infant dea ths.] Ciin Pathol 1998; 51 :589-94.
13 Wigglesworth JS , Keelin g JW, Rushton D, Berty PJ. Tech nica l
method. Pathological inves tigations in cases of sudd en infan t
death.] Ciin Pat110/198 7 ; 40:1481-3.
14 Dettm eyer R, Baas ner A, Schlamann M et a!. Coxsa ckie BJ
myocarditis in 4 cases of suspected sudden infant death
syndrome: diagnos is by immunohistochemical and molecular­
pa thologic invest igation. Patlwl Res PraU 2002; 198 :689-96.
15 Bajanowski T, Rolf B, Jorch G et al. Detection of RNA viruses in
sudden infant death (SID), Int] Legal Med 2003; 117 :237- 40.
16 Carmichael EM, Gold wate r PN, Byard RW. Routin e
microbiological testing in sudden and un expected infant
death. ] Pediatr Child Health 1996; 32:412-15.
17 Mitchell E, Krous HF, Donald T, Byard RW. An ana lysis of the
usefuln ess of spec ific stages in the pathol ogic investigation of
sudd en infant death . Am ] Forensic Med Pat flol 2000;
21 :395-400.
18 Sims MA, Collins KA. Fetal death: a 10-year retrospective
study. Am] Forensic Med Pathol 2001; 22:251 - 5.
19 British Med ica l Associatio n. British Natiol1al Formulary,
Paediatric. Londo n: Roya l Pharm aceutical Society of Great
Britain, 2005, p. 225.
20 Berty J, Allibone E, McKeever P et al. The pathol ogy study:
The contribution of ancillary pathology tests to the
investiga tion of unexpected infant death . In Fleming P, Bacon C,
Blair P et al (eds) Sudden Unexpected Deaths in Infancy the
CESDI SUDI Stu dies 1993-1996. London: The Statio nery
Office, 2000, pp. 97-11 2.
21 Ackerman MJ, Tester DJ , DriscoU DJ. Molecular autopsy of
sudd en unexplained death in the you ng. Am] Forensic Med
Patho1 2001; 22:105-11.
22 Wilc ox RL, Nels on CC, Stenzel P, Steiner RD. Postmol1em
screening for fatty ac id oxidati on di sord ers by analysis of
Guthrie ca rds with tand em mass spec trom etry in sudden
unexpected death in infancy.] Pediatr 2002; 141 :833-6.
23 Yudkin PL, Aboua lfa M, Ey re JA et al. The influence of
ci rcu mference sta ndard s. A rch Dis Child 198 7; 62: 24-9.
24 Chambers HM, Knowles S, Staples A et al. Anthropo metric
measurements in the seco nd tri mester fetu s. Early Human Dev
1993; 33:45-59.
25 Maroun LL, Graem N. Autopsy standards of body pa rameters
and orga n weights in non-macerated an d macerated human
fetuses. Pedia tr Deuel Patho/2005; 8:204-17.
26 Freeman N , Co le TJ, Chinn S, Jones PR M, White EM, Preece
MA. 1995 Cross-s ecti ona l stature and weight refere nce CUlves
for the UK, 1990. Arch Dis Child 73:17- 24.
27 Singer D, Sung CoR, Wiggleswonh JS. Feta l growth and
1998, pp. 8-40.
28 Hansen K, Sung CJ , Huang C et al. Referen ce values for
second trimes ter fetal and neonata l organ weights a nd
meas urements. Pediatr Develop PatI101 2003; 5:160-7.
29 Thompson WS, Cohl e SD . Fifteen-year retrospective stud y of
infant organ weights a nd revisio n of sta ndard weight tabl es.
] Forensic Sci 2004; 49:1-11 .
30 Shan kle WR, Landing BH, Gregg J. Normal organ weights of
infa nts and children: Graphs of va lue by age, with confide nce
intervals. Pediatric Patho l 1983; 1:399-408.
31 Haddad S, Res tieri C, Krishn an K. Ch aracte rization of age­
related cha nges in body weight and orga n weights from birth
to ado lescence in humans. ] Toxic ol Enviroll Health 2001 ;
54:453-64.
32 Altma n PL, Dittmer DS. Growth: In cluding Reproduction alld
Morp hological De velop men t. Washington DC: Federation of
Ame rican Socie ties for Exp erime ntal Biology, 1961.
33 Handy TC, Nichol s GR , Buchin o JJ. A pediatric fo rensic
medicine prog ram. In Dimniick JW, Singer DB (eds) Forens ic
A spects ill Pediatric Pathology. Perspectives in Pediatri c
Pathology. Basel: Ka rger, 1995, pp. 87-95.
34 Kaplan JA, Fossum RM. Patterns of facial resuscitation injury
in in fan cy. Am ] ForellS ic Med Path o/199 4; 15:187-91.
35 Downes AJ, Crossland DS, Mellon AF. Prevalence and distribution
of petechiae in well babies. Arch Dis Child 2002; 86:291-2.
35 Byard RW, Krous HF. Petechial hemorrhages and unexpec ted
in fant death. Legal Med (Tokyo) 1999; 1:193-7.
37 Krous HF, Nad eau JM, Byard RW, Blackbourn e BD . Oronasal
blood in sud den infant death. Am ] Fore nsic Med Path ol
200 1; 23 :346-51.
38 Mayes C, Macleod C. When 'N AJ' means not act ually injured.
BM] 1998; 318:1127 - 8.
39 McCann J, Reay D, Siebert J et al. Postmortem perianal findings
in children. Am] Foren sic Nled Patho/2006; 17:289-98.
40 Valdes-D ape na M, McFeel ey PA, Damus KH et al.
Histopa thology Atlas fo r the Sudden Infant Dea th Syndrome:
Fi ndings derived from the Natio l1 al/llstitute of Child Health
and Hum an Develop ment Cooperati1!e Epidemiological Study
of Sudden Infant Death Syndrome (SIDS) Risk Factors.
Washington : Armed Forces Institute of Pathology, 1993.
41 Kro us H. The interna tional sta ndardize d autopsy protocol for
sudden unexpected infant death. In Rognum TO (ed.) Sudden
Infant Death Syndrome. New Trends in the Nineties. Os lo:
Sca ndin avian Uni ve rsity Press, 1995, pp. 81-95.

';1 van Baarlen J, Schuurman H-J, Huber J. Acute thymus
involution in infancy and childhood: a reliable marker for
duration of acute illn ess. Hum Pathol 1988 ; 19: 1155-60.
";3 Beckwith JB. Intrathoracic petechial hemorrhages; a clue to
the mechanism of death in sudd en infa nt death sy ndrome?
Ann Acad NY Sci \988; 533:37-47.
~4 Muir R, Niven JSF. The local formation of blo od pigments.
J Pathol Ba cteriol 1935; 41 :183-97.
-lS Keeling JW. The perinatal n ecropsy. In Keeling JW, Khon g TY
(eds) Feta l and Neona ta l Pathology, 4th edn. London:
Sprin ger, 2007, pp. 20-53.
46 Hardman JM. Cerebrospinal trauma . In Davis RL, Ro bertson
DM (eds) Textbook of Neuropathology, 3rd edn. Baltimore, MD:
Williams Et Wilkins, 1997.
47 Parsons MA, Smart RD. Necropsy techniques in ophthalmic
pa thology. Best Practice 164 J Clin Path 01 2001; 54:417-27.
48 Lee WR (ed.). Examination of the globe. Techni cal aspects. [n
Oplithalmic Histopathology, 2nd edn. London: Springer­
Verlag, 200 2, pp. 1-33.
49 Reichard RR, White CL Ill , Hogan RN et at. Beta-amyloid
precursor protei n immunohistoch emistry in the evaluation of
pediatric traumatic optic nerve inju ry. Ophthalmology 2004;
111 :822-7.
50 Wigglesworth JS. Performa nce of the perinatal autopsy.
In Wiggleswowrth J S (ed.) Peri/lOtal Pathology, 2nd edn.
Phil adelphia , PA: Sau nders, 19 86, pp. 23-40.
51 Russell JGB. Radiolog ical assessment of fetal maturi ty.
J Obstet Gynae Brit Cwlth 1969; 76:208-19.
52 Ste mpfle N, Huten Y, Fondacci C et al. Fetal bone age
rev isited: proposal of a new radiographic score. Pediatr Radiol
1995 ; 25:551-5.
53 Biankin SA , Arbu ckle SM, Graf NS. Autopsy findings in a
series of five cases of feto maternal haemorrh age. Pathology
2003; 35:319-24.
54 Ha rlow FHD, Spencer JAD. Obstetrics for the neonatologist. In
Rennie JM, Roberton NRC (eds) TeXTbook of Neonatology.
London: Churchill Livin gston e, 1999, pp. 157-77.
55 Morris JF, Hunt AC. Breaking streng th o f the umbilical cord.
J Forensic Sci 1966; 11:43-9.
56 Wigglesworth JS, Husemeyer RP. Intracra nial birth trauma in
vaginal breech delivery: the continued importa nce of injury
to the occipital bone. Br J Obstet Gynecol 1977; 84:684-91.
57 Dorovini-Zis K, Dolman CL. Gestational develo pment of brain.
Arch Pathol Lab Med 1977 ; 101 :192-5.
58 Feess-Higgins A, Lan'oche J-c. De1Jelopmel1 t of the Human
Foetal Brain. An Anatomical Atlas. Pa ris: Inserm CNR S
Masson, 1987.
59 Ya tes PO. Birth trauma to the vertebral arteries. A rch Dis
Child 1959; 34:436-41.
60 Marchet ti D, Belviso M, Marino M, Gaudio R. Evalu ation of
the pl acen ta in stillborn fetuses to estimate the time of death.
A m J Forensic Med Pat/wi 2007; 28:38-43.
61 Pinar H, Sung CJ , Oyer CE, Si nger DB. Reference values for
singleton and twin placental weights. PeliiMr Patlwl Lab Med
1996 ; 16 :901-7.
62 Naeye RD. Umbilical cord length: clini cal significance.
J Pedi atr 1985; 107:278-81.
63 Genest DR. Estimating th e tim e of death in stillborn fetuses:
II. Histologi c evaluation of the pla centa; a study of 71
stillborns. Obstet Gyneco11992; 80:585-92 .
64 Benirschke K, Kaufmann P, Bae rgen R. Pathology of the
Human Place nta , 5th edn. New York: Springer, 2006.
65 Fox H, Sebire N. Pathology of the Placen ta, Major Problems in
Pat/wlogy, 3rd edn. London: Saunders, 2007.
66 Khong TY. The placenta and umbilical co rel. In Keeling JW,
Khong TY (eds) Fetal and Neonatal Pathology, 4th edn.
London: Springe r, 2007 , pp. 54-89.
References I 165
67 Altshuler G. Placenta w ithin the medicolegal imperative. Arch
Pathol Lab Med 199 1; 11 5:688 -95.
68 Macphe rson 1. Fact a nd fancy. What can we really tell from
the placenta? Arch Path Lab Med /991; 115: 672 -81.
69 Wigglesworth JS. Pathology of intrapartum and early
neon atal death in the normally form ed infa nt. In
Wigglesworth JS, Singer DB (eds) Textbook of Fetal and
Perinatal Pathology, 2nd edn . Ma lden , MA: Blackwe ll Science,
pp . 251-68.
70 Keeling JW. Intrapartum problems. In Keelin g JW, Khong TY
(eds) Fetal and Neonatal Pathology, 4th edn. London:
Springer, 2007, pp. 273-96.
71 Teng FY, Sayre JW, Vacuum extraction: does duration predict
scalp if\iury? Obstet Gynecol 1997; 89 :281-5.
72 Amar AP, Aryan HE, Meltzer HS, Levy ML. Neonatal subgaleal
hematoma causing brain compression: results of two cases and
review of the literature. Neurosurgery 2003; 52: 1470-4.
73 Robinson RJ , Rossiter MA. Massive subaponeurotic hae monhage
in babies of African origin. Arch Dis Child 1968 ; 4 3 :684-7.
74 Doward W, Sgouros S. Acute su bdural haematom as follo wing
Ventouse-a ssisted delivery. Pediatr Neu rosurg 2001; 35:336,
75 Towne r D, Castro MA, Eby-Wilkens E e t al. Effect of mode of
delivery in nulliparous women on neonatal intra cranial
injury. N Engl J Med 1999; 341: 1709-14.
76 Vachharajani A, Mathur A. Ultrasound-guid ed needle
aspiration of cranial epidura l hema toma in a neonate:
trea ting a ra re complication of vacu um extraction. A m J
Perinatoi2002; 19:401-4.
77 Perri n RG, Rutka JT Drake JM et al. Ma nagement and
outcomes of posterior fo ssa subdural hematomas in neonates .
Neu rosu rgery 1997; 40:1190-9; discussion 1199- 2000.
78 Whitby EH, Griffiths PO, Rutter S et al. Frequency and na tural
histo ry of subdural haemorrhages in ba bi es a nd relation to
obstetric factors . Lancet 2004; 363 :846-51.
79 Salman M, Crouch man M. What ca n caus e subdura l
haemorrhage in a term neo nate? Paediatr Today 1997 ;
5:42-5.
80 Davies MJ , Pomerance A, Lamb D. Tec hniques in exa mination
and a natom y of the heart. In Dav ies MJ , Pomerance A (eds)
The Path ology of the Hea rt. Oxford: Blackwell Sc ientific
Pu blica tions, 1975, pp. 26-31.
81 Michaud K, Romain N, Taroni F et al. Evaluation ofa
simplified method of th e conduction system a nalysis in 100
forensi c cases . Forensic Sci Int 2002; 130:13-24.
8 2 Ashworth MA. The cardiovascular system. In Keeling JW,
Khong TY (eds) Fetal and Neonatal Path ology, 4th edn.
London: Springer, 2007 , p. 613.
83 Ho SY, Anderson RH. Conduction tissue and SIDS. Am NY
Acad Sci 1988; 533:176-90.
84 Ma tturri L, Otta viani G, Ramos SG, Rossi L. Sudden infant
death syndrome (SID S): A study of cardiac conduction
system . Cardiovasc Pat1lO1 2000; 9:137-45.
85 Keeling JW, McKenzie KJ, Ironside J et al. Does detailed
neuropathological ex amina tion o f the fi xed brain in sudden,
un expected dea th in infa ncy contribute to the evaluation of
the death? J Pat/wi 2002 ; I 98(Suppl.) :2 4A.
86 Report of a Working Group . Sudden Unexpected
Death in Infancy. London: Royal College of Path ologists
and Royal College of Paedi atrics a nd Child Hea lth,
2004.
87 Fleming P J, Blair PS, Sid ebotham P, Hay ler T. Investiga ting
sudden unexpected deaths in infancy and childh ood and
ca ring for bereaved families: an integra ted multiagency
appro ach. BMJ 2004; 328:331-4.
88 Genest DR , Singer DB. Es timatin g the time o f death in
stillborn fetuses: JII. Extern al fetal examination; a study of
86 s tillborns. Obstel Gynecol 1992 ; 80:593-600.
I CHAPTER 9 I
PATHOLOGY OF NEUROLOGICAL
ABNORMALITY IN EARLY LIFE
Waney Squier
Introduction 166
Clinical manifestations of early brain damage:
cerebral palsy 167
Timing of injuries by histology 167
Acquired intra-uterine damage 169
INTRODUCTION
may be very strongly held. The findings in the brain are of
Examination of the nervous system forms a critical part of
the autopsy exami nation of any infant who dies, whether
suddenly and unexpectedly or following recognized illness.
The cause of sudden infant death is not always proximate
insult or disease; pre-existing diseases of, or damage to, the
nervous system may cause sudden death. Babies with neuro­
logical disease, for example, are more pron e to aspiration of
feeds and to epilepsy. Even quite extensive congenital brain
damage acquired in utero or at the time of bilih may not
present clinically until months or years later, while still
hav ing the potential to cause sudden death . It may be that a
'secon d insult' is required, for example velY mild brain mal­
formation (microdysgenesis) is seen in patients with epilepsy
but may not cause seizures until there is another insult such
as trauma. Similarly, babies with metabolic or mitochondrial
disease may become profoun dly unwell a nd die when chal­
lenged by a respiratory infect ion. In these babies careful
exa mination of skeletal muscle with histochemistlY and
review of pre-mortem biochemistry is necessary if an accu­
rate diagnosis is to be made. Chronic subdural haemorrhage
may go quite unrecognized for weeks or months until dis­
covered by routine head circumference measurement or
even by parental observation of increasing head size.
The pathologist making an autopsy exa mination in sud­
den infant death is faced with an awesome task. Th ere will
often be great pressure from those who have cared for the
child in life to find a pa rticular cause of death . Opinions
concerning the possibility of natura.1 or unnatural causes
Birth-related injury 173
Stroke in the developing brain 176
Metabolic disorders 177
Infections 177
References 178
paramount importance in such cases but may also be very
subtle. If trauma is suspected the possible mechanisms
must be fully considered and due attention paid to exami­
na tion of the stru ctures of the neck and spinal cord as well
as the brain. Meticulous neuropathological examination
can identify no t only the cause of neurological disease in
childhood but, in the case of acquired damage, can also
assist in evaluating when the damage has occurred. How­
ever, the effects of the terminal condition of the baby must
be taken into consideration. Hypoxia and resuscitation, as
well as a period of ven tilation , will lead to brain swelling
and ge neral organ failure, disturbance of blood clotting,
leaking from blood vessels and slowing or cessation of cel­
lular reactive processes, serious ly hampeling the ab ili ty to
time the injUIY. Careful correlation with brain scans take n
closer to the time of coll apse will assist in distinguishing
primary injUly from these secondary changes.
Neurological abnormalities may resu lt from genetic or
meta bolic disease, diseases acquired in utero, during deliv­
ery, in the pelinatal period or in early life.
Intra-uterine damage may not be symptomatic until
weeks or months after birth, for example cerebral palsy is
not usually definitively diagnosed until 5 years of age. Con­
versely, babi es damaged immediately before or during bilih
tend to exhibit signs or symptoms such as floppiness,
depression and asphyxia in the first days of life.
Once the brain has been damaged there will be tissue
loss followed by atrophy. Tissue regeneration and plasticity
due to compensatory hyperplasia has been demonstrated in
 Timing of injuries by histology I 167

the immature human brain.l Following even static and fragments of cerebellum in the su barachnoid space at mul­
non-progressive insult, clinical signs may evolve due to tiple levels of the spinal cord.
atrophy of associated and connected brain areas. Histologically, oedema is better seen in densely packed
white matter tracts than in grey matter.

CLINICAL MANIFESTATIONS OF EARLY BRAIN Cell Death

DAMAGE: CEREBRAL PALSY
Cerebral palsy remains the mo st common form of chronic
motor disability in children (1-2 cases per 1000 live births).
The full effects of early damage may not be apparent
until several years of age. Signs include abnormal control of
movement or posture, cognitive impairment, seizu res and
blindness. Sudden death may be due to seizures or swallow­
ing disorders and aspiration.
Timin g the damage is crucial not only to the under­
standing the possible aetiology, but also it will be critical in
any potential litigation.
The cause of cerebral palsy remains unknown in the
majority of cases. In term babies the most impoltant causes
are stroke (17 per cent), maternal infection (12 per cent), mul­
tiple pregnancy (10 per cent) and birth asphyxia (6 per cent).
Birth injury, genetic and metabolic diseases, fetal infections
and toxins account for a small percentage of cases. 2
TIMING OF INJURIES BY HISTOLOGY
Th e two best-described patterns of cell death are necrosis
and apoptosis (Fig. 9.1).
Necrotic neurones develop intense cytoplasmic eosino­
philia, the nuclear membrane lyses and the chromatin dis­
perses into a fine web. Cytoplasmic eosinophili a alone may
be a reversible or artefactual change, often seen in surgical
material. Nucleolysis is part of the irreversible process. These
changes take less than 12 hours, in the adu lt brain they are
thought to take 5-6 hours to develop but timing in the infant
brain has not been accurately documented.
Apoptotic cells undergo pyknosis, when the nucleus
becomes shrunken, rounded and intensely basophilic. Later
the nucleus forms multiple rounded densely staining masses
(kalyorrhexis). Apoptotic cells are readily phagocytosed and
cause no inflammatory response. This process probably
takes 12 hours to complete but, again, accurate fig ures are
not available for the infant.
Apoptosis is a more frequent in the immature brain than
in the adult brain.

Histological timing cannot be regarded as precise and must Macrophage Response

always be interpreted in the light of all the other aspects of
the case, partiCUlarly the clinical history. Much of the data
During the first 24 hours after hypoxic-ischaemic injury
below has been previously presented with source refer­
(HIl) , microglial cells (the intrinsic phagocyte population of
ences.] The earliest reactions in the infan t brain are oedema,
neuronal death and cellular reactive changes (Table 9.1).
Timings must be interpreted with caution, palticularly if the
Table 9.1 Early responses to injury within the brain
baby was venti lated prior to death.
During ventilation the brain may become very swollen Cerebral oedema Minutes to 1 week
and the blood supply compromised (respirator brain). In Cell death
these circumstances the reactive processes may be slowed Necrosis 5-12 hours
or otherwise modified. The dura retains a better blood sup­ Apoptosis <12 hours
ply via the emissary veins of the skul.! bones and in the case Macrophages
of subdural haemorrhage dural histology may prove to be Microgl ia <24 hours
a better indicator of timing. Phagocytosis 3 days-yea rs
Gliosis 12 hours-6 days
Cerebral Oedema Capillary reaction
End othelial swelling 1-3 days
Capillary proliferation 5-7 days
Brain swelling may begin within minutes of injUlY, reach­
Haemosiderin 48 hours-years
in g a maximum after 1 week. The speed of swelling is sub­
Mineralization 8 days-years
ject to huge individual variation. It causes narrowing of the
Axonal damage"
sulci and compression of the ventricles. On slicing the
Beta amyloid precursor protein 0- 3 hours
brain, the cortical ribbon may appear prominent, with soft­
expression
ening and grey discolouration of the underlying white
Silver staining bulbs 12-18 hours
matter. Flattening of the gyri is uncommon as the unfused
Haematoxyli n and eosin staining 15-24 hours
infant skull can accommodate considerable brain swelling
bulbs
without compression. However, it is not uncommon to find
cerebellar tonsillar herniation and necrosis with displaced "Timing from ad ult tissue studies.
168 I Pathology of neurological abnormality in early life
Figure 9.1 Acute severe hypoxic-ischaemic injury. Section of
cerebellum. Asingle remaining Purkinje cell shows intense
cytoplasmic eosinophilia and nucleolysis - typical features of
necrosis. In contrast, the granule cells are pyknotic with rounded,
densely basophilic nuclei. Many contain rounded apoptotic bodies
(haematoxylin and eosin).
the brain) are prominent and are recognized by their long,
rectangular nuclei. These are replaced by macrophages in up
to 2 days after injury and by 3 days these cells may contain
phagocytosed debris or red cells within their cytoplasm.
Damaged tissue is removed by these phagocytes, which
frequ ently cluster around blood vessels where they may
remain for months or years . If there has been haemorrhage
then red cells are seen in the cytoplasm of macrophages in
the early stages; they are later converted to haemosiderin,
which is brown in haematoxylin and eosin (HE) stain but
after about 48 hours may be detected with Peds' stain.
Gliosis
Swollen astrocytes are seen from 12 hours after insult. Fib­
rillary gliosis takes about 6 days to occur. On the basis of
radiological observations, it has been assumed that gliosis
does not occur in the first half of gestation; however, fetal
injury during amniocentesis at 17 weeks elicits a brisk gli­
otic response. 4
Capillary Reaction
Endothelial cell swelling is a helpful and sensitive marker
of tissue damage, becoming prominent 1-3 days after
Figure 9.2 Acute severe hypoxic-ischaemic injury. Reactive
capillaries in the white matter showing branching and
proliferation. The endothelial cells are plump, several layers
thick and many are apoptotic (haematoxylin and eosin).
injury (Fig. 9.2). Endothelial markers CD31 and CD34 assist
in identifying reactive vessels.
Capillary proliferation takes 5-7 days and is seen at the
periphery of necrotic areas of brain. It is often particularly
marked in the ischaemic cortex, particularly in the depths
of the sulci, and is associated with the magnetic resonance
imaging (MRI) appearance of 'cortical highlighting'.
Haemosiderin
This is stainable in Perls' reaction 48-72 hours after haemor­
rhage. In late stages, iron is seen in astrocytes or free within
tissues as well as in macro phages. It is common to find
specks of haemosiderin within the dural layers in infants.
Cyst Formation
Tissue breakdown and removal by macro phages causes cysts
to develop, usually \0 days to several weeks after injury.
Mineral ization
Mineral is readily deposited in damaged areas of the fetal
brain. It begins as early as 8 days after injury and persists
for years (Fig. 9.3). The pattern of mineralization may be a
helpful diagnostic guide. Widespread irregular mineraliza­
tion is seen after infections such as herpes; periventricular
mineralization typically occurs in cytomegalovirus (CMY).
Patches of mineral in areas vulnerable to ischaemic injury
suggest this cause.
Axonal Damage
Damage to the axonal membrane causes interruption of the
intra-axonal transport of proteins, leading to irregular
 Acquired intra-uterine damage I 1 69

Figure 9_3 Periventricular leucomalacia. The edge of the lesion

consists of a band of mineralized cell bodies and axonal
fragments. Within this area, the scarred tissue contains glial cells (b)
and many macrophages with pigmented cytoplasm (haematoxylin
and eosin).

swellings or varicosities of the axon; if sufficiently severe,

rounded swellings or axonal spheroids develop. These
have been traditionally demonstrated with HE staining or
with silver stains, but immunohistochemistry using anti­
bodies to the protein 'beta amyloid precursor protein'
(0APP) has proved a highly sensitive and specific marker of
axonal damage (Fig. 9.4) . Axons express 0APP after a vari­
,
ety of insults, including HII and metabolic damage as well
as trauma;5 however, this stain cannot distinguish the
causes of injury - this depends upon the specific pattern
and distribution of expression, which, again, alone, is not
diagnostic.
Studies in the adult brain have shown that 0APP may be
expressed from less than 30 minutes after injury6 but it is
not a reliable indicator of timi ng. Silver staining takes some
12-18 hours to develop and axonal retraction balls are not
seen with HE staining until about 15-24 hours after injuryJ
Beta amyloid precursor protein staining persists for
about 2 weeks in spheroids but may persist in irregular
granular deposits for several months . Microglial cells accu­
mulate in areas of axonal damage within 4-5 days.

ACQUIRED INTRA- UTERINE DAMAGE

Early .-.

Damage in the first 20 weeks of gestation causes malfor­
mation. Specific malformations may indicate the timing of
the causative insult as structures are malformed only dur­
ing th eir period of formation . Once formed, structures may
show evidence of destruction. An example is seen in the
corpus callosum, which develops from front to back
.
Figure 9.4 Periventricular leucomalacia. Small focus of
periventricular leucomalacia in an infant who died 11 days after
severe hypoxic-ischaemic injury. There is a small collection of
glial cells and macrophages, among which are proliferating
capillaries and axonal spheroids. (a) HEtE; (b) C034, showing the
newly formed capillaries; (c) ~ APP stain, showing the axonal
swellings in the infarcted area.
170 I Pathology of neurological abnormality in early life

between 11 and 17 weeks of gestat ion . If development is

interrupted the posterior part will be deficient, whereas
lesions in an terior or mid-p arts with an intact posterior
part suggest later insults.

AGENESIS OF THE CORPUS CALLOSUM

The corpus callosum may fail to form at all. In some cases

this is due to severe und er-development of the cerebral
cortex, wit h insufficient fibres generated to cross the mid­
line. In others, a bundle of fibres forms but fails to cross the
midline and mns instead in an anterior-p osterior direction .
This is known as a bundle of Probst and causes distortion
of the ventricles, which take on the typica l 'Viking's horn'
appeara nce that is recognized on scans (Fi g. 9.5).
Agenesis of the co rpus callosum may be an isolated
findin g, but is more frequently associated with other mal­
formations such as cerebella r hypoplasia. Its presence
should prompt a search for these.

NEURONAL MIGRATION DISORDERS

One of the better-understood developmen tal processes

is neuro nal migration. All neurones migrate from the place
of their genesis in the germinal matrix to their final posi­
tion in the mature brain. Neuronal migration to the cere­
bral cOltex takes pl ace in the human brain by radial an d
tangential pathways at between 6 and 23 weeks of gesta­
tion. Interruption with this process leads to cOltical dyspla­
sia and neuronal ectopias. These malformations may be
identified on MR scans of the brain and are a frequent find­
ing in patients with intractable epilepsy. They should be
sought in any child who dies unexpectedly as they may
themselves be responsible for seizures 8 Neuron al migra­
tion disorders may be genetic in origin, but many are the
result of early acquired damage occurring before migratio n
is complete.

PORENCEPHALY

Porencephaly means a hole in the brain. The term is applied

to any parenchymal defect, resultin g, for example, from
old trauma, haemorrhage or infa rction. The term is also
applied to asymmetric outpouching of the ventricle due to
deep white matter hae morrhagic infarction. Heterotopic
grey matter in the walls of a porencephalic cyst ind icates
an origin before the end of neuronal migration.
SCHIZENCEPHALY/POLYMICROGYRIA
Schizencephaly and polymicrogyri a are sometimes grouped
together due to their frequent coexistence. 9
Schizencephaly is a defect invo lving the whole thick­
ness of the brain wa ll with communication between the
ventricle and the brain surface. The cleft is frequently lined
Figure 9.5 (a) Age nesis of the co rp us callosum. Coronal slice of
fetal brain at 22 weeks' gestati on. A large bundle of fibres is
running parallel to the midline on each side and displacing the
ventricles laterally. There is no corpus cal losum; instead, the
med ial borders of the hem ispheres curl in wa rds and a bu ndle of
Probst (P) is seen in the medial wa ll of the lateral ventric le.
(b) Section of the same bra in stained with haematoxylin and
eosin. Note that there are small bilateral germinal matrix
haemorrhages. P, bundle of Probst. (c) Hi gher power ima ge shows
the latera l ventri cle and germ inal matrix. In the right wall of the
ve ntricle, fibres of the bundle of Probst (P) have been transacted
(haematoxylin and eosin) . GM, germinal matrix.
 Acquired intra-uterine damage I 171

by ectopic neurones, and the adjacent cortex may be
dysplastic. Schizencephalic clefts are often bilateral,
and most commonly found in the region of the Sylvian fIs­
sure. Although originally considered to be a malformation
due to a focal defect in the germinal matrix,1O schizen­
cephaly is now generally considered to be a destructive
lesionY Rare familial cases are described. Mutations in the
EMX2 gene were implicated but this association has not
been confIrmed. 9
Polymicrogyria is a cOliical malformation resulting
from damage prior to 28 weeks of gestation. iJ.12 Macro­
scopically, gyri are irregular and small , but the diagnosis
can be confirmed only by histology. Two histological pat­
terns are described: unlayered and four-layered, but their
distinction has no aetiological significance and they com­
monly coexist. A common and determining feature of
polymicrogyria is abnormal fusion of adjacent gyri, lead­
ing to continuity of the superficial cortical surface while
the neuronal laminae form festoons below. Large blood
vessels are trapped in the seams of fused gyri. It is not a
uniform entity; several specific patterns of polymicrogyria
are described, both by histology and macroscopic distribu­
tion. Acquired polymicrogyria occurs randomly throughout
the cortex or on the borders of destructive lesions, whereas
symmetrical and bilateral patterns may correlate with spe­
cific genetic disorders. I] Polymicrogyria and schizen­
cephaly are both described after ischaemic events and
infections such as CMY.
Classification systems of cerebral malformations based
on MRI studies are not always accurate, owing to the limi­
tations of MRI resolution. What may appear to be specific
malformations on imaging may be both structurally and
aetiologically heterogeneous.
It has long been accepted that white matter disease results
from ischaemia of the developing white matter. However,
in recent years intra-uterine infection and inflammation
have been implicated in its pathogenesis. Epidemiological
studies have demonstrated an association between placen­
tal and amniotic infection and maternal pyrexia and
neonatal neurological abnormality. IS Histological evidence
of inflammation in the placenta is associated with brain
lesions and later cerebral palsy. J9
Three pattems of white matter damage are described. Com­
mon to all is gliosis and capillary proliferation. When seen
alone this is described as telencephalic leucoencephalopathy.2o
The most common form, pellventllcttiar leucomalacia, is
recognized by multiple infarcts in the deep periventricular
white matter that can be seen by the naked eye as tiny cysts,
often rimmed by a yellow mineralized zone (Fig. 9.6). The
infarcts tend to occur in the deep frontal white matter and
around the trigone. Histologically, they appear as focal glial
scars - centrally cystic and surrounded by macro phages and
damaged axons, which are frequently calcifIed (see Fig. 9.3,
p. 169). Recent lesions are accompanied by capillary prolif­
eration (see Fig. 9.4).
The most dramatic manifestation of white matter dam­
age is multicystic leucoencephalopathy (Fig. 9.7), in which
large areas of damaged white matter are replaced by cysts,

Late

WHITE MAnER DAMAGE

This is the commonest cause of cerebral palsy. It typically

occurs before the brain is mature, usually between 24 and
36 weeks of gestation, but this is not exclusive and some
white matter damage almost invariably accompanies severe
hypoxic-ischaemic injury at term. I Studies of infants
undergoing cardiac surgelY have shown white matter dam­
age to develop after term. Magnetic resonance imaging
studies at day 5 have confirmed development of white mat­
ter injury, 14 and it has been determined by neuropathological
examination that it can develop up to 11 months
postnatally. IS
White matter damage causes diplegic cerebral palsy by
involvement of the motor tracts as they pass through the
deep white matter of the cerebral hemispheres. This form of Figure 9.6 Periventricular leucomalacia. Fixed brain slice with
cerebral palsy is typically seen in very-Iow-bilih-weight tiny areas of periventricular leucomalacia in the deep white
and premature infants. There is little doubt that grey matter matter, close to the ventricular wall; although some have yellow
damage, particularly the deep nuclei, accompanies most flecks of mineralization, others are more subtle and show as dusky
white matter damage. 16 • 17 patches adjacent to vessels.
172 I Pathology of neurologica l abnormality in early life
Figure 9.7 Multicystic leucoencephalopathy. Almost the entire
white matter is replaced by large cysts, some is traversed by glial
strands. The overlying cortex is thin and gliotic. Microscopical
examination showed extensive gliosis through grey and white
matter.
particularly in the subcortical areas. They stretch the over­
lying cortex, which becomes gliotic with loss of neurones
and there is often associated gliosis in deep grey nuclei.
Infections such as herpes, varicella and listeria may pro­
duce similar white matter damage. 21 ,22
HYPOXlC-ISCHAEMIC INJURY AT TERM
There are three basic patterns of term hypoxic-ischaemic
injury, which depend on the nature, intensity and duration
of injury. Evidence largely depends on animal models, as
well-documented data from human series is limited.
Predominantly Brainstem Damage (Cardiac
Arrest Encepha lopathy)
This pattern results from an acute or total interruption of
cerebral blood supply, probably for a perio d of 10-25 min­
utes, such as may occur in a cardiac arrest. Owing to the
speed of interruption of circulation there is no time for
adaptive redistribution of blood supply, and damage occurs
in the most highly metabolically active brain areas - the
nuclei of the brainstem and thalamus (Fig. 9.8). Babies so
affected either die or show absent gag and corneal reflexes,
and swallowing and breathing difficuities. 2J
Basal Ganglia{Tha lamu s
Severe acute, profound asphyxia causes damage predomi­
nantly to the deep grey nuclei, thalamus, internal capsule,
hippocampus and central cortex 24 ,25 (Fig. 9.9). The mecha­
nisms include sudden severe episodes of asphyx ia : for
example cord prolapse, uterine rupture or placental separa­
tion, acute or chronic hypoxia, or multiple severe episodes
Figure 9.8 'Cardiac arrest encephalopathy'. Section of medulla
showing well-defined symmetrical areas of infarction in the
dorsal medulla, close to the floor of the fourth ventricle
(haematoxylin and eosin).
of asphyxia in utero. 26 In the human infant, almost all
occur after labour has begun, suggesting that these insults
also represent acute on damage on a brain that has already
been subjected to episodes of relatively mild hypoxia.
Cortical and Watershed Da mage
Chronic or repeated asphyxia or hypotension of 30 minutes
to several hours' duration allows adaptive changes to
25
OCCUr. Blood is shunted form the forebrain to the vital
and highly metabol ically active brainstem areas. Although
this preserves the deep grey nuclei, the cerebral cortex
becomes vulnerable, particularly in the distal, watershed
territories. The most frequently involved is the parasagittal
cortex. Gyral crests are relatively spared, as well as
the more vulnerable depths of the sulci atrophy. The
gyri become mushroom-like in appearance - 'ulegyria'
(Figs 9.10 and 9.11).
GERMINAL MATRIX HAEMORRHAGE
This is most common before 36 weeks of gestation while
the germinal matrix persists in the ven tricular wall. The
thin-wa lled capillaries of the germinal matrix may rupture
after hypoxic damage and surges in cerebral blood pres­
sure. The most common time for haemorrhage to occur is
within the first 48 hours of postnatal life. If the bleed
remains confined within the matrix then the prognosis is
good (Fig. 9.5a, p. 170). However, if blood ruptures into the
ventricles it causes shedding of the ependymal lining, reac­
tive gliosis and ependymal proliferation. If this occurs in
the narrow aqueduct or drainage channels of the fourth
- - .- - ~

•
Figure 9.9 Acute profound hypoxic-ischaemic injury. Basal
ganglia damage. (a) Fresh brain slice showing dusky
discolouration of the deep grey nuclei, particularly the lentiform
nuclei (arrows). The cortex appears as a wh ite ribbon and the
ventricles are narrow, owing to oedema. (b) High-power image of
the thalamus, many months after hypoxic-ischaemic injury,
showing neuronal loss and dense gliosis. Many remainin g
neurones are mineralized (haematoxylin and eosin).
ventricle it will cause obstruction of cerebrospinal fluid
(CSF) flow, resulting in ventliculomegaly or hydrocephalus.
The histological sequel is formation of ependymal rosettes,
gliosis and Perls' positive material, which may persist for
months or years.
PARENCHYMAL VENOUS INFARCTI ON
About 15 per cent of germinal matrix haemorrhage leads to
wedge-shaped infarction of the adjacent deep white matter,
fanning out from the haemorrhage. The draining veins are
congested and macrophages containing red cells pack the
white matter. Resorption of the damaged tissue leads to
porencephalic cyst formation in the deep white matter or
unilateral dilatation of the ventricle. The unilateral damage,
Birth·related injury I 1 73
Figure 9.10 Old hypoxic-ischaemic injury. There is loss of
cortex from the depths of parasagittal sulci, leading to focal
ulegyria (arrowheads).
most commonly frontal, distinguishes this lesion from
leucomalacia, which is much more commonly bilateral.
Infants who suffer venous parenchymal infarction are at
risk from later hemiplegia due to destruction of motor fibres
as they pass through the deep white matterY
BIRTH-RELATED INJURY
Brain injury acquired at birth is usually due either to trauma
or to asphyxia. Today, birth trauma is relatively uncommon
owing to improved obstetric care and high rates of Cae­
sarean section. Despite a fivefold increase in Caesarean sec­
tions over the last 30 years, the rates of cerebral palsy have
remained constant in industrialized countries? with many
of those affected being low-birth-weight or premature
babies who survive with white matter damage.
Birth Trauma
One of the largest pathological studies of bilth il1iury is that of
Towbin, published in 1970, in which he studied 600 still births
and neonatal deaths. 28 In this study, trauma accounted for
some 1-2 per cent of newborn deaths, but many more infants
suffered non-lethal injuries. Birth trauma is associated with
both precipitous and prolonged labour, with large infant size,
fetal malpresentation and difficult instrumental extraction. 28
Scalp and Skull
Scalp and, in particular, subgaleal haemolThage may follow
instrumental delivery, particularly vento use extraction.
174 I Pathology of neurological abnormality in early life
Figure 9.11 Old hypoxic-ischaemic injury. Cerebellar cortex
showing old hypoxic-ischaemic injury with large areas of
neuronal loss and gliosis. (a) The damage is most severe in the
depths of folia. (b) At higher powe r only one Purkinje cell
remains. The granule cell layer is replaced with glial cell s.
Subgaleal haemorrhages may be very large and cause cra­
nial compression as well as significant blood loss from the
circulation. 29 Skull fracture is seen occasionally after normal
delivery but, more usu a lly, this follows instrumental deliv­
ery.30 Extracranial birth injuries are considered in Chapter 8.
Intracranial Haemorrhage
The commonest form of birth injury is intracrani al haemor­
rhage, of which subdural haemorrhage is the most
Figure 9.12 Intradural haemo rr hage in 2-day-old infant
fo llowing birth asphyxia. There is extensive fresh haemorrhage
into the falx adjacent to the sagittal sinus.
frequent. Among 41 consecutive cranial bilih injuries in term
infants, Pollina et al31 found that 73 per cent had subdural
haem OiThage, 20 per cent had subarachnoid haemorrhage
and 20 per cent had parenchymal haemorrhage. Overall, 24
per cent had cephalhaematomas, 5 per cent had intraventlic­
ular haemorrhage and 5 per cent had skull fracture. The fre ­
qu ency of birth injury in the institution was not stated.
Method of delivelY is important. Towner et al 32 found an
increased incidence of intracranial haemorrhage after instru­
men tal delivery with ventouse or forceps and emergency
Caesarean section. Incidence was lower after Caesarean sec­
tion before labour. Pollina et al found a higher incidence of
intracranial injury after forceps and vacuum deliveries,
whereas Chamnanvanakij et al 33 found subdural haematoma
on computerized tomography (CT) in 26 symptomatic term
infants, half of which followed instrumental delivery and
half after s ponta neous delivery. In a UK-wide study,
O'Mahoney et al 34 reported a lower incidence of cranial birth
trauma. Amongst 873 intrapartum deat hs of mature babies,
they found only 37 cases out of 181 in which cranial trauma
a ppeared likely from the history. Instrumental delivery was
clearly associated with trauma; multiple pulls and use of
multiple instruments were particular risks.
INTRADURAL HAEMORRHAGE
When the dura is examined, intradura l haemorrhage is fre­
quently observed - particularly in the falx and tentorium. It
is seen in the n eonate in many circumstances, often when
there is evidence of hypoxic-ischaemic injury. It is also
described in spontan eous abOliions and intra-uterine deaths
when the fetus is likely to have experienced hypox ia. 35 It is
particularly obvious in the dural folds and may be confined
to the dura itself or leak on to th e subdural surface, although
this is not common (Fig. 9.12).
 Birth-related injury I 175

SUBDURAL HAEMORRHAGE

Subdural haemorrhage is the most common form of birth­

related intracranial haemorrhage. Recent studies using MR
scans in asymptomatic neonates have shown a surprisingly
high incidence of subdural haemorrhage. Looney et al 36
;o und small subdural haemorrhages in 26 per cent of vagi­
na l deliveries and Whitby et al 37 in 9 per cent, the differ­
ence probably due to use of scanners of different
resolution . Infants with congenital heart disease had an
incidence of asymptomatic intracranial haem orrhage of 62
per cent after normal delivery.38

Method of Delivery
Most subdural haemorrhages follow normal delivery, but
are proportionally more frequent after forceps and ve ntouse
deliveries,37 although, in the above studies, not after Cae­
sarean section.

Location
The source of bleeding in bilth-related subdural haemor­
rhage remains uncertain. It is striking that the great maj ority
of birth-related subdural haemorrhage is found in the poste­
rior fossa or around the occipital poles, close to the tento­
rium. 36 .37 The preferential location at the back of the brain is
likely to relate to the origin of bleeding in the dural folds.
The dural sinuses and their tributaries have traditionally
been considered the origin for bilth-related subdural haem­
orrhage,J9 even in the absence offrank tentori al tearing. The
Figure 9.13 Chronic subdural haemorrhage. (a) Very vascular
falx has particularly prominent venous sinuses within it 40
membrane (M) adherent to deep surface of dura (D) . Note the
and bleeding here is associated with asphyxia. 41 lntratentor­
large numbers of capillaries at the junction wi th the dura and
ial bleeding in asphyxiated infants has been demonstrated
at the free edge (CD34 immunocytochemistry). (b) Subdural
by CT scan and autopsy correlative studies. 42
haemorrhage 4 months after head injury. There are collections of
Another consideration in the location of these bleeds is
fresh red cells and macrophages with pigmented cytoplasm. Peris'
gravitational redistribution when babies are nursed, and
positive material is seen within the macrophage cytoplasm and
indeed scanned, on their backs.43 Blood also tracks into the
free in the tissue, deposited on fibrous strands. These features
subdural space around the spinal cord and is frequently
indicate bleeding of several different ages (Perls' stain).
identified at autopsy as a crescentic collection over the pos­
terior cord at lower spinal levels.
Subdural haemorrhage is occasionally found overlying
an area of cerebral infarction in the term infant, usually in
Natural History of Infant Subdural Haemorrhage
the middle cerebral artery territory.44 The pathogenesis is
unknown, but Steinbok et al 44 considered the most likely The natural histOIY of subdural haemorrhage is that it
sequence to be cerebral infarction, with secondary subdural resolves by forming a granulating memb ra ne (see Fig. 9.13).
haemorrhage resulting from rupture outwards of a haemor­ This membrane contains wide capillaries that may rebleed.
rhagic infarct. Animal studies have suggested an alternative The characteristic histology of a healing subdural membrane
explanation: that subdural blood damages the underlying is a layer of fibroblasts containing fresh and old red cells,
brain. 45 pigmented macrophages and haemosiderin consistent with
Other rare causes of subdural haemorrhage include vas­ repeated episodes of rebleeding (the chronic active subdural
cular malform ations, meningitis and after neurosurgical haemorrhage) (Fig. 9.13). Birth-related subdural bleeds were
intervention. Very rarely, subdural haemorrhage compli­ seen to resolve on a single MR scan at 4 weeks in 9 cases. J7
cates metabolic diseases associated with brain atrophy, However the membranes may persist, Rogers et al 47 found
such as Menkes disease and glutaric aciduria in older evidence of chronic subdural membranes in up to
infants. 46 31 per cent of sudden infant death syndrome (SmS) case
176 I Pathology of neurological abnormality In early life
subjects not suspected of having chronic subdural haemor­
rhage in life.
Some infant subdural haemorrhages develop into
chronic fluid collections th at are visible on brain scans. The
mechanism is unknown but may be related to immaturity
of the arachnoid villi or rupture of the arachnoid mem­
brane, allowing leakage of CSF into the subdural space,43
osmotic accumulation of fluid or repeated small bleeds. 48
EXTRADURAL HAEMORRHAGE
Extradural haemorrhage can be seen without skull frac ture.
It may result form 'springing' of a suture, causing vascular
tearing without fracture. Away from sutures, it probably
results from impact related deformat ion of the thin infant
skull an d tearing of the dura from the skull without frac­
ture [ping pong fracture). Extradural and periosteal bl eed­
ing is sometimes seen on the edges of bones th at are
adjacent to widely sepa rated sutures where there has been
acute brain swelling and raised intracranial pressure.
SUBARACHNOID HAEMORRHAGE
Subarachnoid haemorrhage over the brain surface is very
common following hypoxic-ischaemic injury, particularly
in preterm infants. More extensive subarachnoid haemor­
rhage is seen at the base of the brain after intraventricular
haemorrhage, when blood tracks down from the ven tri cu­
lar system and leaks out through the exit foramina of the
fourth ventricle. Considerable amounts of blood may fill
the spaces beneath the brainstem in the cisterna magna.
Posterior Fossa Damage
As noted above, the majority of birth-related subdural
haemorrhage is found, at least in part, in the posterior fossa.
Another rare birth-related injury to the posterior fossa is
occipital osteodiastasis, characterized by displacement of the
squamous part of the occipital bone during delivery. It tends
to occur in breech, forceps or prolonged and difficult vagi­
nal deliveries.49 The compliant skull of the premature infant
allows displacement of the occipital bon es forwards and
upwards when under pressure during breech or instrumental
delivery. This can lea d to bleeding by compression of the
venous sinuses causing venous congestion and leakage or by
tearing of the superficial vessels of the cerebellum. Further­
more, direct pressure from the displaced bones wiH compress
and damage the cerebellar tissue. 50
Spinal Cord and Nerve Roots
Early studies have noted spinal cord and nerve root injury
after birth.28 ,5 1 Yates 51 described damage to the cervical
spine in almost one-half of neonatal autopsies, with nerve
•
.
I
.'
Figure 9.14 Axonal injury. Cervical nerve root showing many
axonal swellings stained with beta amyloid precursor protein. This
is a very sensitive method of identifying axonal damage in this
site (beta amyloid precursor protein immunocytochemistry).
root damage in one-third of these. He ascribed the damage
in most cases to vertebral artery injury.
Spinal nerve root tears and haemorrhages and spinal cord
injury were seen in 10 per cent of infants, due to stretching
or tearing during delivery. Cument incidence is unknown as
the cord is not frequently removed and examined. Infants
with very severe brain swelling and tonsillar herniation
through the foram en magnum will have considerable dis­
ruption of the cervica l cord and nerve roots, which may,
itself, cause local nerve root damage.
Immunohistochemistry with antibodies to 0APP is
extremely helpful in demonstrating nelve root injury (Fig.
9.14) but is not sp ec ific for the cause of injury. Distinction of
hypoxic-ischaemic damage from tru e axonal tearing may be
difficult, particularly as the two almost invariably coexist.
Towbin 28 noted epidural haemorrhage as the most fre­
quent injury in his series of neonatal post-moliems. However,
small amounts of epidural bleeding are a common finding in
infants who have died from non-traumatic causes, suggest­
ing that in some cases they may represent post-mOIiem
artefact - perhaps related to method of cord remova1. 52
STROKE IN THE DEVELOPING BRAIN
Unilateral cerebral artery infarction is common, occurring
in 1 in 4000 term infants. The middle cerebral artery is
most commonly involved. The damage is due to failure of

blood supply through a m ajo r v essel, which may be due to
its blockage by a b loo d clot or tissue fragment, for exam­
ple fronl the placenta, or by externa l compression. Ex tern al
co mp ression is rare. The cause of vascular obstruction is
not often demonstrated at autopsy.
Stroke may occur in the an ten atal or perinatal periods
period or at the time of birth, and the timing of the damage
may be reflected in the clinical presentation. Ba bies injured
in the perin atal perio d tend to present early with neo natal
seizures, whereas infants damaged in utero present later
and develop hemiplegia and intractable epilepsy after sev­
eral months of pos tna tal life. 53 However, many strokes are
clinica lly sile nt at the time w hen they occur.
The aetiology is obscure and prob ably multifactoriaP4.55
Overall, 30 per cent of neonatal stroke is due to v eno us sinus
thrombosis. 56 Factors associated with neonatal stroke
include: blood clo tting disord ers in fec tions, placental dys­
fun ction, growth retardation, pre-eclampsia, tw in pregnancy,
peripartum asphyxia and maternal thrombophilia with
thrombotic lesions in the placen ta. 54.57 ,58 Hypoxic-ischaemic
injlllY is the most common association.59
Cerebral Sinovenous Thrombosis
Venous thrombosis in sudden deat h is probably under
reported and not always sought. Ante-mortem clo ts are
distinguis hed by their dry or lamina ted appearance, an d
are often adherent at some point. There is a high incidence
of sinovenous thromb osis in the first month of life. Dehy­
dration, sepsis, polycythaemia a nd protein C deficiency
have been implicated as causes but are no t always pres­
ent. 5O ,61 In to ta l, 84 per cent of neon ates with sinove no us
thromb osis have perinatal complications, of w hi ch hypoxic
encephalopathy is the most common. 56 ,62
METABOLIC DISORDERS
Hypoglycaernia
Cortical an d subcortical damage in the parieto-o ccipital
region has been described by MRl studies of infants who suf­
fered hypoglycaemia in the neonatal period ,6),64 but studies
of the neuropa tho logy of neonatal hypo gJycaemi a are few. 65
There is little informa tio n re garding the effects of
maternal diabetes on fetal brain development, although
white matter damage is described and considered to result
from delayed brain matura tion 66
Mitochondrial Diseases
Impairment of mitochondrial fun ction can lead to sudd en
death. Babies with abn ormalities of mitochon drial DNA
Infections I 177
usually h ave raised lactate in their blood and CSF, which
may be exacerbated on exercise or when stressed by inter­
curre nt infections. Clin ical presentation includes poor
feeding, vomiting and exercis e-induced lactic acidosis, res­
piratory deficiency and lethal ap noea. 67
Autopsy examination of skeletal muscle may s how typ­
ical ragged red fibres with Gomori's tr ichrome stain, abno r­
mally increased succinate dehydrogenase ac tivity and
cytochrome oxidase deficiency in many fibres. However,
th ese changes a re less frequ ent in infant than in ad ult cases
and diagnosis depends on mitoch ondrial DNA an alys is.
Brain examination may show a number of abnorma lities.
In Leigh's syndrome there are typ icall y symmetrical lesions
of tissue rarefaction, gliosis and capiUary proliferation in
the basal ganglia and brainstem. In other cases, for example
pyruvate dehydrogenase deficiency, there may be migration
disturbances, congen ital cerebral at rophy an d ventricular
dilatation. 67
Lipid Storage Diseases
These diseases are rarely a cause of sudden death. Presenta­
tion is usually between 6 and 24 months, and includes vo m­
iting or coma triggered by fasting. The most commo n lethal
fo rm is long-chain methyl CoA deficiency, bu t medium- a nd
mul tiple-ch ain CoA deficiencies also occur. Skeletal muscle
and li ver sho uld be sampled and snap-frozen for histochem­
ical study. In froze n sections of skel eta l muscle, abnormally
stored fat forms multiple large globules within muscle fibres;
this is in contrast w ith normal neonatal muscle, which usu ­
ally contains very li ttle fa t. Fat may also be excessive in the
liver and other organs.
INFECTIONS
Bacterial infections of the fe ta l brain are extremely rare.
The most common organisms to damage the fetal brain
by direct invasion are toxoplasmosis, rubella, CMV, herpes
simpl ex and li sterios is. Human imm uno defic ien cy v irus
(HN) an d herpes simplex v irus (HSV) may be tra nsmi tted
to the infa nt during delively. Viral infections may result in
calcification and cystic damage . Listeria causes granulo­
mata aro und blood vessels in t he meninges and ch oroid
plexus. CMV may be asso ciated with polymicrogyria (Fig.
9.1 5) , possibly due to its a ffini ty for the end oth elial cell s,
causin g interferen ce w ith vascular supply to the cortex in
its later stages of development.
Bilateral, multifocal cysti c lesions an d calcifica tion ,
haemorrhage and cerebral oedema have been described on
MR and CT sca ns of babies with v iral encephalitis.22 ,27
Extracranial manifestations, for example limb hypoplasia,
cataracts, micro-ophthalmia, choreoretinitis, organ calcifi­
cation o r skin scars supp ort the diagnosis of a congeni tal
viral infection.
178 I Pathology of neurological abnormality in early life
Figure 9.15 Polymicrogyria. Fixed brain of fetus with
cytomegalovirus. There are areas of polymicrogyria in the middle
regions bilaterally, an appearance which is simi lar to perisylvian
polymicrogyria of genetic or ischaem ic origin. The distribution abnormalities in congenita l cytomegalovinls infection.
can be misleading.
REFERENCES
Rutherford MA, Pennock JM, Cowan FM et at. Does the brain
regenerate after perinatal infarction? Europ J Paediotr Neural
1997; 1:13-17.
2 Nelson KB. Can we prevent cerebral palsy? N El1gl .J Med
2003; 349:1 765- 9.
3 Keeling JW. Feta l lind Neonatal Pathology, 2nd edn. London:
Springer-Verlag, 1993.
4 Squier M, Chamberlain P, Zaiwalla Z et al. Fi ve cases of bra in
injury follo win g amniocentesis in mid-term pregnancy. Deu
Med Child Neuro/2000; 42:554-60.
5 Geddes JF. What's new in the dia gnosis of head inj ury? J Ciill
Pothol 1997; 50: 271 -4.
6 Hortobagyi T, Wise S, Hunt N, Cary N et al. Traumatic axonal
damage in the brain can be detected using beta-APP
immun ohistoc hemistry within 35 min after hea d injUly to
hum an adu lts. Nellropati1ol Appl Neurobi ol2007; 33:226-37.
7 Oehmichen M, Meissner C, Schmidt Vet al. Axo nal injury - a
diagnostic too l in forensic neuropathology? A review.
Forensic Sci Inr 1998; 95:67-83.
8 Hannan AJ, Servotte S, Katsnelson A et al. Characterization
of nodul ar neuronal heterotopia in ch ildren. Brai l7 1999;
122:21.9- 33.
9 Barkovich AJ, Kuzniecky RI, Ja ckso n GD et al. Classification
system for malformations of cOl1ical develop ment: update
2001. Neurology 2001; 57: 2168-70.
10 Yakovlev PI, Wadsworth RC. Schizencep halies: a study of the
congenita l clefts in the cerebral mantle. l. Clefts with fused
lips. J Neuropathol Erp Neurol 1946; 5: 116-30.
11 Barkovich AJ, Kjos BO. Schizencephaly: correlation of clinical
findings with MR characteristics. AJNR Am J Neuroradiol
'1992; 13:85-94.
12 Bar1h PG. Migrational disord ers of the brain. Curl' Opin
Neuro! Neurosurg 1992; 5:339-4 3.
13 Ja nsen A, Andermann E. Genet ics of the polymicrogy ria
syndro mes. 4. J Med Genet 2005; 42 :369-78 .
14 McQuillen PS, Barko vich AJ, Hamrick SE et a l. Temporal and
anatomic risk profile of brain injury with neo natal repair of
congenital heart defects. Stroke 2007; 38 (Suppl.):736-41.
15 Kinney HC , Panigrahy A, New burger JW et al.
Hypoxic-ischemic brain il'\iury in infants with congenital
heart disease dying after cardiac surgery. Acta Neuropathol
(Berl) 2005; 110:563-78.
16 Krageloh-lVlann I, Toft P, Lunding J, Andresen Jet al. Brain
lesions in preterms: origin, consequenc es and compensation.
Acta Paediatr 1999; 88 :897 - 908.
17 Krageloh-Mann I, Helber A, Mad er I et al. Bilateral lesions of
thalamus and basal gangl ia: origin and outcome. Del! Med
Child Neural 2002; 44:477-84.
18 Peebles DM, Wyatt JS. Sy nergy between antenatal exposure
to infection and intrapal1um events in ca usation of perinatal
brain injury at term. BJOG 2002; 109:737-9.
19 Redline RW. Severe fetal placental vascular lesions in term
infants with neurologic imp airm ent. I. Am J Obstet Gynecol
2005; 192:45 2-7.
20 Gilles FH, Leviton A, Dooling ED . The Delle/oping Human
Brain: Growth and Epidemiolog ic Neuropathology. Boston:
John Wright, 198 3.
21 de Vries LS, Gunardi H, Barth PG et al. The spectrum of
cranial ultra so und and magnetic resonance imaging
Neuropediarrics 2004; 35:113-19.
22 Kurtz J, Anslo w P. Infantil e herpes simplex encephalitis:
diagnostic featu res and differentiation from non- accidental
injury. J !njcct2003 ; 46:12-16.
23 Pasternak JF. Hypox ic-ischemic brain dama ge in the lerm
in fa nt. Lesso ns from th e laboratory. Pediatr Ciin North Am
1993; 40:1061-72 .
24 Sie LT, van der Knaap MS, Oosting J et al. MR pattern s of
hypoxic-i schem ic brain damage after prenatal , perinatal or
postnatal asphyxia. Neuropediatrics 2000; 31: 128-36.
25 Naeye RL, Lin HM. Determination of the timing of fetal brain
damage from hypoxemia-ischemia. Am J Obstet Gyneeol
2001 ; 184:217-24.
26 Myers RE. Two patterns of perinatal brain damag e and their
conditio ns of occurrence. Am J Obstet GYllecol 1972;
112:246-7 6.
27 Rutherford M. Magnetic resonance imaging of injury to the
imm atu re brain. In Squier W (eeJ.) Acquired Damage to the
Developing Brain: Timing and Causation. London: Arnold ,
2002, pp. 166-92.
28 Towbin A. Central nervous system damage in the human fetus
and newborn infant. fvlechanical and hypoxic injury incurred
in the fetal-neonatal period. Am J Dis Child 1970;
119 :529-42.
29 Amar AP, Aryan HE, Meltzer HS, Levy ML. Neonatal subgaleal
hematoma causing brain compress ion : repol1 of
two cases and review of the literature. Neurosurgery 2003;
52:1470-4.
30 Dupuis 0 , Silveira R, Dup ont C et at. Comparison of
'instrument-associated' and 'spo nta neous' obstetric depressed
sku ll frac tures in a cohort of 68 neo nates. Am J Obstet
Gyneeol 2005 ; 192: 165-70.
31 Poll ina J, Dias MS, Li V, Kachurek D, Arbesman M. Cranial
birt h injuri es in term newborn infants. Pediat/' Neurosu rg
2001; 35:1 13-19.
32 Towner D, Castro MA, Eby-Wilkens E, Gilbert VVM. Effect of
mode of delivery in nulliparous women on neonatal
intracranial injury. N Engl J Med 1999; 341:1709-14.
33 Chamnanvanakij S, Rollins N, Perlman JM. Subdural
hematoma in term infants. Pediotr Neural 2002; 26:301-4.

34 O'Mahoney F, Settatree R, Platt C, Johnson R. Review of
singleton fetal and neonatal deaths associated with cranial
trauma and cephalic delivery during a national intrapartum­
related confidential enquiry. BJOG 2005; 112:619-26.
35 Geddes JF, Tasker RC, Hackshaw AK et a\. Dural haemorrhage
in non-traumatic infant deaths: does it explain the bleeding
in 'shaken baby syndrome'? Neuropathol Appl Neurobiol
2003; 29:14-22.
36 Looney CB, Smith JK , Merck LH et a\. Intracranial
hemorrhage in asymptomatic neonates: prevalence on MR
images and relationship to obstetric and neonatal risk factors.
Radiology 2007; 242:535-41.
37 Whitby EH, Griffiths PD, Rutter S et a\. Frequency and natural
history of subdural haemorrhages in babies and relation to
obstetric factors. Lancet 2004; 363 :846- 51.
38 Tavani F, Zimmerman RA, Clancy RR et al. Incidental
intracranial hemorrhage after uncomplicated biI1h: MRI
before and after neonatal heart surgety. Neuroradiology 2003;
45:253-8.
39 Volpe JJ. Intracranial haemorrhage. In Neurology of the
Newborn, 3rd edn. Philadelphia, PA: WE Saunders, 1995,
pp. 377-8.
40 Tubbs RS, Louis RG , Acakpo-Satchivi L, Salter EG. Anatomy
of the falcine venous plexus. J Neurosurg 2007; 107:155-7.
41 Friede RL. Hemorrhages in asphyxiated premature infants. In
Friede R (ed.) Developmental Neuropathology. Gottingen:
Springer-Verlag, 1989, pp. 44-58.
42 Kibayashi K, Shojo H, Sumida T. Dural hemorrhage of the
tentorium on postmortem cranial computed tomographic
scans in children . Forensic Sci lilt 2005; 154:206-9.
43 Vinchon M, Noizet 0 , Defoort-Dhellemmes S et al. Infantile
subdural hematomas due to traffic accidents. Pediatr
Neurosurg 2002; 37:245-53.
44 Steinbok P, Haw CS, Cochrane DD, Kestle JR. Acute subdural
hematoma associated with cerebral infarction in the full-term
neonate. Pediatr Neurosurg 1995; 23 :206-15.
45 Durham SR, Duhaime AC. Basic science; maturation­
dependent response of the immature brain to experimental
subdural hematoma 1. J Neurotrauma 2007; 24:5-14.
46 Kemp AM. Investigating subdural haemorrhage in infants.
Arch Dis Child 2002; 86:98-102.
47 Rogers CB, Itabashi HH, Tomiyasu U, Heuser ET. Subdural
neomembranes and sudden infant death syndrome. J Forensic
Sci 1998; 43:375-6.
48 Markwalder TM. Chronic subdural hematomas: a review.
J Neurosurg 1981; 54:637-45.
49 Currarino G. Occipital osteodiastasis: presentation of four cases
and review of the literature. Pediatr Radio12000; 30:823-9 .
50 Volpe JJ. Intracerebellar haemorrhage. In Neurology of the
Newborn, 3rd edn. Philadelphia, PA: WE Saunders, 1995,
pp. 384-9.
References I 179
51 Yates PO. Birth trauma to the vertebral at1eries. Arch Dis
Child 1959; 34:436-41.
52 Rutty GN, Squier WM, Padfield CJ. Epidural haemorrhage of
the cervical spinal cord: a post-mortem artefact? Neuropathol
Appl Neurobiol 2005; 31 :247-57.
53 Squier W, Salisbury H, Sisodiya S. Stroke in the developing
brain and intractable epilepsy: effect of timing on hippocampal
sclerosis. Dev Med Child Neurol 2003; 45:580-5.
54 Arias F, Romero R, Joist H, Kraus FT. Thrombophilia: a
mechanism of disease in women with adverse pregnancy
outcome and thrombotic lesions in the placenta. J Maternal
Fetal Med 1998; 7:277-86.
55 Mercuri E. Timing and aetiology of neonatal cerebral
infarction (Letter). Paediatrics 2000; 106 :615-16.
56 DeVeber G, Andrew M, Adams C et al. Cerebral sinovenous
thrombosis in children. N Engl J Med 2001; 345:417-23.
57 Mercuri E, Cowan F, Gupte G et a\. Prothrombotic disorders
and abnormal neurodevelopmental outcome in infants with
neonatal cerebral infarction. Pediatrics 2001; 107: 1400-4.
58 Marrett S, Lardennois C, Mercier A et a\. Fetal and neonatal
cerebral infarcts. Bioi Neonate 2001; 79:236-40.
59 Evans D, Levene M. MRI Imaging of injury to the immature
brain. In Squier W (ed.) Acquired Damage to the Developing
Brain; Timing and Causation. London: Arnold, 2002.
60 Shevell MI, Silver K, O'Gorman AM et a\. Neonatal dural sinus
thrombosis. Pediatr Neurol 1989; 5:161-5.
61 Rivkin MJ, Anderson ML, Kaye EM. Neonatal idiopathic
cerebral venous thrombosis: an unrecognized cause of
transient seizures or lethargy. Ann Neurol 1992; 32:51-6.
62 Fitzgerald KC, Williams LS, Garg BP et a\. Cerebral sinovenous
thrombosis in the neonate. Arch Neurol 2006; 63 :405-9.
63 Traill Z, Squier M, Anslow P. Brain imaging in neonatal
hypoglycaemia. Arch Dis Child Fetal Neonatal Ed 1998;
79:F145-7.
64 Murakami Y, Yamashita Y, Matsuishi T et at. Cranial MRI of
neurologically impaired children suffering from neonatal
hypoglycaemia. Pediatr Radiol 1999; 29:23-7.
65 Rossiter JP, Anderson LL, Yang F, Cole GM. Caspase-3
activation and caspase-like proteolytic activity in human
perinatal hypoxic-ischemic brain injury. Acta Neuropathol
(Berl) 2002; 103:66-73 .
66 Lynch N, Roland EH, Poskitt K et a\. Gestational diabetes
mellitus: risk of newborn brain injury. Dev Med Child Neurol
2007; 49(Supp\. 108):7.
67 Brown GK, Squier MV. Neuropathology and pathogenesis
of mitochondrial diseases. J Inherit Metab Dis 1996;
19:553-72.
I CHAPTER 10 I
FETAL AND PERINATAL DEATH
Jean W Keeling
Introduction 180
Defi n itions 180
The law 181
Background information 182
Concealed pregnancy 182
Unattended delivery 182
Was the baby born alive? 183
Is the baby of sufficient maturity to survive? 187
INTRODUCTION
been translated as 'no risk'. Thus, many consider that there
There are several circumstances in relation to fetal deaths,
still births and deaths in early life when an investigation is
likely to faU within the rem it of a coroner, a procurator fis­
calor a medical examiner. These include concealment of
pregnancy; unattended delivery, with or without abandon­
ment of the newborn infant; and suspected infanticide. The
size of the problem is difficult to assess, although the num­
ber of infanticides (neonaticides) has fallen in some jurisdic­
tions, as has the number of maternal fatalities attributable to
criminal abortion, probably for similar reasons.i Neverthe­
less, abandoned fetuses and infants are still discovered in a
variety of circumstances. Fetal death following maternal
injury, sudden fetal death and those fetal deaths in which the
attendant midwife or obstetrician is unable to complete a
still birth certificate also require formal investigation in
some jurisdictions. 2
A further circumstance in which it has become increas­
ingly likely that the pathologist's report will acq uire
medicolegal significance is following intrapartum still bilth
or early neonatal death (perinatal deaths) after hospital
delivery. The perinatal mortality rate has fallen to lo w
levels in many countries, particularly Western Europe,
North America, and Australia and New Zealand; it was less
than 8 in 1000 total births in the UK in 2002 with an intra­
partum death rate of 0.62/1000 total births 3 This low risk of
labour-related mortality has, in the public's perception,
-
Is there evidence of prolonged or difficult labour? 187
Are there any significant injuries? 188
Fetal death following maternal injury 190
Is there a natural cause for death? 193
Can I give a cause of death? 193
Should the intrapartum still birth be a 194
medicolegal autopsy?
References 195
must be fault attached to any death occurring during or
shortly after labo ur.
DEFINITIONS
Definitions relating to death in early life differ in different
countries and in different jurisdictions. This section lists some
important and relevant definitions and draws attention
to different usages of the same term. These are related to both
offences that may have been committed and charges arising
therefrom. It is important to be aware of them when
comparing statistics or papers from different countries.
• Fetal death is defined as death before 22 completed
weeks of gestation (World Health Organiza tion [WHO])
or before 24 completed weeks of gestation (UK), when
the conceptus exhibits no sign of life after complete
separation from the mother. Fonnal registration of feta l
death is not required, although many maternity units
issue a certificate of fetal death, both as an
acknowledgement of existence4 and to facilitate
funeral arra ngeme nts in accordance with parental
wishes. Concealment of the death of a fetus is not
an offence (in England).
• Still birth is death after 22 completed weeks of
gestation (WHO) or after 24 completed weeks of
gestation (UK) of a conceptus who exhibits no signs of
---------------~
....

life after complete separation from the mother. A
certificate of still birth must be delivered to the
appropriate registration authority. It can be completed
by either a midwife or a medical practitioner who was
present at the birth, or who examined the body after
delivery, and is able to certify that the child was not
born alive. If there is any doubt, the procurator fiscal,
coroner or medical examiner should be informed.
• Neonatal death is death following live birth, i.e.
exhibition of signs of life after complete separation
from the mother, irrespective of the length of gestation,
in the first 28 days (UK) or 30 days (USA) of life. Both
birth and death certificates are required.
• Signs of life, from a medical standpoint, are heartbeat,
spontaneous respiration and voluntary movement.
Only one of these is necessary to fulfil the definition
of live birth.
• Complete separation from the mother is defined as
complete delivery of all parts of the baby (fetus) from
the body of the mother. Severance of the umbilical
cord or delivery of the placenta has no part in
definitions relating to completeness of separation from
the mother or to live or still birth. The concept of what
constitutes a separate existence is both important and
difficult; cardiac function and the ability to breathe
have been variously considered the gold standard. 5
However, it is not only the initiation of respiration, but
the ability to sustain respiration, which is essential for
survival,6 that matters.
• Abortion is the termination of pregnancy before the
legal limit of viability. Unqualified, the term
encompasses natural pregnancy loss as well as
del iberate termination of pregnancy.
• Induced abortion is the deliberate termination of a
pregnancy before the legal limit of viability whether by
medical or surgical means.
• Child destruction is the destruction of the life of a child
who is capable of being born alive, i.e. after the legal
limit of viability.
• Infanticide (England and Wales) occurs when a mother
(acting alone) causes the of death of a child who is less
than 12 completed months of age by a wilful act or
omission.
• Infanticide (USA and elsewhere) is used more generally
to mean causing the death of a child in the first week
of life.
• Neonaticide has no legal definition but is used
variously to mean the unnatural death of a liveborn
child within 24 hours of birth 7 or within the neonatal
period.
THE LAW
The law, in relation to the unnatural death of the fetus and
newly born infant, is considered in detail by Bowen,8
The law I 181
Kellett 5 and Mason. 6 They review the effects of sequential
changes in legislation in this area in terms of the definition
of particular offences and restrictions on medical practice.
They set out differences in legislation in Scotland com­
pared with England and Wales. The law will be considered
only briefly here.
The Offence of Procurement of Abortion
This is defined in the Offences Against the Person Act 1861,
Sections 58 and 59. This Act is still extant. The Abortion
Act 1967 serves mainly to define the term 'unlawfully' in
the 1861 Act.
Child Destruction
Child destruction is an offence under the Infant Life
(Preservation) Act J 929, except when performed in pur­
suance of 'preserving the mother's life'. It applies to the
killing of a viable fetus both before and during labour, thus
offering protection to doctors faced with performing cran­
iotomy during labour for cephalo-pelvic disproportion.
This procedure is rarely performed now in developed coun­
tries because of the infrequency of contracted pelvises and
the availability of prenatal diagnosis and Caesarean sec­
tion. The Infant Life (Preservation) Act of ]929 does not
apply in Scotland.
Concealment of Birth
Concealment of birth, originally a criminal offence, is leg­
islated in the Offences Against the Person Act 1861, Section
60. It encompasses both live and still births. It is not neces­
sary to show how death occurred, or even to produce the
infant's body, provided that the mother can be shown to
have delivered a baby in the recent past. It thus dispenses
with the thorny problem of demonstration of a separate
existence and is a less serious charge than infanticide.
In Scotland, the Concealment of Birth (Scotland) Act
1809 additionally requires that the mother conceals both
pregnancy and delivery, and that the child was viable. As
in England and Wales, it is not necessary to produce a body
for successful prosecution.
Infanticide
This offence was introduced in England and Wales by the
Infanticide Act 1922 as an alternative to a charge of mur­
der or manslaughter. It related solely to death of an infant
caused by the mother acting alone, because of disturbance
of her mind following birth. It applied originally only to
the death of a newly born child. This was amended by the
182 I Fetal and perinatal death

Infanticide Act 1938 which specifIes 'any wilful act or

omission , ca uses the death of her child being less than 12
months of age' and 'the balance of her mind is disturbed '
'not havin g fully re covered from the effects of giving birth'
or 'lactation'. This is, in effect, a charge of manslaughte r. In
Scotland, where there is no similar act, a mother who kills
her infant under similar circumstances is charged with cul­
pable homicide.

BACKGROUND INFORMATION

Before commencing post-monem examination, It IS most

important to be fully informed about the circumstances of the
case. A detailed police report, including how the baby was
found and by whom, together with details of the JOCLIS, is
essential for bilihs and deaths occurring outside hospital. In
some circumstances photographs of the scene may provide
information to suggest modification of necropsy techniques.
However, important information may be missed if a patholo­
gist does not attend the locus, and so such a vis it should be
encouraged. When the mother is known, as much informa­
tion as possible about the pregnancy should be obtained from
her genera l practitioner and hospital records.

CONCEALED PREGNANCY

There are a number of important areas that must be co nsid­

ered in relation to concealed pregnancy, unattended delivery
and abandonment which are peliinent to charges that may
be brought. It is important, then, that these are addressed as
precisely as possib le. These include fetal maturity, time of
death in relation to delivery, whether the baby was livebom,
cause of death and evidence of trauma.
Women who conceal their pregnancy have many fea­
tures in common. 9 In the UK and USA, they are frequently,
but not invariab ly, experiencing a first pregnancy. Th e
mother may never be identifi ed, although some seek medical
assistance because of retained placenta, postpartum haemor­
rhage or infection. They are usually young, of low educa­
tional attainment and g ive birth unaided, or at best with
inexpert assistance. The neonatal monality follo wing these
pregnancies is more than twice that of hospital deliveries;lO
the still birth rate is likely to be much higher. Labour may be
protracted and there is usually denial of live birth . Efforts are
made to concea l the body (Figs 10.1 and 10.2).
It is, perhaps, not surprising that concealment of preg­
nancy and delivery continue in conservative societies
where sex ed uca tion is su ppressed . II It was not unco mmon
in UK rural communities within the last two generations 12
and is still occasionally encountered in Europ e following
the chance finding of fetal remains. 13 - IS
Serial concealment of pregnancy and infanticide is not
uncommon in Japan, despite the ready availability of abor­
tion . The motive seems to be largely economic. 16 ,17 The
Figure 10.1 Conce aled pregnancy and delive ry. Mature baby
wrapped in bedding and discovered in a cupboard. (a) The skin is
covered in blood and mecon ium. (b) The umbilica l cord is t orn
close to its fetal insertion.
practice is not confined to single women alth ough perpe­
trators are frequently single. 18 Usuall y, husbands of mar­
ried perpetrators deny knowledge of the pregnancies. 19
UNATIENDED DELIVERY
Legislation in the UK restricts thos e persons who may
attend a woman in childbirth to practising midwives and
medical practitioners. 2o These indiv iduals must be regis­
tered with their appropriate professional body and be able
to demonstrate both that they have underta ken appropriate
training and that they are fit to practice.
In general, unattended birth can occur in two different
circumsta nces and the findin gs on examin ation of the baby
mi ght be expected to be different in each case. Either
birth may follow a concealed pregnancy, or emergency
professional help is unavailable because labour is either
short or has not been perceived in time and delivery is
 Was the baby born alive? I 183

(a)

Figure 10.2 Mummified remains of a baby found behind

panelling during renovations. Body length and amount of
subcutaneous fat indicate that the baby was mature.

rapid and occurs before appropriate help is availab le

(Fig. 10.3). In the case of mature baby, the mother is
often of high parity and has usually, but not invariably,
registered for antenatal care and planned confinement.
In unattended delivery in general, perinata l mortality is
higher than in the case of local hospital delivery or planned
home birth .21

Delivery into Toilets

AJleged spontaneous delivery of a mature baby into a lava­

LOry pan shou ld be viewed with circumspection. 22 Although
perinea l sensations preceding delivery may be confusing,
only the rapid delivery of previable fetuses in this way is
redible. The subsequent actions of the mother and other
persons present in the domicile are important. Prompt and
eta iled documentation of these is required. Disposal of Figure 10.3 Precipitate delivery onto the bathroom floor.
unwanted, liveborn infants in toilets is not uncommon in (a) A baby of about 26 weeks' gestation. The skin is extensively
so me parts of Japan. 17 - 19 reddened as a resu lt of hypothermia. (b) A section of skin
showing extensive haemorrhage in the deep dermis.

WAS THE BABY BORN ALIVE? findings alone, but there are some findings that can answer
this question absolutely. When there is doubt, this must
t may not be possible to distinguish with absolute be clearly expressed and the likelihood of still birth
<:enainty between live birth and still birth from pathological promoted.
184 I Fetal and perinatal death
Figure 10.4 The earliest sign of maceration, skin slippage, over
an extremity.
Figure 10.5 Early maceration with focal epidermal loss over
bony prominences. Drying has resulted in dark-red discolouration,
which might be confused with inflicted injury.
MACERATION
Changes in the skin due to maceration preclude live birth.
Sequential changes owing to maceration are set out in Table
8.5, p. 159. Cutaneous maceration is unlikely to be present
until about six hours after death, when focal loss of superfi­
cial epidermis is seen over extremities and bony prominences
(Figs 10.4 and 10.5). This loss of epidermis (skin slippage)
becomes more widespread with the increasing length of time
that the dead fetus spends in utero. Maceration of face, back
or trunk develops in about 12 hours. When more than one
Figure 10.6 Dark-red discolouration of the fetal insertion of the
umbilical cord is a sign of maceration which takes about six hours
to develop.
Figure 10.7 A 4-day-old neonate. Desiccation of the umbilical
cord is an indication of live birth with several days' survival.
(Courtesy of Professor N Mcintosh, Edinburgh.)
area is involved, we can estimate that death occurred about
18 hours before delivery.23 These times are only approximate.
Estimates of the time of death may be modified by the
appearance of the umbilical cord and internal organs. It is
important to remember that cutaneous maceration is retarded
in the presence of intra-uterine growth retardation (fUGR)
and accelerated in the presence of oedema (fetal hydrops).
UMBILICAL CORD INSERTION
Deep-red discolouration of the umbilical cord at its fetal
insertion is an indication that fetal death has taken place at
least six hours before delivery (Fig. 10.6). Loss of Wharton's
jelly from this part of the cord, often accompanied by
marked twisting of the cord, takes longer to develop and is
also a marker of still birth. Desiccation or separation of
the umbilical cord stump indicates live birth with survival
for a few days. The cord stump progressively desiccates
(Fig. 10.7) over 3 or 4 days and then separates from the

umbilicus. A vital reaction is apparent w ithin the
cord stump w ithin 2- 3 hours of birth on microscopic
exam ination; this indicates live birth with, at least, brief
survival.
STOMACH
Mil k curd (or other foreign ma terial) in the sto mac h indi­
cates live birth with survival. Care must be taken no t to
confuse mucoid secretions, which can be turb id in the
matu re fetus, particularly in the presence of ascending
infection w hen the fet us has swallowed infected liquor
am ni i, w ith milk, which is usually flo ccu lent. Smears of
gastri c co ntents stained w ith haematoxylin an d eosin,
Gram 's method for bacteria l organisms and oil red a for fat
will often clarify any uncertainty.
Rad iographic examination may demonst rate gas in th e
stomach and intestines in liveborn infants. This, however,
is not abso lute proof of live birth as contamination with
gas- form in g organisms by swallowing infected liquor
am nii ca n result in a gast ric gas bubble.
Meconium may be present in the stomach when there
has been acute hypo xic stress to the fetus either just befo re
or during labour. Its presence or absence does not help to
distinguish between live birth an d still birth.
STOOLS
Before birth, the large intestine contai ns meconium, a mix­
ture of amniotic squames, intestinal secretions, desqu a­
ma ted intestinal epithelium and bile acids. 24 Many babies
defecate in the first few minutes after birth ,25 although
many mature ba bies who are subjected to hy pox ic stress
before or during labour w ill often defecate before birth.
Meconium is dark green and sticky, but sterile. Over the first
few days after birth, a mixture of meconium and normal
stool is passed. Breast-fed babies have very soft, acidic,
mustard-yellow stool; bottle fed babies have a firmer, paler
stool, w hich is less acid ic. 25
Colonization of the intestine wit h bacterial organisms
takes place rapid ly. There is geographic variation . More
than one-hal f of vaginally delivered infants in Pakistan
have positive cu ltures for faec al organisms at 1 day of
age,26 but the proportio n in Swedish infan ts was found to
be much lower. Stool bacteria l flora was well estab lished at
4 days of age in babies born in the UKY The stool of milk­
fed newborns wi ll cont a in abundant fat globules; bacterial
organisms will be seen diffusely through the stool sample.
There are di fferences in the bacterial flora of breast-fed,
compared with form ula-fed, bab ies. 26.27
UPPER AIRWAYS
Plugging of the upper airways by thick meconium pre­
cludes estab lishment of respirati on. Foreign material, for
Was the baby born alive? I 185
examp le tissues, cotton w ool or fabric, is occasionally
fo und in the mouth or pha ry nx in deliberate a ttempted a ir­
ways obstruction. Abso rbent material is so metimes found
in or over the mouth, presumably to take up unwanted
fluid secretions, when other findings suggest still birth. The
presence of frothy fluid in the upp er a irways, which has
been fo und to withstand freezing of the infant's body, 28 is
a lso an indica tion of li ve birth.
LUNG
Gross Appearance
When the lun g has been fully expanded by establishment of
respiration in a mature baby, it w ill be pink and crepita nt.
It is virtually impossible to expel a ll of the air by external
pressure. The lungs are unexpanded in still birth. They are
smaller, darker in colour and occupy the posterior, paraver­
tebral parts of the thoracic cavity. Knight 29 found the naked­
eye appearance and feel of lu ngs a better dis criminant
between live a nd still birth than the lung flotati on test.
Petech ial hae morrhages are often visible throu gh the
pleural surface, but they may be confined to the pleural
fissures. They are considered a sign of acu te hypoxia .
Petechial h aemo rrhages are p art icularly numerous when
death is du e to placental abruption.3o They lose definitio n
and then fade afte r a fe w days, both in the fetus reta ined
in utero and in the delivered baby.
Flotation Test
The use of the property of lu ngs to float in water (or
buffered formalin) as a determinant of live birth is fra ught
w ith diffi cul ty. 29 It is un w ise to rely on it as the only deter­
minant of live birth even when so me or any of the pub­
lished modifications, which all egedly impro ve reliability,
are introdu ced. It may be falsely positive because of putre­
faction, even to a minor degree. When putrefactio n has
occurred, gas may be in terstitial or both in terst itial and
within terminal ai r spaces (see Fig. IO .Sc and d). The value
of the flotatio n test is negated by mouth-to-mouth or other
p ositive-press ure ventilation J
The flotation test may be falsely nega tive in the pres­
ence of well-developed hyaline membrane disease, a fre­
quent find in g in the livebo rn preterm baby as well as
liveb orn babies born at term who have experienced severe
hypoxic stress, or follo w ing ventilation with 100 per cent
oxygen. Despite these dra w backs, Moar 31 eva lu ated lung
flotation along with liver flotation, lun g compression, the
gross appea rance of the lungs and histol ogica l examina­
tion. He fo und it of valu e in carefully controll ed co nditions
but stressed the need for corroborative tests.
Histological Appearance
Both the presence of large amou nts of squamous debris and
mu cus fro m the amnioti c fluid and the presence of meconium
186 I Fetal and perinatal death

.':....
Figure 10.8 Patterns of lung distension. (a) Antepartum still birth at term: there is partial expansion of terminal air spaces up to the
pleural surface as a result of hypoxia-induced inspiratory efforts. There is meconium aspiration. (b) Delivery after motor vehicle collision
(MVC) at 34 weeks' gestation, survival for 13 hours. There is overdistension of term inal bronchioles and alveolar ducts with collapse of
intervening lung parenchyma; these changes are seen in early respiratory distress syndrome. (c) Irregu lar, focally marked, distension of air
spaces and interstit ial tissue as a result of putrefaction, concealed delivery. (d) Putrefaction has resulted in overdi stension of terminal air
spaces and accumulation of gas in pleural connective tissue.

plugging of large intrapulmonalY airways and groups of ter­
minal air spaces indicate severe hypoxic stress before birth.
The former are readily demonstrated using Alcian
blue/phloxine staining, promoted by Attwood 32 to demon­
strate amniotic fluid embolism in the maternal pulmonary
vasculature, or an immunohistological marker against high­
molecular-weight cytokeratin, such as CK1. 33
When the fetus has made strong and repeated inspiratory
efforts before birth, partial expansion of terminal air spaces
is apparent and the alveolar walls create a saw-tooth pattern
(Fig. 1O.8a). The presence of hyaline membranes within the
lung on histological exa mination is a marker of live birth .
They take around 6 hours to develop and up to 12 hours to
become widespread within the lung. Before that time, a char­
acteristic pattern of overdistension of terminal bronchioles
and alveo lar ducts is usually apparent (Fig. 1O.8b). Although
hyaline membranes are classically associated with the
immature lungs of a preterm infant, hyaline membranes may
be seen in the mature fetus following hypoxic injury to the
type II pneumocytes.
When respiration has become fully established the lung
will be uniformly expanded (Fig. 10.S). This can be demon­
strated by histological examination. The terminal air spaces
will not be completely expanded after death, but some
expansion should be present right out to the pleura. Pul­
monary interstitial emphysema has been found to be a use­
ful additional marker of live birth. 34 Septal distension due to
interstitial air trapping and distension of lymphatics can be
distinguished using an endothelial marker. It is also impor­
tant to consider that this might be part of the putrefaction
process and indicate the presence of gas-folming organisms
(Fig. 1O.8c and d).

IS THE BABY OF SUFFICIENT MATURITY
TO SURVIVE?
In the UK, the legal definition of viability was, in 1992,
revised downwards to 24 completed weeks of gestation from
the previous level of 28 completed weeks of gestation. 35
Whilst some babies born at low gestations survive in a hospi­
tal setting where modem facilities and highly skilled carers
are available, should they be born outside hospital they are
unlikely to survive for very long in a domestic environment.
The hazards of unassisted labour and delivery are particularly
great for the immature baby. Subsequently, hypothermia (see
Fig. 10.3, p. 183) and respiratory problems compromise infant
well-being. Opinions about the chances of survival of a par­
ticular baby should take into account both the level of care
required, above that likely to be available at the time of birth,
and the proportion of babies who survive at that gestation
when given appropriate care. This is between 17 and 49 per
cent at 24 weeks' gestation and between 70 and 74 per cent
at 27 weeks' gestation, taking into account fully informative
studies performed in the 1990s. These are reviewed by Evans
and Levine. 36 Likelihood of survival should take into account
the standards of care current at the time of death, when fetal
or infant deaths are investigated or re-investigated several
years after death occurred. 3?
Assessment of fetal maturity is best made using a com­
bination of different factors. These are external measure­
ments, fetal body weight, organ weights (see Appendix 1),
the complexity of the cerebral gyral pattern and histologi­
cal assessment of some organs.
Basic measurements of crown-rump, crown-heel and
foot length, together with head circumference and bipari­
etal diameter, are compared with normal standards and will
provide a reasonable working assessment of maturity;38,39
body weight can be assessed for normality using the
measurement-derived gestational estimation. This is rele­
vant to the likelihood of survival, as both growth-restricted
and macrosomic fetuses are less likely to survive than
appropriately grown infants. The assessment of appropri­
ateness of growth is relevant to the overall interpretation of
the case. Anthropomorphic standards are also available for
comparison with skeletonized fetal remains. 4o
The appearance of the cerebral gyral pattern is particularly
useful in the second half of pregnancy, when the complexity
of pattern develops in a regular fashion. It is particularly use­
ful when there is severe growth restriction. Body lengths and,
in very severe cases, head circumference may be reduced
below gestation-related norms. It can be assessed against
illustrations of gestation-related normal appearance. 39 ,41.42
The histological appearance of the renal cortex is a useful aid
to assessment of gestation from 18 to 36 weeks of gestation
using a combination of the number of generations of
nephrons and the appearance of the nephrogenic zone. 39
A whole-body radiograph will provide additional informa­
tion about fetal maturity based on the presence of ossification
centres and length of long bones. 43 ,44
Is there evidence of prolonged or difficult labour? I 187
LUNG MATURITY
A key factor in the newborn's ability to achieve an inde­
pendent existence is lung maturity. If the lungs are imma­
ture or abnormally developed, then the baby may not
survive despite being of such dimensions that survival
would be expected. Lung weight, despite confounding fac­
tors such as haemorrhage, oedema and inflammation, is a
good, simple indicator of lung growth. The appropriateness
of lung size can be conveniently assessed by suspending
the thoracic viscera by the larynx. The lower borders of the
lungs should be level with the apex of the cardiac ventri­
cles. 45 Histological examination provides further informa­
tion. It is not usually necessary to resort to morphometric
assessment, but sometimes this may be thought desirable in
a particular case (see Baak and Oort 46 for methods).
DID THE BABY HAVE A SEPARATE EXISTENCE?
To establish live birth, it is not sufficient merely to rule out
still birth - the baby must have exhibited signs of life, i.e.
breathing, a heart beat or spontaneous movement, when
completely outside its mother's body. Neither division of
the umbilical cord nor expulsion of the placenta enters into
the definition of live birth. Of these signs the pathologist
can comment only on the likelihood of established respira­
tion. Even this must be done with caution as some respira­
tory effort can occur during the birth process, before
separation from the mother is complete.
The skin of a term baby will usually be covered with
vernix caseosa, often mixed with blood. When no vernix is
present at all, not even in flexures, deliberate action to clean
the baby's skin has occurred. As set out earlier, desiccation
or separation of the umbilical cord, a vital reduction in the
cord stump, the presence of milk curd in the stomach, faeces
(not meconium) in the terminal colon or the presence of hya­
line membranes on histological examination of the lungs
indicates not just live birth but survival for some hours or
days. Thus, they are clear markers of a separate existence.
IS THERE EVIDENCE OF PROLONGED OR
DIFFICULT LABOUR?
Concealed pregnancies are often first pregnancies. Without
appropriate assistance, labour may be particularly long and
delivery itself achieved with difficulty. The stresses experi­
enced by the fetus during labour may be compounded by
prolonged pregnancy. It is well established that perinatal
mOliality is lowest between 39 and 41 weeks of gestation
(term) and is higher outside those limits.4?
When subjected to severe hypoxic stress, a term baby
often passes meconium into the amniotic fluid. This will
adhere to the skin and will also stain the chorionic plate of
the placenta a greenish brown colour (see Fig. 10.16, p. 194).
If the skin has been cleaned, meconium staining can still
1 88 I Fetal and perinatal death
Figure 10.9 Elong at ion of the verticomental diameter of the
cranium can be an indication of long labour.
usually be found beneath fingernails , in skin creases and in
and behind the ears.
In cephalic fetal presentation, prolonged labour is likely
to produce excessive moulding of the head. When the ver­
tex is the presenting palt, this resu lts in elongation of the
verticomental diameter of the cranium (Fig. 10.9). With
vertex presentation there is often an area of localized
oedema, usually with marked congestion, but sometimes
with frank haemorrhage, over the posterior fontanelle or
adjacent parietal or occipital bones. In antepartum fetal
death, very marked congestion of the whole of the scalp
can occur, simulating subgaleal haemorrhage which may
complicate instrumental or, very occasionally, spontan eous
delivery4B (see Chapter 8, p. 160).
Examination of the head and face for localized conges­
tion and oedema may suggest an unfavourable presenta­
tion, which might be expected to add to the difficulty of
delivery. It is not unusual to find cutaneous petechial haem­
orrhages over the presen ting part. They are particularly
likely when placental abruption has occurred (Fig. 10.10).
Another cause is difficulty in delivery of the shoulders.
In this situation, there is increased venous pressure in the
delivered head and neck whilst the thorax is still constricted
by the birth canal. Petechial haemorrhages are commonly
found over the face, h ead and n eck and occasionally on the
chest wall. Such haemorrhages should not be interpreted as
evidence of strangulation or deliberate airways obstruction
without corroborative evidence.
Shealing tears in the tentorium, unassociated w ith dam­
age to cerebral sinuses, indicate excessive and/or rapid dis­
tortion of the cranium. A baby presenting in the breech
position is at particular lisk during unattended delivery. The
- --
Figure 10.10 Blotchy facial petechiae in a non-macera ted
antepartum still birth as a result of retrop lacental haemorrhage.
breeched baby can pass through an incompletely dilated
cervix, which then impedes passage of the head. Breech pres­
entation is suggested by the presence of cutaneous petechial
haemorrhages over the lower trunk and legs. Sometimes,
haemolThage into underlying muscle is extensive. Unskilled
effOlts to complete delivery by traction on the baby's trunk
can compound the probl em causing injury to the liver, which
will be tense because of congestion. A baby's liver is poorly
protected by the rib cage, particularly when delivered
preterm. When a term baby appears to have presented in the
breech position, it is important to look for factors that predis­
pose to breech presentation, such as neuromuscular problems
or renal agenesis (because of oligohydramnios), as they may
make a major contribution to failure to survive; death is then
attributable to natural disease.
The umbilical cord shlmp should be examined carefully
and any discolouration or dehiscence noted (see ab ove). The
appearance of the free end of the cord may indicate the
method of its division (see Chapter 8, p. 160). It may corre­
spond with the free en d of the cord attached to a placenta
located elsewhere (Fig. 10.]]). The presence or absence of a
clip or tie on the cord should be noted and described.
ARE THERE ANY SIGNIFICANT INJURIES?
Before ascribing significance to any injury observed, it is
important to identify commonly described bilth injuries
and to be aware of the circumstances in which they occur.
They are discussed in Chapter 8 (p. 161).
- .a_ ~_~

Figure 10.11 Torn umbilical cord, concealed delivery.
It is clearly important, when examining a baby found
dead after concealed pregnancy or unattended delivery, to
look carefully for evidence of inflicted injury. Not every
injury found will be deliberately inf1icted. It is impOitant to
identify injuries that might occur as a result of maternal
efforts to self-deliver as well as any which might arise during
the course of prolonged labour. Other injuries may occur after
death, during or followin g efforts to conceal the body. Evi­
dence of post-mortem animal-inf1icted injury is sometimes
seen in babies who have been aba ndoned out of doors.
The presence and nature of any irUury is recorded and
photographed. Injuries are infrequent, death is more usually
due to lack of cares Drowning and hypothermia from expo­
sure are other causes of death in these circumstances. 49
However, in one study from South America,50 violent injury
was fou nd in a large proportion of in fant icide cases.
EXTERNAL FEATURES
Linear or slightly curved, parallel, abrasions on the neck
are sometimes seen. Examination of skin tags will show
them to run from the shoulder to t he head. These can be
infli cted by the mother trying to grasp th e chin/occiput in
an effort to expedite delivelY of the shoulders.
Early skin slippage, which has dri ed out, co mmonly
seen over bony prominences such as the brow, cheek bone
or chi n, can be mistaken on firs t glance for abrasion due to
deliberate injury (see Fig. 10.5, p. 184). Other changes of
maceration should be carefully sought.
Cutaneous bruises are unusual bilth injuries. Any found
should be sampled for histological examinatio n to distin­
guish t hem from injuries caused post-m oltem.
'Bru ising' of buttocks and legs may be see n in infants
presenting in the preterm breech position as a result of
hypoxic/hypostatic haemorrhage into underlying muscles.
A Mongolian blue spo t (see Fig. 8.7, p. 152), usually pres­
ent at the apex of the gluteal fold, can sometimes be very
extensive and should not be interp reted as trauma.
Althou gh more common in Afro-Caribb ean infa nts and
Are there any significant injuries? I 189
Figure 10.12 Complicated skull fra ct ure of the parietal bone
following maternal abdominal trauma. (Courtesy of Dr KJ
McKenzie, Ed inbu rg h.)
those whose parents are from the Indian su bcontinent, it is
somet imes seen in babies of European parentage.
FRACTURES
Skull fractures are unusu al foll owing non-instrumental
delivery, so any fracture should be documented both photo­
grap hically and radiologically, and adjacent tissues exam­
ined with care. As in older individuals, fractures which
extend across suture lines result from impact trauma (Fig.
10.12). Fractures resulting from the process of birth are dis­
cussed in Chapter 8 (pp. 161-162).
INTRACRANIAL HAEMORRHAGE
Intracran ial haemorrhage should be carefully documented
and photographed. Much information about causatio n can
be addu ced by careful observation and recording of the site
of haemorrhage 48 (Fig. 10.1 J). Extradural hae mato ma is
excessively rare in newborns, apart from a trace of blood in
the vicinity of a fracture.
Subdural haemorrhage is almost always due to trauma.
Heavy blood sta ining of the cerebrospinal fluid (C SF). or
haemorrhage in the posterior fossa may be the result of sub­
dural haemonhage over the convexities but, particularly in
preterm infants, may result from intraventricul ar haemo r­
rhage, when it emerges into the subarachnoid space through
the fourth ventricular fora mina. It may sometimes have a
local cause. Shearing tears of the falx or, more usu ally, the
tentorium indicate excess ive moulding of the head.
190 I Fetal and perinatal death
Asphyxia
Cutaneous
petechiae
Haemorrhagic
infarction
Subarachnoid
haemorrhage
Falcine
haemorrhage
Intraventricu la r
haemorrhage
Germinal matrix
(subependymal)
haemorrhage
r
Lep<om",;::: J IIII
Figure 10.13 Sites of pericranial and intracranial haemorrhage in the newborn by causation. (From
edn, 2007, p. 287, Fig. 13.16, with permission, Springer Science and Business Media.)
Subarachnoid haemorrhage is usually focal but can be
confluent, usually over the temporal poles (but see Fig. 8.19,
p. 160). It is usually hypoxic in origin. Focal subarachnoid
haemorrhage is often found at the base of a torn bridging
vein when the vessel has retracted beneath the membrane.
Cerebral haemorrhage is rarely traumatic. It is usually
related to bleeding into an area of prior hypoxic injury.
The same is true of intraventricular haemorrhage (NH),
which arises from hypoxic injury to small vessels in the
periventricular germinal matrix or choroid plexus.
Although NH most commonly occurs in the preterm baby
postnatally, there are well-documented examples of NH
occurring in utero. 51 Therefore, its presence in a perinatal
death should not be interpreted as a marker of live birth.
FETAL DEATH FOLLOWING MATERNAL INJURY
Fetal deaths following maternal injury are not common,
although trauma complicates 6-7 per cent of pregnancies. 52
The extent of legally directed investigation of the baby in
these circumstances depends on a number of factors: the
causation of the injury, whether the baby was liveborn or
still born and the jurisdiction under which it falls. The procu­
rator fiscal (Scotland) is more likely to instruct an investiga­
tion into a fetal death or still birth than is the coroner
(England) if the status of the fetus as a person seems some­
what ambiguous. Fetal maturity is an important factor and is
most impOliant when prosecution of any sort is being con­
sidered?8 Studies of the outcome of minor trauma in preg­
nancy are few. Fort and Harlin5J identified women from the
Trauma
Caput
succedaneum
~--------------------
Subaponeurotic
---"-.A:---'I------------------ (galeal) haemorrhage
""""""""'c------------- Cepha Ihaem atom a
,,-'Ic-'\--\-\--\\----------- Ex trad ura I
haemorrhage
Brain
rf ----+\-\-\--+-\-\-------- Su bd ura I
haemorrhage
:::0'"'"' (9;''''1
Bone Periosteum
Fetal and Neanatal Pathalagy, 4th
American Collaborative Study who had suffered non-cata­
strophic trauma during pregnancy and looked at the out­
comes. Traumatic episodes comprising blows to the
abdomen, falls and minor road traffic accidents complicated
210 (5.7 per cent) pregnancies, half occurring in the third
trimester. No increase in pregnancy loss was found when
compared with non-traumatized control subjects.
Fetal loss is more likely following major maternal trauma
(61 per cent) than following minor trauma (27 per cem)54
and when the maternal injury severity score is high 55 ­ 57
and the mother is in shock. When trauma is non­
catastrophic, complications are unusual. Fetal loss occurred
in about nine per cent of cases in one study.58 The reason
is most commonly premature labour, but placental abrup­
tion sometimes occurs. Some advocate careful observation,
even after minor trauma, for this reason. 59 However, a
prospective study of nearly 50000 cases found pregnancy­
related complications in only 0.001 per cent.50
BLUNT ABDOMINAL TRAUMA
The most common cause of trauma to the abdomen of a preg­
nant woman is motor vehicle collision (MYC).51 In one study,
maternal injury following MYC accounted for 80 per cent of
cases of abdominal trauma to pregnant women, of whom 10
per cent were pedestrians. 55 Other injuries, falls, assaults and
self-harm, are much less common. One institution found that
serious assault accounted for 31.5 per cent of trauma to preg­
nant women over a 9-year period 52 and for 22 per cent of
injury in another study over a similar period. 52

With expanding vehicle ownership, pregnant women are
increasingly involved in MVCs whether as a driver, passen­
ger or pedestrian. One study found the highest incidence of
fetal death following MVC in teenage mothers. 63 Occasion­
ally pregnant women are involved in accidents as cyclists. In
my experience, fal ls from bicycles can provoke premature
onset of labour and fetal death as a result of prematurity or
placental abruption. Although accidents can occur at any
stage of pregnancy, deaths occurring in the last trimester of
pregnancy (after 28 weeks' gestation) are the most likely to
be subjected to medicolegal investigation, although the
deaths of less mature fetuses are sometimes investigated.
When assessing the likelihood of feta l death being the result
of any accident, an acquaintance with the incidence of spon­
taneous pregnancy loss at that particular time of gestation is
useful. Spontaneous fetal loss is less common as pregnancy
advances. In the absence of ma terna l predisposing factors,
some of which may be revealed on exa mination of the pla­
centa, co incidental fetal demise or spontaneous onset of
labour is unlikely after 32-34 weeks of gestatio n.
The prime concerns after materna I involvement in an
MVC are for maternal and fetal well-being, the former bear­
ing critically on the latter. 64 Observation of mother and fetal
monitoring are necessary for several hours, even when
maternal injury is slight. 61 Any maternal injuries present
should be carefu lly recorded . A careful examination of the
abdomina l wa ll for bruising is essential. When the accident
involves the mother as a pedestrian or cyclist, the full cir­
cumstances and extent of injury to the mother are recorded
and are usefully made available to the pathologist before
fetal examination commen ces.
When the mother was the driver or passenger in a vehi­
cle, information about the type of restrain ts and whether
they were used is important. Fetal mortality and morbidity
has been associated with lap strap use. Early reports of uter­
ine rupture 65 and fetal inju ry66 have deterred some women
from using restraints. One study found that fetal mortality
increased from 14.4 per cent to 16.7 per cent when this type
of restraint was used, a small and comparatively non­
significant increase, but maternal mortality was reduced by
half67 Despite more recent advice that the risks to the fetus
are reduced with appropriate use of three point restraints,68
the message is difficult to get across to some women. 55
Impact severity is the best prediction of adverse fetal out­
come, but appropriate seatbelt use (three-point restraint)
improved fetal outcome following impacts of mild or mod­
era te severity.69 The position of the restraining straps across
the woman's body is criti ca l to fetal well-being. Defo rma­
tion of the uterus from forcible contact with the restraint is
the cause of retropla cental haemorrhage producing a shear
strain across the uteroplacenta l interface. There may be
additiona l stresses from tensi le failure during rapid deceler­
ation of the uterus following impact. 69 Restraints shou ld be
positioned above and below the gravid uterus. 6S
Premature onset of labour is a common complication
(28 per cent) when trauma occurs between 25 and 37
Fetal death following maternal injury I 191
weeks' gestation 70 but is not usually a cause of fetal loss.
Uterine rupture is uncommon/! but when it occurs fetal
mortality is close to 100 per cent. Amniotic fluid embolism
is an occasional compJication n
Other causes of blunt abdominal trauma are falls and
assault, such as ki cking or heavy blows to the maternal
abdomen. Whilst the velocity component of trauma
enco untered in MVC is missing, nevertheless these insults
can result in pregnancy loss and fetal injury even when the
maternal injury severity score is zero. 62 The in cid ence of
assault during pregnan cy is difficult to assess and likely to
be under-estimated. Systematic questioning of mothers
attending antenatal clin ic found that an assault had
occurred at some time during pregnancy in 14.1 per cent of
384 women. 73 In another study, physical violence occurred
in 11.1 per cent of 6143 women in the year prior to deliv­
ery, with an increase in materna l morbidity including
preterm delivery, but no other adverse fetal outcome was
noted l4 A study from Sa udi Arabia reported physical vio­
lence in pregn ancy in 21 per cent of over 7000 records;75
there was an increased rate of placental abruption and pre­
mature delivery. Confidential enquiries into maternal
deaths in the UK found that more than one-third of domes­
tic abuse cases start during pregnancy and the risk of fetal
loss or still birth is doubled. 76 The subsequen t triennial
report cites continuing vio lence to pregnant women.
Eleven deaths in that triennium were the result of murder
by the woman's partner, higher than the number of deaths
due to MVC or amniotic fluid embolism and equal to
deaths from sepsis. 77 Fatalities are the tip of the iceberg of
domestic viol ence in pregnancy. Adverse feta l outcome fol­
lowing domestic violence, including premature delivery
and a 3.5-fold in crea se in neonatal death, is described in a
North American study.78 A literature review of 30 informa­
tive studies of intimate partner v iol ence 9 found an
increase in fetal and maternal mortality and morbidity,
with a feta l death rate around 16 in 1000 in affected
pregnancies.
Chronic subdural haematomas have been demonstrated
in the neonatal period in offspring of women subjected to
physical abuse, which included kicking and blows to the
abdomen dUling pregnancy.80,8! In the Stephens et aisl study,
cerebral atrophy and haematomata of different ages were
apparent on imag ing shortly after birth. Repeated assault in
the context of domestic violence increases the risk of peri­
partl.lm complications. B2
Subdural haematoma have been desClibed in 47 offspring
of immigrant Pacific Islanders in New Zealand. s3 Forty-four
of these babies were still born; some had intraventricular or
intracerebral haemorrhage as well as a subdural haematom a.
Some had sca lp or significant subgalea l haematoma. The
haemon'hages were not accompanied by skull fra ctures or
tears of the dural folds. Coagulation studies were normal in
all mothers tested; a further eight had normal platelet
counts. The ca use of the intracranial and pericranial haem­
orrhages in these babies was thought to be particularly
192 I Fetal and perinatal death
forceful abdominal massage, resulting in slow head com­
pression. It is probably significant that more than half of the
injured babies were presenting in the breech position.
ELECTRICAL INJURY TO THE PREGNANT WOMAN
The literature pertaining to electrical injury, including lig ht­
ning strikes, is small, with a heavy emphasis on single case
reports. The outcome is very variable. Lieberman et al 84
report six cases of electrical injury related to domestic acci­
dents occurring between 20 weeks' gestation and term. In
two cases in which there was an immediate reduction in fetal
movements, fetal death occurred within one week of the
accident. A furth er fetus died 12 weeks after an insult at 2 1
weeks' gestation and was growth restlicted at birth. A causal
relationship here is tenuous. The other three were born at full
term and survived. The same groups report abortion after
an electrical accident. 8s Jaffe et al 86 report fetal death at
14 weeks occurring within hours of a domestic mishap.
Reports of fetal death from the late nineteenth and early
twentieth century (cited by Mehl 87 ) describe immediate
cessation of fetal movements following the incident.
Interval deaths and chronic problems such as oJigo­
hydramnios 84 are more difficult to evalu ate from the point
of view of causation . The frequency of pregnancy loss from
other (natural) causes at that particular stage of pregnancy
must be taken into account. Yoong88 describes massive
placental abruption and fetal death 24 hours after electri­
cal injury.
Fetal mortality following maternal electrical injury,
when based on accumulated reported cases, is high.
Fatovich 89 found a fetal mortality of 73 per cent among 15
reported cases. However, reporting bias is likely w hen the
subjects are single case reports. The only prospective study
of pregnancy outcome following maternal electrical injury
followed 31 women and fo und no adverse outcome related
to the insult. 9o Fetal outcome following matern al lightning
strike is variable. Of the 12 reported cases 91 - 95 half of the
fetuses died. Most mothers sustained burns. Infant survival
bore no relationship to maternal loss of consciousness or
being thrown to the ground . Loss of fetal movements was
observed soon after the incident when the fetus died. There
is a Single case report of taser injury in custody 87 resulting
in miscarriage 7 days after the incident. There were other
inauspicious features in that case.
PENETRATING TRAUMA TO THE MATERNAL ABDOMEN
The frequency of penetrating trauma to the gravid uterus has
wide geograp hic variation. It is, not surprising ly, enco un­
tered more frequently in war zones 96 and those countries
where gun ownership is common. 97 Fetal well-being is com­
promised by penetrating injUly to the uterus, but not all
cases are fatal, even when there is some bleeding into the
amnio tic cavity or minor fetal injury is sustained. 98 ,99
Maternal death is infrequent. The gravid uterus provides
protection to other maternal organs. As the uterus increases
in size, organs, particularly the in testines, are displaced
upwards and may be spared from direct injury. In the case of
gunshot wounds, the uterus, being of dense texture, absorbs
most of the energy, reducing the risk of damage to other
maternal abdom in al viscera. loo Intra-uterine fetal death,
premature onset of labour and fetal injury are more likely
following penetrating abdominal trauma than is the case
following blunt trauma. S5 Fetal injury is seen in 60-90 per
cent of cases, about half being serious, including penetrating
wounds of thoracic and abdominal viscera. Injury to the
placenta and umbilical cord is also described. IOI Pelinatal
mortality of 47-71 per cent is likely.loo Among Awwad et
al's96 cases, half of the fetuses died because of maternal
hypotension or direct fetal injury. Sel f-directed penetrating
trauma has, bizarrely, been used in attempts to terminate
pregnancy. Both self-inflicted gunshot wounding 102 and
vo luntary repeated stabbi ng l03 are recorded .
Fetal Injuries and Cause of Fetal Death
The usual cause of fetal death followin g blunt trauma to the
maternal abdomen is retroplacental haemorrhage. It usually
occurs within hours of the injU1Y and can be predicted by
increased uteline contractions. 61 Delayed (several days) placen­
tal abruption has been recorded. Fetomatemal haemonhage
is more common in pregnancies complicated by maternal
trauma. 104 It is more likely when the placenta is anterior61 and
can result in fetal dysrhythmias, anaemia and fetal death from
exsanguination. Placental abruption follo wi ng MVC is signifi­
cantly associated with higher maternal injulY severity score
a nd higher vehicle speed. 105 Still birth occurred in more than
50 per cent of cases in which abruption followed the collision,
with no fetal deaths in the absence of abruption. Rogers et al 106
also fo und an association between placental abruption and
high maternal severity score. They found that fetal cardio­
graphic monitoring could reduce fetal mortality.
Any injulY resulting in maternal hypotension puts the
fetus at risk. If prolonged or severe, it will result in fetal
death from hypoxic/hypotensive injury.107 These changes
are likely to be seen in the brain and kidney and occasion­
ally the heart, liver and intestines. Pathol ogical findings
are of extensive haemorrhagic infarction. 108 When the
effect on the maternal circulation is less severe, the fetus
may survive, but wit h brain injury. Rapid resuscitation of
the mother gives the fetus the best chance of survival. 59
Blunt injury to the maternal abdomen can result in
direct injury to the fetus, altho ugh it is not common. 58 This
is most likely to occur when maternal injury occurs late in
pregnancy. The fetal head is afforded protection by the
maternal pelvis but is vulnerable when the pelvis is frac­
tured. Sometimes the fetal skull is fractured when maternal
injury is trivial. Skull bones are the most common sites of
fractures susta ined by the fetus (Fig. 10.12, p. 189) and may
 Can I give a cause of death? I 193

be multiple. 109 Not all are fatal. 110 Skull fractures may be Malformations which should be carefully so ug ht when
complicated by intracranial haemorrhage, perhaps the resuscitation is unsuccessful include laryngeal atresia
more significant injury. 54 Intracranial haemorrhage and (Fig. 10. 14) or severe laryngea l stenosis, laryngeal cleft
tentorial tears are also described. 66 (Fig. 10.15), diaphra gmatic hernia and rena l age nes is.
Subdural and pericranial haemorrhage in t he absence of Cardiac disorders do not usually present as fai lure to
skull fracture and tears of dural folds can cause death within respo nd to resuscitation. Those cardiac anomalies responsi ­
hours, although some fetuses apparent ly survive for several ble for peJinatal death are usually ductus-d ependent lesions.
days.B3 Not all antepartum subdural haematomas are fatal. Death usually occurs after sudden deterioratio n at 1 or 2
Some babies are liveborn with significant morbidity owing days of age. Both the cardiac anatomical anoma lies and
to cerebral atrophy, hydrocephaly or porencephalyBI .83. 111 some degree of cl osure of the ductus arteriosus w ill be
Whilst the vast majority of in utero subd ural haemorrhage readily apparent on exami natio n of th e thorax.
can be related to tra umatic insult, there are occasional Causes of su dd en de ath in the neon ate are considered
reports of this type of haemorrhage being found in situations furt her in Chapters l1 (p. 209) and 12 (pp. 248-2 50).
of reduced blood coagulability. Bilateral fetal subdural Placental findin gs asso ciated with sudden death are vela­
haemorrhage in late pregnancy has been described as a com­ mentous cord insertion with tearing of major vessels, tight
plication of maternal warfarin therapy, 11 2 although no infor­ cord knots, meconium staining of the fetal surface
mation about precise timing or repetition of haemonha ge (Fig. 10.16) and retroplacental haemorrhage (Fig. 10.17).
was ava il a ble. It has been described occasionally as a feta l
complication of inherited clotting disord ers, such as factor X CAN I GIVE A CAUSE OF DEATH?
deficiency.113 Usua lly, however, fatal thrombocytopenia and
clotting factor deficiencies give rise to intracranial haemor­
Whilst the answer to t his question before post- mortem
rhage in other sites, predominantly intracerebral and in tra­
should, in many cases of antepartum still birth, be 'no', few
ventricular. 114 See Chapter 4 (pp. 94- 97).
of these deaths are, in fact, reported to a coroner or procu­
Cerv ical injury is rare, bu t lower cervical injury (C5) is
rator fiscal - the ass umption being that examination of the
describ ed follo wi ng ]'vTVC when the mother was using a
placenta, even if fetal exami nation is denied, will provid e
seatbelt. 115 Some authors suggest that protection from
the answer, or at least substantial clues, to t he underl y in g
direct feta l injury in MVC is afforded by an anterior pla­
problems.
centa, but the placenta itself is vulnerable, resulting in
retropl acental haematoma. The posterior situation of the
placen ta was thought to be significant in one infant who
sustained lace ration of the liver, ha emoperiton eum and
subdural and subp leural haemo rrhage. 116 Splenic injury
with haemoperitoneum is also described.54.117 Fetal chest
injuries with h ae mothora x and pulmonary contusion can
follo w MCv. 11 8 Scalp bruising was presen t in one infant
foll owing MVC, although the cause of death was rupture of
the placen tao U9

IS THERE A NATURAL CAUSE FOR DEATH?

When investigating apparent in t rapartum-related deaths it is

important to be awa re of congenital or acquired diseases
which might cause intrapartum death or failure to respond to
resuscitation. It is important, too, that appropriate techniques
are used to best demons trate any birth trauma a nd that
microbiological samples are taken as unsusp ected asce nding
infection can cause intrapartum death. These samples are
most important when group B streptococcal infection is
invo lved as fetal death is due to toxaemia, a nd histological
evidence of response to infection may be slig ht or absent.
The presence and nature of any dysmorphic features are
noted. These may contribute to a syndrome diagnosis, Figure 10.14 Failed resuscitation of a mature, non-dysmorphic
which may be related to death. Remember that the presence baby. Laryngeal atresia: the upper larynx appears normal;
of two or more dysmo rp hic features is an indication for complete airway obstruction by a thick bar of cartilage in the
chromosome examination . lower larynx.
194 I Fetal and perinatal death
Figure 10.15 Posterior laryngeal cleft can facilitate massive
aspiration or, as in this case, repeated displacement of an
endotracheal tube, result ing in hypoxic brain damage. (From Fetal
and Neonatal Pathology, 4th edn, 2007, p. 538, Fig. 20.7, with
permission, Springer Science and Business Media.)
Figure 10.16 Meconium staining of the fetal surface of the
placenta is an indication of fetal distress in mature babies;
it persists for several days.
In the face of an unexpected intrapartum (fresh) still
birth, both obstetrician and paediatrician are more likely to
report the death to the lo cal medicolegal officer, particularly
if there is any expression of dissatisfaction from the family
in respect of the care given during labour. Whatever the
likely cause of death in this situation, it is important that
assistance is sought from a perinatal pathologist. He or she
will be more conversant with the range of likely natural
Figure 10.17 A placenta after concea led delivery. There is
sign ificant blood clot attached to the maternal surface of
the placenta.
causes of death, more likely to be able to recognize any
iatrogenic contribution to death and less likely to compound
the problem by wrongly interpreting trivial findings.
SHOULD THE INTRAPARTUM STILL BIRTH
BE A MEDICOLEGAL AUTOPSY?
Post-mortem examination following intrapartum fetal
death should be done wi th particular care. A forensic
pathologist presented with this problem is advised to seek
advice of, and perhaps even collaborate with, an experi­
en ced perinatal pathologist before embarking on the
necropsy. A full perinatal necropsy is needed with photo­
graphy and radiological examination, bacteriological sam­
ples fro m baby and placenta, complete external
examination, full dissection of organs and fixation of the
brain prior to detailed examin ation. It is important to
examine the placenta, umbilical cord and membranes and
to take routine samples from fatal organs and gestation sac
for his tological examination. 39
The clinical history may indicate the need for Yates' dis­
section to remove the cervical spinal cord en bloc within the
vertebrae. 120 If not, then in the absence of intracranial haem­
orrhage remova l of the spinal cord is recommended. Full
sampling of organs for histological examination is important.
Examination of the umbilical cord, placenta and membranes
is essential in every case as it may demonstrate a natural
cause of death.1 21 Even when the investigation has not been
directed by the procurator fiscal/coroner/medic al examiner's
departmen t, it has become increasingly likely that the pathol­
ogist's report will be scmtinized by lawyers and medical
expelis on behalf of the bereaved parents. Bea ring this in
mind, it is important to check through the report ca refully for
factual or transcribing errors (particularly numbers and deci­
mal points). It is equally important not to promote the per­
ception of wrongdoing by the inappropriate interpretation of
pathological findings, such as ascribing cephalhaematoma
to trauma when this lesion can occur during an easy, normal
vaginal delivery. Similarly, shearing tears of the tentorium
are a marker of defo rmation of the skull but not a cause of

death. They should also selve as a reminder for detailed
neuropathological examination, which is only possible after
fixation of the brain, even in mature babies.
It is current practice for obstetricians or neonatologists
to give a copy of the post-mortem report to bereaved parents.
For this reason, it is our departmental practice not to issue to
clinicians the clinical summary made by the pathologist for
departmental use in case it is based on incomplete, or occa­
Sionally erroneous, information. Such errors can aggravate
parental concerns about the possibility of substa ndard care. It
is sensible to avo id the use of potentially emotive telms such
as asphyxia or a noxia, which might exacerbate perceptions
of malpractice. Hypoxia is a more appropriate and accurate
term. Another potential aggravation is the overinterpretation
of amniotic fluid debris, which is frequently present in the
intrapulmonary airways, as meconium, which may be present
in air spaces in abundance fol1owing hypoxic stress in utero
in a mature baby.
REFERENCES
Adelson L. Pedicide revisited . Am J ForeJ1 sic Med Pathol 1991 ;
12: 16-26.
bospita l-delivered infants . Acta Paediatr SCGnd 1991;
2 Sims MA, Collins KA. Fetal death. A lO-year retrospective
80:602- 10.
review. Am J Forensic IvIed Patlwl 2001; 22:261-65.
27 Balmer SE, Wharton BA. Diet and faecal fl ora in tbe newborn:
3 Confidential Enquiry into Maternal and Child Health. Still
breast milk and infant formula. Arch Dis Child 1989 ;
Birth, Neonata l and PostlleoJ1atal Mortality 2000-2002.
64:1 672- 77.
London: RCOG Press, 2005.
4 Schott J, Henley A, Kolmer N. Pregnancy and the Death oj a
Baby: Guidelines Jor ProJessionals, 3rd edn. London: SANDS
(St illbirth and Neonatal Death Society). 2007, pp. 171-2.
5 Kellett RJ. Infanticide and child destruction - the historical ,
lega l and pathologica l aspects. Forensic Sci Inter 1992; 5 3: 1- 28.
6 Mason JK. Unnatural death of the fetus and the newborn:
abortion, child destru ction and infanticide or child murd er. In
Maso n JK (ed.) ForeJ1 sic Medicine Jor Lawyel·s. Butterworths:
London, 2001, pp. 254-68.
7 Di Maio YJ, Di Maio D. ForeJ1sic Pathology, 2nd edn. Boca
31 Moar JJ. The hyd rostat ic test - a valid method of
Raton: CRC Press, 200 1, pp. 336-7.
determining li ve birth? Am J Forellsic IvIed Pathol 1997;
8 Bowen DA. Concealment o f birth. child destruction and
in fant icide. In Mason .JK (ed.) Paediatric Foren sic MediciJ1e ft
Pathology. London: Chapman and Hall Medical, 1989,
pp. 178-90.
9 Ophoven JJ. Pediatric forensic pathology. In Gilbert-B arness E
(ed.) Potter's Pathology oj the Fetus, InJant alld Child,
2nd edn. Mosby: Elsevier Inc. , 2007, pp. 816-20.
10 Goldenberg RL, Hale CB, Houde J et al. Neonatal deaths in
Alabama. Ill. Out-o f-h os pital births. 1940- 1980. Am J Obstet
Gynecol 1983; 147:687-93.
11 Ong BB, Green M. Infanticide in Malaysia: two case reports
and a review of litera ture. Am J Fo rensic Med Pathol 2003;
24:64-9.
12 Hamilton A, English S. Baby bones uncover an island murder
mystery. The Times, January 18,2002 ; p. 3.
13 English S. Woman charged after children's bodies found.
The Times, April 18, 2001; p. 1.
14 Nurse hid her dead babies in A ttic. The Herald, August 27
2005, p. 6.
15 Boyes R. Mother tells of her dead babies' secret. Th e Times,
August 3 2005, p. 32.
References I 195
16 Funaya ma M, Sagisaka K. Consecutive infanticid es in Japan.
Am J Forensic iVIed Pathol 1988 ; 9:9-11.
17 Yamauchi M. Usami S, Ikeda R et a l. Medico-legal stud ies on
infanticid e: statistics and a case of repeated neon at icide.
Forensic Sci Inter 2000; 11 3 :205-8.
18 Shiono H, Maya A. Tabata N et al. Medicolegal aspects of
infanticid e in Hokkaido District, Japan. Am J Forensic Med
Pathol 1986; 7: 104-6.
19 Funay ama M, Ikeda T, Tab ata N et al. Case repon:
repeated neonaticides in Hokkaido. Forensic Sci Int 1994;
64:147- 50.
20 Davies RJ. The midwife's role in the managem ent of normal
labour. In Cha mberlain G (ed.) Turnbull's Obstetrics.
Edinburg h : Churchill Li v ingsto ne, 1995, pp. 581-90.
21 North ern Region Perinatal Mortality Survey Coordinating
Group. Co ll aborative survey of perinatal Joss in planned and
unplanned home births. Br Med J 1996; 313:1 306-19.
22 Mitchell EK, Davis JH. Spon taneo us bil1hs into toilets.
J Foren sic Sci 1984; 29: 591 -6.
23 Genest DR, Singer DB. Estimating the time of death in still
born fetuses: Ill. Ex ternal fetal examination; a study of
86 sti ll borns. Obstet Gynecol 1992; 80:593-600.
24 Harries JT. Meconium in health and disease. Br Med Bull
1978 ; 34:75-8.
25 Rob erton NRC. Care of the normal term newborn baby. In
Rennie JM, Roberton NRC (eds) Textbook oj Neonatology.
London : Churchill Livingsto ne, 1999, pp. 373 -87.
26 Adlerberth J, Carlsson B, Man de Pet al. Intest inal
colonization with enterobacteriaceae in Pak istan and Swedish
28 Tabata N, Morita M. Azumi J. A frozen newborn infant: froth
in the air-passage after thawing. Forensic Sci Int 2000;
108 :67-74.
29 Knight B (ed.). Infanticide and still birth. In Forensic
Pathology. London: Ed wa rd Arnold, 199 1, pp. 402-JJ.
30 Wi gglesworth JS. Pathology of Intrapartum and early
neonatal death in the no rmally form ed infant. In
Wigglesworth JS, Singer DB (eds) Textb ook oj Fetal and
Perinatal Pathology, 2nd edn. Malden, MA: Blackwell Sci ence,
1998, pp. 251-68.
18: 109-10.
32 Attwood HD. The histological diagnosis of amniotic fluid
em bolism. J Pathol Bacteriol 1958; 76: 211-15.
33 Konstanidou AE, Agapitos E, Korkolopoulou P, Davari s P.
Tracheoesophageal malformation: pathogenetic evidence
provided by two cases. Teratology 2001; 63:lJ-14.
34 Lavezzi WA, Keough KJV! , Der'Oha nnesia n PJD et al. The use
of pulmonary interstitial emphyse ma as an indicator of li ve
birth. Am J Forensic Med and Pathol2003; 24:87- 91.
35 St ill birth (Definition) Act 199 2. Eliz II, 1992, c29.
36 Eva ns DJ, Levene MJ. Evidence of selection bias in preterm
surv ival studies: a syste matic review. Arch Dis Child Feta l
2001; 84:F79 - 84.
37 Huxley AK, Froede RC, Birkby WH. Strangulation of pregnant
woman leads to one first-degree murder indictment for the
death of the mother: A medicol ega l reconsideration of
maternal/fetal/homicide. Am J Forens ic IvIed Pathol 2001;
22:51-4.
38 Singer DB , Sung C-R, Wigglesworth JS. Fetal growth an d
maturation ; with stan dards for body and organ development.
In Wigglesworth JS, Singer DB (eds) Textbook oj Fetal and
196 I Fetal and perinatal death
Perinatal Patho logy, 2nd edn. Malden Ma: Blackwell Science,
1998, pp. 8-40.
39 Keeling JW. The perinatal necropsy. In Keeling JW, Khong TY
(eds) Fetal and Neonaral Pathology, 4th edn. London:
Springer, 2007, pp. 20-53.
40 Huxley AK, Kosa F. Estimation of lunar age from diaphyseal
length from the fetal materials in the fetal collections at
Albert Szent-Gyorgyi Medical Uni ve rsity, Hun gary (abst).
Proc Am Acad Forensic Sci 1999 ; 5:222-3.
41 Dorovini-Zis K, Dolman CL. Gestational developm ent of brain.
Arch Pathol Lab Med 1977; 101:192-5.
42 Feess-Higgins A, Lan'oche J-C. Del!eiopmel1t oj rhe Humlln
Fetal Brain: An Anatomical Atlas. Paris: INSERM CNRS,
Masson, 198 7.
43 Foote GA, Wilson AJ, Stewart JH. Perinatal post-mortem
radiography - experience with 2500 cases. Br] Radiol
1978; 51 :351-6.
44 Russ ell JGB. Radiological assessment of age, retardation and
death. In Barson AJ (ed.) Laboratory In/iestigation oj Feral
Disease. Bristol: Wright, 1981, pp. 3-16.
45 Gould SJ. The respiratory system. In Keeling JW, Khong TY
(eds) Feral and Neonatal Pathology, 4th edn. London :
Springer, 2007, pp. 531-70.
46 Baak JPA, Oort J. A Manual oj Morphometry in Diagnosric
Pathology. Berlin: Springer-Verlag; Heidelberg, 1983,
pp. 131-6.
47 Hilder L, Costeloe K, Thilaganathan B. Prolonged pregnancy:
evaluating gestation -speci fi c risks of fetal and infant
mortality. Br] Obstet GYl1aeco/1998; 105:169-73.
48 Keeling JW. Intrapartum problems. In Keeling JW, Khong TY
(eds) Fetal and Neonatal Patho logy, 4th edn. London:
Springer, 2007, pp. 273-96.
49 Saunders E. Neo nat icid es following 'secret' pregnancies:
seven case reports. Public Health Rep 1989; 104:368-72.
50 Mendlowicz MV, Jean-Louis G, Gekker M, Rapaport MH.
Neonaticide in the city of Rio de Janeiro: forensic and
psycholegal perspective.] Forensic Sci 1999; 44:741-5.
51 de Vries LS, Rennie JM. Preterm brain injury. In Rennie JM,
Roberton NRC (eds) Textbook oj Neonatology, 3 rd edn.
London: Churchill Li vingsto ne, 1999, pp. 1252-7 \.
52 Connolly AM, Katz VL, Bash KL et al. Trauma and pregnancy.
Am] Perillatol1997; 14: 33 1-6.
53 Fort AT, Harlin RS. Pregnancy o utco me after noncatastrophic
maternal trauma during pregnancy. Obstet GYl1ecol 1970;
35:912-15.
54 Rothenberger D, Quattlebaum FW, Perry JF et al. Blunt
maternal trauma: a revi ew of 103 cases.] Trauma 1978;
18:173-8.
55 Shah KH, Simons RK, Holbrook T et al. Trauma in
pregnancy: maternal and fetal outcomes.] Trauma 1998;
45:83-6.
56 Curet MJ, Schermer CR, Demarest GB et al. Predictors of
outcome in trauma during pregnancy: identification of
patients who can be monitored for less than 6 hours.
] Trauma 2000; 49: 18-24.
57 Theodorou DA, Velmahos GC, Souter I et al. Fetal death after
trauma in pregnancy. Am Surg 2000; 66:809- 12.
58 Good w in TM, Breen MT. Pregnancy outcome a nd
fetomaternal haemorrhage after non catastroph ic trauma .
Am ] Obster Gynrcol 1990 ; 162 :665-71.
59 Siebenga J, van del' Schoot JT, Keeman IN. Trauma during
pregnancy. Ned Tijdschr Geneesk.d 1999; 143:1144-8 .
60 Runnebaum IB, Holcberg G, Katz M. Pregnancy outcome after
repeated blunt abdominal trauma. Eur] Obstet GYllecol
Reprod Bioi 1998; 80:85-6.
61 Pearlman MD, Fintinalli JE, Lorenz RP. Blunt trauma during
pregnancy. N Eng/ ] Med 1990; 323: 1609 - 13.
62 Poole GV, Martin IN Jr, Perry KG Jr, Griswold JA, Lambert CJ,
Rhodes RS (1996) Trauma in pregnancy: the role of
interpersonal violence. Am] Obstet Gynecol 174:1873-7.
63 Weiss HB. Songer TJ, Fabio A. Fetal deaths related to
maternal injury. ]ANlA 2001: 286: 1863-8.
64 Sugrue M, Kolkman KA. Trauma during pregnancy. Aust ]
Rural Health 1999; 7:82-4.
65 Rubovits FE. Traumatic rupture of the pregnam merus from
'sea t belt' injury. Am ] Obstet Gynecol 1964; 90:828-9.
66 Raney EH. Fetal death secondary to nonpenetrating trauma to
the gravid uterus. Am] Obstet GynecolJ970; 106:313-14.
67 Crosby WM, Costiloe JP. Safety of lap-belt restraint for pregnant
victims of automobile collisions. N Engl ] Mer! 1971; 284:632-6.
68 Pearce M. Seatbelts in pregnancy: above and below the bump,
not over it. BM] 1992; 304:586-7.
69 Pearlman MD, Klinich KD, Schneider LW et al. A
comprehensive program to improve safety for pregnant
women and fetuses in motor vehicle crashes: a preliminary
report. Am] Obstet Gynecol2000; 182:1554-64.
70 Williams JK, McClain L, Rosemurgy AS, Co lorado NM.
Evaluation of blunt abdominal trauma in th e third trimester
of pregnancy: m a ternal and feta l considerat ion s. Obstet
Gyneco/1990; 75:33-7.
71 Matthews CD. Incorrectly used seatbelt associa ted w ith uterine
rupture follo wi ng ve hicular co lli s ion. Am ] Obstet Gynecol
1975 ; 121:1115-\6.
72 Jud ich A, Kuriansky J, Engel.berg J et aJ. Amniotic fluid
embolism following blunt abdominal trauma in pregnancy.
Injury 199 8; 29:475-7.
73 Covington DL, Diehl SJ. Wright BD, Piner M. Assessing for
violence during pregnancy using a systematic approach.
Matern Child Health] 1997; 1:129-33.
74 Cokkinides VI, Coker AL, Sanderson M et al. Physical
violence during pregnancy: maternal comp li cations and birth
outcomes. Obstet GYl1eco/ 1999; 93:661-6.
75 Rachana C. Suralya K, Hisham AS et al. Prevalence and
comp li cations of phys ica l vio lence during pregnancy. Eur]
Obstrt Gynecol Reprod Bioi 2002; 103:26- 9.
76 Depal1ment of Health. Confidential en quiri es of maternal
deaths in the UK. London: Depal1ment of Hea lth , 2001
(http//www.doh.gov.uk/cmo/mdeaths.htm).
77 Drjfe J. Why mothers die. ] Roy Coil Physicians Edin 2005;
35:332 - 6.
78 Lipsky S, Holt VL, Easterling TR, Critchlow Cw. Impact of
police-reported intimate partner violence during pregnancy
on birth outcome. Obstet Gynrcol2003; 102:557-64.
79 Boy A, Salihu HM. Intimate pal1ner violence a nd birth
outcomes: a systematic review. Int ] Ferti! Wom ens Mer!
2004; 49: 159-64.
80 Soka l MM, Katz M, Lell ME, Fox A. Neonatal sLl lvival after
traumatic fetal subdural hematoma.] Reprod Med 1980;
24:131-3.
81 Stephens RP, Richardson AC, Lew in JS. Bilatera l subdu ra l
hematomas in a newborn infant. Pediatrics 1997; 99:619-21.
82 Pak LL, Reece EA, Chan L. Is ad verse preg nancy outcome
predictable after blum abdominal trauma 7 Am ] Obstet
Gyl/eco/1998; 179:1 140-4.
83 Becroft DMO, Gunn TR. Prenatal cranial haemorrh ages in 47
Pacific Islander infants: is traditional massage the cause?
NZ iVIed] 1989; 102:207-10.
84 Lieberman JR, Mazor M, Molcho J et al. Electri ca l acc iden ts
during pregnancy. Obstet Gynecol 1986; 67:861-3.
85 Ma zor M, Lieberma n JR. Abortion caused by electric curren t.
Arch GYllecol Obstet 1987; 241 :71-2.
86 Jaffe R, Fejgin M, Ben Aderet N. Fetal death in early
pregnancy due to electric current. Acta Obstet GYll ecol SClmd
1986; 65:283.

87 Mehl LE. Electrica l inju ry from tasering and miscarriage.
Acta Obstet Gynecol Swnd 1992; 71 : 118-23.
88 Yoo ng AF. Electrical shock susta ined in pregna ncy fo llowed
by placenta abruption. Postgrad Med ) 1990; 66:563 - 4.
89 Fatovich DM. Electric shock in pregnancy. ) Emerg Med
1993; 11:175-7.
90 Einarson A, Bailey B, ln oce ncion G et al. Accid enta l electric
shock in pregnancy: a prospective cohan study. Am) Obstet
Gyn ecoi l997; 176:678-81.
91 Rees WD. Preg nant woman struck by lightning. BM) 1965;
1: 103-4.
92 Chan Y-F, Sivasamboo R. Lightn ing accidents in pregnancy.
) Obstet Gynaecol Br Commonw 19 72; 79:761 -2.
93 Guha-Ray DK. Feta l dea th at term due to li ghtnin g. Am)
Obstet Gynecol 1979; 134: 103-5.
94 Fl a nnery DB, Wiles H. Follow- up of a survivo r of intra -uterine
light ning exposure. Am) Obstet GYl1eco11982 ; 142:238 - 9.
95 Pierce MR , He nde rson RA , Mitchell JM. Cardiopulmonary
arrest secondary to lightning inj ury in a pregnant woman.
Ann Emerg Med 1986; 15: 597-9.
96 Awwad IT, Azar GB , Seoud MA et a1. High-velocity
penetrating wounds of the gravid uterus: review of 16 yea rs
of civil war. Obstet GYl1eco l 1994; 83 :259-64.
97 O'Shaughn essy MJ. Conservat ive obstetric management of a
gunshot wound to the seco nd-trimester gravid uterus: a case
repon. J Rep rod Med 1997; 42:606-8.
98 Piers on R, Miha lovits H, Thomas L, Beatty R. Penet rating
abdominal wou nd s in pregnancy. Ann Ell1erg iVIed 1986;
15: 1232 - 4.
99 Grubb OK. Nonsurgica l management of pe ne trating ute rine
trauma in preg nancy: a case report. Am ) Obstet Gyn ecol
199 2; 166 :583-4.
100 Franger AL, Buchsbaum HJ, Peaceman AM. Abdomina l
gunshot wo und s in pregnancy. Am ) Obstet Gynecoi 1989;
160: 1124-8.
101 Lavin JP, Polsky SS. Abdom in al tra um a during pregnancy.
Ciin Perinotol1983; 10: 42 3- 38 .
102 Buchsbaum HJ, Staples PJ. Self-infli cted gunshot wound to
the pregnancy uteru s: report of two cases. Obstet Gyneco/
1985; 65:32S.
103 Sakala EP, Kort DD. Management of stab wounds to the
preg nant uterus: a case report and a review of the literature.
Obstet Gynecol Sum 1988; 43:319-24.
104 Rose PG , Strohn PL, Zus pan FP. Feto maternal haemorrhage
follo wing trauma. Am) Obsrer Gynecol 1985; 153:844- 7.
105 Reis PM, Sander CM, Pearlman MD. Ab rup tio place ntae after
auto accidents. A case-control stu dy. ) Reprod Med 2000;
45:6-10.
References I 197
106 Rogers FB, Rozycki GS, Osier TM et al. A multi-in st itutio nal
study of facto rs associated with feta l death in injured
pregnant pati ents. Arch Surg 1999; 134:1274-7.
107 Drost TF, Ros emurgy AS, Sherman HF et al. Major trauma in
pregnant women: materna l/fetal outcome. J Trauma 1990;
30:574-8.
108 Fries MH, Hankin s GOY. Motor vehicle accident associa ted
with minimal maternal tra uma but subsequent fetal demise.
AIIII Em Med 1989: 18:30 /- 4.
109 Evrard JR, Stmner WQ , Murray FJ. Fetal skull fracture from
an automobile accident. Am ) Forensic Med Patho l 1989;
10:232-4.
110 Hartl R, Ko K. In utero skull fra cture: a case report. ) Trauma
1996 ; 41 :549- 52.
III GUIlIl TR , Mo ra JD , Becroft DM. Co nge nital hydrocephalus
seco ndary to pre nata l intracrania l hae morrhage. Aust NZ )
Obstet Gyl/aecol 1988; 28: 197 - 200.
11 2 Robinson MJ, Cameron MD, Smith MF, Ayres AB. Fetal
subdural haemorrh ages presenting as hydrocephalus. Br lVIed
J 1980; 281:35.
]] 3 De Sousa CD, Clark T, Bradshaw A. Ante natally diagnosed
subdural hemorrhage in congen ita l factor- X deficiency. Arch
Dis Child 1988; 63:1168 - 70.
114 Shere r DM, Anyaegbunam A, Onyeije C. Antepartum
fetal intracranial hemorrhage, predisposing factors and
prenatal sonography: a review. Am ) Perinatol 1998;
15:431-41.
11 5 Weinberg L, Wyatt P, Busuttil A. Tra umatic intrauterine fetal
sp in al fracture follo wing seatbelt use: a case repon.
J Trauma 200 1; 51 : 1195- 16.
1/ 6 Fakhoury GW, Gibson JRM. Seatbelt ha za rd s in pregnancy: a
case report. Br J Obs tet Gynaeco/ 1986; 93:395-6.
11 7 Siddall-Allum IN , Hughes JH, Kal er S, Reginald pw. Splenic
rupture ill utero following a ro ad traffic acc ident. A case
report. Br ) Obstet Gynaecol 199 1; 98:3 18-1 9.
11 8 Litmanovitz I, Dolfin T, Arnon S et al. Feta l intrathoracic
injuri es following mild maternal motor vehicle acc id ent.
J Perill at ivIed 2000; 28: 158 - 60.
119 Griffiths M, Hillman G, Usherwood MM. Seat belt injUly
in pregna ncy resulting in fetal death. A need for
ed ucation? Case reports. Br J Obstet Gynaeco/ 199 1;
98:320- 4.
120 Yates PO. Birth tra uma to the verte bra l arte ri es. Arch Dis
Child 1959; 34 :43 6-41.
121 Ito Y, Ts ud a R, Kimura H. Diagnostic va lue of the
placenta in med ico-legal practice. Forensic Sci lilt 1989;
40:79 - 84.
I CHAPTER 11 I
SUDDEN UNEXPECTED DEATH IN
INFANCY: SUDDEN INFANT DEATH
SYNDROME OR SOMETHING ELSE?
Jean W Keeling
Introduction 198
The definition of SIDS 199
Epidemiology 201
Sleeping environment 203
INTRODUCTION
In the UK, SIDS became an option as a registrable cause
Sudden, unexpected deaths are a large and important
group of deaths in the post-neonatal infant age group
(1 month to 1 year). These deaths require a consistent high
standard of investigation, which goes well beyond
anatomic dissection and microscopic examination of tissue
samples from a few major organs. Among the deaths pre­
senting to the pathologist as sudden and unexpected will
be those that are fully explained by a recognized disease
process (see Chapter 12), some that are non-natural,
some explained by minor or long-standing disease and
others that occur in the absence of any pathological
abnormality. Cases in the last three groups are often
problematic and require the fullest, multidisciplinary
investigation.
The term 'sudden infant death syndrome' (SIDS) was
devised to be used in death certificates in cases of (sudden)
infant death in which no disease was identified at post­
mortem examination and death was deemed natural. It was
convenient for pathologists - no need to ascribe the death to
tenuous 'respiratory tract infection' or 'gastroenteritis' - and
it avoided the use of terms such as 'unasceI1ained', which
necessitated fulwer inquiries and an inquest or, worse, 'aspir­
ation of vomitus', which not only provoked an inquest
(non-natural death) but implied poor infant care practices by
parents or carers.
Pathological findings in SUDI 205
Death certification 218
References 219
of death in 1971, and the number of registrations increased
over the following decade as the option was taken up by
more pathologists. The International Classification of Dis­
ease separated 'SIDS' (789.0) from 'Sudden Death - cause
unknown' (795.0) in 1979; the latter term was included for
the first time in the previous revision in 1968.
Whilst the term 'cot death' has been used since the early
1950s and was refined in 1965 to encompass only unex­
plained deaths, I it was not until the mid-1970s that the
general acceptance of Beckwith's definition 'The sudden
death of any infant or young child, which is unexpected by
history and in which a thorough post-mortem examination
fails to demonstrate an adequate cause of death',2 enabled
meaningful comparisons between different studies. Clearly,
there are ambiguities that beset this definition: How sud­
den? Unexpected by whom? Which historical events
exclude deaths from the SIDS definition? What constitutes
a 'thorough' necropsy examination? What type or extent of
pathological abnormality comprises an adequate cause of
death?
Despite these reservations, this definition remains the
focus of our approach to the problem. However, the
importance of the immediate circumstances of death has
become increasingly appreciated. J Berry et al 4 found a
detailed history, particularly of the precise circumstances
of death, to be a most imp0!1ant pal1 of the evaluation of

sudden unexpected death in infancy (SUD!). The formal
consideration of the circumstances of death as part of the
definition of SIDS has been proposed,5 but not widely
implemented. The greatest benefit of the SIDS designation
has been that it has permitted much better estimates to be
made of the size of the problem than has hitherto been pos­
sible. The number of cases has been very much greater than
had been supposed initially and constitutes an important
public health problem.
There are, however, disadvantages in the use of the term
SIDS. One has been the development of an assumption that
this 'syndrome' has a single aetiology - 'died of SIDS' is an
unfortunate phrase seen in many publications. Another has
been a tendency among some in child care to stop thinking
about the problem because it is an accepted, registrable
designation. The term has been accepted uncritically and in
any particular case, until recently,6 few attempts were
made to unravel any predisposing factors. A fUl1her prob­
lem is the unwillingness of some pathologists to investigate
cases fully 'because there is nothing to find'! This is a most
unfortunate view as some of the specific causes of sudden
death, which may be missed by slipshod investigation, are
relevant to the medical management of other family mem­
bers or of siblings yet unborn. Of equal concern is that
some of these poorly investigated deaths may be the result
of homicide. Bajanowski et al 7 found 5 per cent of non­
natural deaths among 339 externally normal victims of
sudden infant death. Non-natural deaths accounted for 7.9
per cent of all deaths in the Confidential Enquiry into Still­
births and Deaths in Infancy Sudden Unexpected Deaths in
Infancy (CESDI SUDI) studies. 6
Gilbert-Barness and Barness 8 suggested that the use of
SIDS be repJaced by 'sudden infant death' with case-specific
qualification. Green 9 has argued for the abandonment of the
terms 'cot death' and 'SIDS', and for wider use of 'not ascer­
tained' on the death certificate. He points out that with the
considerable reduction in infant deaths following the suc­
cessful 'Back to Sleep' campaigns, the proportion of non­
natural deaths is likely to be higher among sudden infant
deaths at the end of the 1990s than it was 10 years earlier.
Meadow lO supports this view and considers that unexpected
infant deaths require a multidisciplinary investigation of
high standard and that the term SlDS should be more dosely
defined or else abandoned. The opposite view is put by
Limerick II who argues that the number of non-natural
deaths is small, and that there is a need for a dear category
of sudden, unexplained natural deaths. Her statement that a
case review following post-mortem examination would
avoid any need for the pathologist to use the term 'cause
unascertained' is, in my view, incorrect.
SUDI can fall outside the SlDS definition for reasons
related to the baby's family history, past medical history,
circumstances around the time of death, or because of the
results of post-mortem investigations carried OUt. 12 Some
of the inconsistency of terminology can be related to the
provision of inadequate information to the pathologist prior
The definition of SIDS I 199
to the commencement of post-mortem examination,13
although evaluation of the effect of bed sharing poses prob­
lems for individual pathologists. Fleming et al 14 emphasize
the role of the multidisciplinary team in the overall evalu­
ation of any SUDI, using a non-hierarchical grid.
Krous et al 15 have proposed a plan of subdivision of
SUDI!SlDS with precise definitions. They acknowledge the
problems of evaluation of cases after prolonged resuscita­
tion or incompleteness of investigations, both of which
introduce serious problems for reviewers.
This chapter will consider those questions arising out of
Beckwith's definition of SIDS. It will explore family back­
ground and discuss why prior events and some patho-logical
abnormalities require a different, sometimes more specific,
designation. It will also describe pathological criteria which
supp0l1 the SIDS categorization and discuss concerning fea­
tures in presentation or history and asses the appropriateness
and significance of pathological findings.
THE DEFINITION OF SIDS
How Sudden?
A very tiny minority of SUDI occur whilst the infant is
being observed. it is, however, an important group, as a
high proportion will be explained. This important fact must
be conveyed to the pathologist as additional investigations
are necessary. The three likely groups of underlying path­
ology which must be carefully considered are trauma, acci­
dental and non-accidental, cardiac pathology and upper
respiratory tract obstruction.
A larger, but still relatively small, group of babies are
those who collapse or die suddenly during the course of a
recognized acute illness, but before a specific diagnosis has
been reached. In this group, an explanation for death is
likely to be found if an adequate examination, induding
microbiology and investigation for possible genetic meta­
bolic disease, is carried out.
The majority of infants dying unexpectedly are found
dead in their usual place of sleep when the rest of the
household wakes. The CESDI SUDI studies 6 found that 64
per cent of deaths were discovered betvveen 05.30 and
11.30 hours. A further 17 per cent died during a daytime
sleep. Few deaths in that study occurred between 18.00 and
24.00 hours. Time of death should not be an excuse for dis­
missing the death as natural and for incomplete investiga­
tion of the death. Important, treatable pathology is found
in some infants in this group, and it is important that both
the parents and family practitioners are made aware of
such findings.
A group of deaths that do not fit easily within the defini­
tion of sudden are those in which the infants, found col­
lapsed, are maintained on life support systems for more than
12 hours. These should, in my view, be investigated with the
same vigour. Clinical management can result in findings at
200 I Sudden unexpected death in infancy
necropsy that are difficult to interpret and a full interven­
tional history is essential. Discussion with the medical staff
involved may assist evaluation of findings (see Chapter 19).
Clearly, there is no necessity to repeat those investigations
done in hospital, but care should be taken that the full range
of investigations is completed.
How Old?
National statistics define the infant period as from birth to
1 year, but deaths in the first week of life are generally birth­
related and not sudden. SUDI studies, anxious not to miss
cases, may set age limits widely, 1 week to 2 years in a UK
multicentre study.IG However, most SUD! occur within a
more restricted age range. Many consider 1 month to 1 year
to be appropriate limits. Most deaths occur prior to the
eighth month. Unexplained sudden deaths before 1 month of
age are unusual. They are discussed fuliher on p. 209. Sud­
den neonatal (0-4 weeks of age) deaths accounted for only
three per cent of deaths in a Scottish study. 17
Unexpected deaths over I year of age are also unusual
(three per cent). Most sudden deaths among infants over
6 months of age are likely to be explained. Fleming et al 6
found the median age at death of SIDS to be 13 weeks within
an interquartile range of 7 weeks, 6 days, to 21 weeks,
3 days. Aggregated data from several large studies derived
the 5th and 95th percentiles as 3 weeks and 9 months,
respectively. IS
Unexpected by Whom?
An important part of the definition of SIDS is the lack of
antecedent symptoms. This absence of symptoms requires
critical evaluation and is a forceful argument for the
involvement of a multidisciplinary team in the invest iga­
tion of SUD!. An experienced paediatrician or health visitor
may unearth symptoms whose significance was not appre­
ciated by carers or police. This might be because of lack of
experience of the carer or the non-specific or insidious
nature of signs and symptoms. The apparent lack of symp­
toms may, in a few cases, be explain ed by the velY rapid
course of the illness, particularly septicaemia or meningitis.
The lack of specific signs of illness in the infant may be
another contributory factor.
In some cases, the history is deliberately misleading, to
conceal either real or perceived inappropriate treatment of
the infant. Even after thorough, sensitive and timely ques­
tioning, it is apparent that some infants exhibit no or only
trivial symptoms. la A careful and thorough history is
particularly important to the pathologist to enable him or
her to include additional investigation pertinent to the
particular case 4 and in the overall evaluation of the death.
It is not infrequent for families to give a history of a recent
mild upper respiratory tract infection or of immunization.
Which Historical Events are Significant?
The most significant points to elucidate from the history
are previous infant deaths (explained or otherwise) in the
immediate or extended family, previous admission of the
infant to hospital and any documentation of prior injury to
the deceased or to siblings. These and other family events
are considered in more detail later in the chapter.
What is a Thorough Necropsy Examination?
The purpose of necropsy examination in SUDI is to distin­
guish between natural and non-natural death, to elucidate a
cause of death if one is present and to further the under­
standing of sudden death in the infant period. Published
guidelines include those of Wigglesworth et al,J9 the Royal
College of Pathologists,20 ValMs-Dapena et al,21 Krous 22 and
Howatson.23 It is impOltant that SUDI are investigated to a
consistently high standard. This includes the completion of
investigations even in those cases where major pathological
abnormality is discovered early in the investigation. How
else is the pathologist to understand the significance of his
findings in other cases? The need for consistent and thor­
ough investigation of the family background as well as the
post-mortem examination itself were emphasized by an
intercollegiate working group24 and are supported by Landi
et al, 25 who found that more and appropriate investigations
were done in cases certified as SIDS than those considered
explained. Centres unable to offer full radiological and
microbiological support as well as a high technical standard
for processing histological samples, including neuropatho­
logical samples, should not investigate these deaths. Sadler 26
emphasizes the importance of early and thorough microbio­
logical investigations. The relative contributions of different
components of the investigation of SUDr are evaluated by
Berry et al 4 and Mitchell et alY
Toxicological investigations have to be justified in
many jurisdictions because of a low return of positive find­
ings. Langlois et al 28 detected drugs in 16 per cent of SUD!
investigations with drug administration an important, but
unsuspected, factor in the death in 2.6 per cent. Arnestad
et al 29 found that toxicological investigations made a
positive contribution to the evaluation of four per cent of
SUD!.
What is an Adequate Cause of Death?
The interpretation of the pathological findings in respect of
adequacy of a cause of death is frequently difficult in SUD!.
The problem does exist in relation to deaths at other ages,
but because of the prevalence of degenerative diseases and
the relative infrequency of tmly unexplained sudden death
in older children and aduIts 30 it is less discussed. Over­
emphasis of minor pathological findings, particularly in

poorly investigated cases, is a particular problem in this
age group.4,]I,]2 The significance, or otherwise, of partiClI­
lar pathological findings is discussed later in the chapter.
Multidisciplinary Case Review
A case review following a SUDI is the usual practice in many
areas]] and is rapidly becoming the norm in many countries.
Review groups involved the family doctor and health visitor
(with midwifery input, if appropriate), a researcher, when
formal parental interviewing is normal practice, a patholo­
gist and a paediatrician. Its purpose is to examine both risk
factors and the actions of parents or health professionals,
which might usefully be modified either in respect of a par­
ticular family or in professional practice.
Review of the pathology findings encourages consistent
investigation and goes some way towards introducing uni­
formity of interpretation of pathological abnormality. It
forms the basis of discussion with each family. It is the
ideal venue to develop appropriate local strategies to sup­
port bereaved families. 24
EPIDEMIOLOGY
Epidemiology of SUDI and SIDS
It is through its epidemiological features that SIOS is best
defined, but the interrelationships of these are complex.
Investigation of these complexities demands that study
methodology is clearly set out and the objective well
defined 34 Many factors are as relevant today as they were
in the earliest studies, but there have been some changes
over the past five decades. One factor that has resulted in
changes in the epidemiology of SIOS is better case defin­
ition, as efforts are made to reach a consensus about what
constitutes an adequate explanation for death. However,
there is no doubt that changes have occurred; seasonality
(i.e. a higher incidence in winter months) is currently much
less marked than two decades ago, whilst maternal smok­
ing exerts a stronger effect. 35 - J7 Most cases occur between
6 and 20 weeks of age with a median at around 12 weeks.
There is a male predominance and a continued correlation
with higher birth order and low maternal age. The associ­
ation with low maternal age and birth order has become
stronger in some studies. 38 Low birth weight and pre-term
delivelY in the index and previous pregnancies 39 and twin
pregnancy are recognized risk factors, as is maternal smok­
ing during pregnancy.37 Blair et al 37 found that pre-term
delivery among victims of SIOS increased from 12 per cent
(Q 34 per cent over a 20-year period. Malloy40 looked
at SIOS in pre-term infants and concluded that the
association was the result of socioeconomic rather than
medical characteristics. Mothers of SIDS infants have more
Epidemiology I 201
miscarriages and still bilths 41 and shorter interpregnancy
intervals than control subjects. 42 Other pregnancy-related
problems such as urinary tract infection and anaemia are
weaker associ-ations. An elevated maternal serum alphafe­
toprotein level in the second trimester of pregnancy is
associated with a higher risk of SIOS .43 Maternal recre­
ational drug use is also a risk. 6,17 In Bristol, UK, the propor­
tion of SIOS victims from deprived socioeconomic
backgrounds rose from 47 per cent to 74 per cent over a
20-year period. 37
Simultaneous SUDI in twins is most unlikely to be
SlOS. It is an unusual event, but likely to have an environ­
mental cause. A detailed death scene investigation is
mandatory. 44 Sequential SUD! in twins is likely to result
fro m infection.
SIOS victims have lower Apgar scores at birth than con­
trol infants and a re more likely to be admitted to a special
care baby unit. They are also more likely to have dysmor­
phic features, malformations and positional deformities.
Low socioeconomic class, lower parental education level
and poor housing are consistent findings in the families of
SIDS infants. The increase in unemployment and of cohab­
itation makes it difficult to compare CUlTent and earlier stud­
ies, but the CESDI SUDr studies clearly demonstrate that,
irrespective of the parameters used - occupation, family
income, receipt of income support or employment status ­
an excess of SIOS deaths is seen in families with socio­
economic problems. 17
Predicting SIDS - and What Else?
With the recognition of pregnancy and perinatal factors asso­
ciated with SIDS, scoring systems have been devised with the
aim of identifying those infants at high risk of SIDS with a
view to intervening, in the forms of family support, education
and surveillance aimed at prevention of SIOS.45,46 Much
resource/research has been put into a variety of schemes
based on sCOling systems or targeted at families where SIDS
had occurred, with only modest return. A much more suc­
cessful intervention, widely applied, has been the 'Back to
Sleep' campaign.
SIOS scoring systems are not specific but identify babies
with other problems such as respiratory infection, the need
for hospital admission and non-accidental injury46 and
explained infant deathY Some SIOS risk factors are also
predictors of infant homicide.48
Recent Medical History in SUDI and SIDS
Interrogation of parents following sudden infant death fre­
quently elicits a history of symptoms of illness in the baby.
It is important to remember that these parents are likely to
be searching for something to which they can attribute
blame for the death, or which might have provoked action
202 I Sudden unexpected death in infancy
on their part and, perhaps, have averted death. It is also
important to remember that young infants often suffer a
succession of minor ailments. This background of frequent,
minor morbidity must be taken into account when conside­
ring the relevance of reported symptomatology.
Stanton et al 49 found that snuffles and colds were
reported frequently by both SUD! and control groups.
However, they found that more serious respiratory symp­
toms, such as cough or wheeze, were more frequent in the
SUD! group. Infants who died were also more likely to have
exhibited, in the day or so prior to death, unusual drowsi­
ness, irritability or excessive crying. They were also more
likely to be off feeds, pyrexial or sweaty. Almost 60 per cent
of SUD! victims had symptoms in the 48 hours before death ,
half of which were considered major. More babies who died
than control babies had seen a doctor both within 24 hours
of death and earlier in the course of perceived illness. Half of
the infants with system-related symptoms had evidence
of significant terminal disease at necropsy whilst only one­
quarter of those with non-specific symptoms had signifi­
cant pathology. 16 In that study, about one-t1Fth of victims of
unexpected deaths (no recognized symptoms) had evidence
of serious acquired disease at necropsy.
In another study, babies dying unexpectedly in the infant
period were more likely to have had minor signs of illness in
the previous seven days than control infants and to have
been seen by their general practitioner. 18 This study found
no difference between babies who died and control infants
with respect to major signs of illness. Stanton and OakJeyS°
found that SUD! victims were more likely to have been
admitted to hospital previously and required longer admis­
sion than control (surviving) babies. Although eight infants
in their study were admitted to hospital because of suspicion
of abuse, no babies had unexplained apnoea, but 9 of 71 suf­
fered loss of consciousness - repeated in three cases.
In New Zealand, SIDS babies were found to have more
severe illness resulting in more GP or hospital contact than
control infants.51 The clinical signs that prompted parental
action were not specified. A more recent study47 found that
the clinical characteristics of explained SUD! and SIDS vic­
tims were similar. Apparent life-threatening events
(ALTEs), defined as an episode in which the infant became
apparently lifeless, were reported by parents of both groups
(9/61 , 14.8 per cent, and 37/317. 11.7 per cent) and were
more frequent than among control subjects. However. only
two-thirds of these episodes were brought to medical atten­
tion and less than one-halfwere seen in hospital. Infants in
both groups were more likely to have been unwell than
control infants in the 24 hours before death. Signs seen
more frequently in the SIDS group were drowsiness.
wheezing and reduced fluid intake in the 24 hours prior to
death. Neonatal problems and hospital admissions were
much more common in the SUD! group than among SIDS
babies.
In the Netherlands, signs of iJJness were more frequent
in explained SUDI than SIDS or borderline SIDS. 52
Table 11.1 Factors in the history causing concern in SUDt
Many GP/accident and emergency attendances without clear
evidence of illness
Apparent life-threatening events (ALTEsl. especially if more
than one and starting at less than 4 weeks of age and if
unwitnessed or if the same person finds the baby each time
Age > 1 year
Epistaxis
Seizures - especially unexplained
Death occurs during the day when the baby was apparently
well earlier
ALTE in siblings
Previous SUDI in sibling
Death on anniversary of previous SUDI
Parental involvement with the media after child's death
Adverse Historical Factors (Table 11.1)
The occurrence of a previous infant death in the family or
in the extended family should alert the pathologist to the
possibility of an inherited disorder. which might not have
been fully investigated in the earlier death, as well as to the
possibility of non-natural death 10.53 More than half of
Meadow's non-natural deaths were initially certified as
SIDS deaths although the thoroughness of investigation is
not set out. A detailed background investigation is required,
and it is helpful to retrieve the necropsy and police reports
on the previous death prior to commencing examination of
the second as well as accessing any clinical notes.
A history of unexplained collapse with pallor or cyanosis
and apnoea is, for me, a cause for concern in SUDI babies.
Concern increases when acute AlTEs have occurred on more
than one occasion 10,54 and when episodes commence after
the neonatal period. 55 A further cause for anxiety is when
investigation of these episodes in a hospital setting fails to
find an underlying problem, particularly if ALTEs cease in
hospital. 54 .56 It is important that the circumstances under
which the episodes occurred are fully investigated. Ariagno
et al 57 found that AlTEs in pre-term infants occurred earlier
and were more likely to be witnessed by professionals (54 per
cent) than those in term infants (23 per cent). Similar episodes
described in siblings is a further concem. 53- 55 Byard and
Krous 58 list potential natural causes of AlTEs and highlight
the need to consider imposed suffocation as a basis for such
events. In the past, such events have even been called ' near­
miss cot deaths', but currently many are thought to be the
result of imposed upper airways obstruction. 10,55.59,60
Such a history is most unusual in what, both epidemio­
logically and pathologically, is a typical SIDS death.
Having been somewhat sceptical about the frequency of
imposed airways obstruction as the cause of ALTEs, per­
sonal experience of repeated infant death with or without
AlTEs has changed my view. I wonder whether the history
~ -_.­

of even a single unexplained ALTE should p reclud e use of
the SIDS designation. Unless death was clea rly explained,
fulfilling a sp ecific diagnosis, then, in these circumsta nces,
death sho uld be considered unascertained.
Another unusual symptom in the infant perio d is epi­
staxis, defi ned as bl eeding from both nostrils.55 This should
not be confused with blood-streaked mucus from one nos­
tril, usually the resul t of self-inflicted injury. Epistax is is
particularly concerning when accompanied by choking
and/or tra nsient change in breathing patte rn as it raises the
possibili ty of imposed ainvays obstruction. 58 ,61 The pres­
ence of blood-tinged mucu s at the nostrils in a su dden
baby death is to be treated suspiciously.
A hi story of seizures is an indication fo r particularly
careful central nervous system (CNS) exa minati on, which
can be done only fo llowing fixation of the brain. Detailed
information about pregnancy and the neonatal peliod is
essentia l when seizures have been reported. Howeve r, the
possibili ty of a non-natural process should not be forgo tten,
Seizures (due to induced cereb ral hypoxemia) are probably
the most common presentation of fabri cated illness 62
Necropsy examination should include careful h arves ting of
sampl es for toxi co logical examination ,63 particul a rly urine,
blood and vitreous fluid (only if the eyes do not need to be
examined histo logica lly).
Unusual findin gs amo ng parents of babies who had
been smothered were som atizing diso rders, involvement
with arson or medical litigation and actively seeking media
publicity in relation to their infant's death.1O
SLEEPING ENVIRONMENT
Infan t care practices are culture dependent a nd have
evolved gra du a ll y over very ma ny years as a result of fam­
ily experience and loca l contacts. In recent years, opinions
from many sources have been wid ely dissemin ated; fash­
ion, fads and consumerism are pressures on you ng, often
inexpe rienced, mothers who compete with advice from
health professionals.
Good practice
The reco mmend ed sleeping environment for young infants is
a cot with firm mattress, with no lise of soft pillows or
bumpers, Whilst a firm, permeable pillow may have had merit
when used wi th an impermea ble mattress cover when many
infants were put down prone,64 it is less useful when most
babies sleep supine. Preferred cot bedding is a sheet wit h
lightweight blan ket(s), which can be tucked in at the sides,
rather than duvets or sleeping pods. It is recommen ded that
the enviro nment be neither too hot or cold (at about 16-18°C)
and well ven tilated. Pl acing a sm all infant at the end of t he
cot (foot to foot) rather than in the middle, to avoid burrow­
ing under covers and overheating is current advice.65
Sleeping enviro nment I 203
Placing the infant in the 'supine' sleep ing position 66 and
avoidance of prone and, now, 'side sleeping'65 is advised,
Avoidance of infa nt restrai nts design ed to maintain the
initial sleep ing position is current best practice. More
details abo ut safe sleeping and appropriate avoidance of
co-sleeping have been formulated. 67 Co-sleeping is con­
si dered further on p. 204.
Encouragement of bed sha ring app lies only to non­
smoking parents who refrain from alcohol and social
drugs. Excessive tiredness and use of prescribed sedatives
or tranquillizers a re also contraindications to bed sha ring
as they may result in overlaying of the baby. Sleeping in a
cot in the parental bedroom reduces the risk of SIOS. 68
Suboptimal sleeping environments
Stanton 69 drew atte ntion to overheating of sleeping infants
as a lisk factor in cot death. He found a number of contribu­
tory fact ors likely to promote overheating. These includ ed
maternal atti tude, especially duri ng co ld weather, use of
daytime, outdoor clothing in bed, swa ddling in blankets,
use of folded adult bed blankets on cots, use of bonnets or
hoods and a hot environment, for example central heati ng
on a ll night. In some cases, pyrexial illness in the infant also
seemed to compound the problem, Excessive bedding was
found to be an indep endent risk factor in south-west
England. 70 Wigfield et al 71 fo und that a low outdoor tempera­
ture sti ll prompted mothers to increase infa nts' bedding
when room tempe rature was adequate a decade after Stan­
ton's observations were made.
The associatio n of prone sleeping position and cot death
has been reported in the case-control studies of infant deaths
since 1970.72 Reports of the association accu mulated with
increasing importance of risk from this practice from several
centres, as other risk facto rs for SlDS have changed?O,73-76
Further supp ort for the association has been the marked
reduction in SlDS followi ng publicity campaigns to increase
professional and public awa reness. Mitchell et al 77 consider
that the practice of 'Back to Sleep' accou nts for about 50 per
cent of the fa ll in SlDS deaths in New Zealand ,
Interactions between sleep ing position and bedding have
been explored. Pron e sleeping and high tog values of bed
coverings were independent risk facto rs in a UK Study.70 The
use of quilts was a risk factor for SlDS among supine or side­
sleepers in Tasmania. 78 An additional adve rse factor for
prone-sleeping infants may be the adoption of face straight
down or face near-straight down positions. These positions
have been shown to introduce an element of airways
obstruction that makes death more likely in some infants,79
Further examinati on of the infants' sleep ing environ­
men t demon stra tes that an increased risk is related to side­
sleepin g pOSition and to head covering as an impediment
to maintaining thermoneutrality.76,8o Living in a household
where carers smoke is a risk factor for SIOS. This risk is
related to the number of cigarettes smoked in the local
204 I Sudden unexpected death in infancy
environment. 35 ,75,76 Sleeping in a room separate from par­
ents was a risk factor for SIDS in a New Zealand study.
Inappropriate sleep surfaces for infants
The use of unstable sleep surfaces is unsuitable for small
infants. The danger of placing infants prone on surfaces such
as adult-size water beds was flagged up by Ramanathan
et al. 81 They reported that 6 out of 34 SIDS cases investigated
over a short period had been found face down on water beds.
Bass 82 reported 13 infant deaths on water beds and drew
attention to the water bed design (free flow, both original and
modified designs). Eleven of these cases had been originaJly
certified as SIDS. Deep-pile sheepskin rugs may present a
similar, unacceptable hazard to young infants.83 In these situ ­
ations, the infant is not able to exercise the head movement
possible when prone on a firm surface.
The danger to infants of the use of polystyrene-filled
bean-bags as a resting or sleeping place is clearly
recorded. 84 ,85 At least 35 deaths have taken place on these
cushions, the majority being initially registered as SIDS.
Babies were predominantly placed face down on the soft,
deformable surface and found dead after an interval of
0.5-10 hours. Thirty-four of the 35 infants were below the
median SIDS age of 13 weeks. Thirteen were only 1 month
of age. A combination of factors is thought to be respon­
sible for these deaths, including rebreathing in a confined
area, obstruction of nose and mouth, altered permeability
of cushion coverings by wetting with secretions, hyper­
thermia and the instability of the bag.
The use of 'V'-shaped pillows for young infants should
also be avoided. 86 Movements during sleep can result in
wedging in the pillow angle from which the infant cannot
retreat and obstruction of the upper a irway can supervene.
Other unsafe sleeping environments
Other unsafe sleeping environments are some types of
swing-cradles87 and cots with ill-fitting mattresses,88-90
which may permit wedging of the baby between mattress
and cot frame with impedance of respiratory movement. The
continued use of broken cots also constitutes an unaccept­
able hazard for infants,91 who can become entrapped
through gaps between broken or loose cot components.
Parental modification of cots can also be problematic. Byard
et a1 92 reported positional asphyxia in two infants whose
parents had introduced an additional mattress into mesh­
sided cots as the original mattresses were thin. This had the
effect of raising the sleep surface so that it abutted onto the
mesh section of the side-walls. This area is much more read­
ily distensible by the infant and permitted fixation of the
head between cot wall and mattress. Inappropriate cot
design is an additional hazard. Bass 88 reported 10 asphyxial
deaths seen over an I8-month period where cot design was
inappropriate for young infants, who were able to push their
heads between cot slats or rails. Smialek et al 89 and Gilbert­
Bamess et aJ83 have described similar cases and a correspond­
ing problem when babies sleep on adult beds. Deaths in
similar circumstances were still occurring 12 years later
because families still used old or broken cotS.93
The use of bedside cots, with bars along only one long
side of the cot, introduces similar potential for the infant to
become wedged between the cot and the parental bed. 94
Bed rails, designed to stop young children from falling out
of toddler or adult beds, pose a similar hazard for infants
less than 1 year of age. 95 There is a danger of strangulation
among older infants who can pull themselves up using cot
sides when loose clothing or dummy cords catch on cot
p rotru bera nces, 83,88,89,96
Byard et al 90 draw attention to the risk of asphyxial
death when infants are left sleeping in car seats, baby
bouncer or pushchairs (stroller-prams)97 with ill-fitting
restraints, resulting in suspension of the infant. There is a
danger of asphyxia when impermeable mattress coverings
in cots are loose rather than being closely fitted. This
applies even when babies are very young and consequently
less mobile.
Some hazardous sleeping environments are likely to be
appreciated only when timely death scene investigations
are conducted. 98 Excessive room heating and probable car­
bon monoxide or carbon dioxide poisoning were problems
discovered by Bass et a[3 when thorough investigations
were done. Leaving infants sleeping in poorly ventilated
vehicles on sunny days poses a similar hazard.
The possibility of accidental asphyxia, when mothers,
who are very tired or sedated, have been breast-feeding in
bed, has been explored by Bass et al 3 and Byard. 99 The for­
mer considered that 3 of 26 SIDS cases resulted from acci­
dental asphyxia in this situation. The latter found that 3 of
28 infants who died unexpectedly during a I-year period
in south Australia were in bed with their mother, who had
fallen asleep during breast-feeding. Although 'overlying' is
no longer considered a major cause of SUDI, circumsta nces
may dictate its serious consideration in some cases. 83
Co-sleeping
Bed sharing with non-smoking parents is enco uraged by
many health carers as a means of facilitating breast-feeding,
despite a small increased risk of death among co-sleepers. 100
UNICEF 67 stresses the need for a clean, firm mattress, avoid­
ance of overheating from excessive clothing and hazardo us
situations which might lead to wedging or falls, as well
as advising against bed sharing during maternal illness.
Concern is expressed by some about the increased propor­
tion of SUDI found after a co-sleeping night. Luke 101 drew
attention to this problem when he found that an increasing
proportion of SUDI within his jurisdiction occurred following
a co-sleeping night. This was four times higher than usual

practice 76 in a Scottish study. The increase in co-sleeping on
the night of death might be a marker of minor illness in the
baby, well recognized in SIDS victims, but the spectre of
positional asphyxia cannot be dismissed. Unusual bed shar­
ing was also reported in another (small) study52
lt is most important that the sleeping environment is
examined with great care when an infant is found unexpect­
edly dead in this situation. There are two potential problems,
asphyxiation and overheating. As well as the precise pos­
ition of the infant - prone/supine - the level of the infant in
relation to the parents and covers and the llse of pillows and
bed coverings, particularly heavy duvets, should be explored.
The state of the mattress and its coverings, particularly the
presence of blood-stained secretions and evidence of sweat­
ing, as well as the relative weights of the parents need to be
considered. This is in adclition to a clear history about
drug/alcohol ingestion and levels of tiredness. Parental alco­
hol consumption and co-sleeping on the night that the baby
died were frequent findings in two Scandinavian stud­
ies,I02,103 although this was not a factor in one American
study.98 I find that lack of detail about many of these factors
hampers evaluation of individual cases. Rushton 104 has
expressed similar concerns and draws attention to the
inevitable lack of important detail related to relative pos­
itions of parents and baby during sleeping hours. The loca­
tion of fixed lividity may be informative in some cases.
Sleeping on a sofa with an adult is a particularly haz­
ardous situation. l05 In the CESDI SUDI studies, this factor
contributed the greatest risk, with an odds ratio (OR) of
31 :25. It was confirmed in a Scottish study (OR 66.9, 95
per cent confidence interval [CI] 2.8-159.7) 106 and
re-emphasized in a more recent study from BristoL 37 In my
experience, shared by other pathologists, risk is enhanced
when alcohol has been consumed by the carer. 83
PATHOLOGICAL FINDINGS IN SUDI
As the number of unexplained, presumed natural deaths in
the infant age group has fallen during the last decade, so the
proportion of explained deaths in medicolegal autopsies
might be expected to rise, Although the increased range and
rapidity and improved accessibility of di,agnostic tests may
be expected to enable a pre-mortem diagnosis to be reached
in more babies. In a UK multicentre study, 1976-79, 18 per
cent of victims of unexpected death had evidence of serious
acquired disease,16 although this may not have explained
death in every case. In a retrospective study by Czegledy­
Nagy et aI, 107 there was an explanation for SUDI in 20 per
cent of cases. Nine per cent were non-natural deaths. Only
nine per cent of SUDI in an Adelaide study were thought to
be unexplained. 108 Subjective interpretation of pathological
findings is, to some extent, inevitable. The CESDI SUDI
investigations during 1993-96 emphasize the effect of sub­
jectivity in the interpretation of findings. In that study, only
17.8 per cent of the deaths were considered explained after
Pathological findings in SUDI I 205
review of pathological findings by an expert panel,
although death was considered explained by the examining
pathologist in 30 per cent of cases. 4 Parham et al 109 found
previously unsuspected pathological abnormality in more
than half of babies less than 6 months of age in a review of
sudden deaths investigated by the Medical Examiner's
Office; death was considered to be explained in 35 per cent
of cases. Only 39 per cent of SUDI were considered typical
SIDS cases in a South Australian study, no although a further
21 per cent were deemed 'undetermined'. Knowledge of the
circumstances of death may alter the pathologist's expect­
ation of positive pathological findings, but facilities for full
investigation should be available at the outset.
The findings following the sudden death of older infants
and children are discussed in Chapter 12.
The Pathology of Explained Natural Death
The range of pathological findings encountered in SUDI are
shown in Table 11.2111 (see also Byard and Krous 58 ). Patho­
logical findings in SUDI are discussed in detail by Byard
and Cohle, 112 There are no specific pathological markers for
SIDS despite the best efforts of many investigators over the
years.
INSTANTANEOUS DEATH
When infant death is instantaneous, then the type of patho­
logical findings that are likely come from a more restricted
group. Most of the deaths will have a cardiovascular cause.
Cardiac malformations, particularly ductus-dependent
abnormalities, are the most common group. One-third of
babies with cardiac malformations were awake at the time of
collapse in one study. 113 These are particularly common in the
first month of life. Nevertheless, infants with some malforma­
tions that are particularly associated with sudden neonatal
death, such as hypoplastic left heart syndrome, do survive
into the post-neonatal period. A surprising number of infants
with isolated ventricular septal defect die suddenly.114 Up to
one-third of affected children have arrhythmias. Detailed
examination of the cardiac conduction system has shown
fibrous interruption of conduction tissueY5 Sudden death in
infants who have undergone surgical correction of congeni­
tal heart disease (CHD) some months previously demands
detailed examination of the cardiac conduction system. The
distribution of the cardiac conduction tissue is frequently
aberrant in CHD, and inflammation around suture material
can have fatal consequences. A range of other cardiac pathol­
ogy has been identified in SUDI llG but, in an individual case,
requires careful consideration of its relevance to causation.
Coronary arteritis, usually a manifestation of Kawasaki's
disease, is encountered in the infant age group, more com­
monly in infants over 6 months of age. ll7 As well as diffuse
thickening of the coronary arteries (Fig. 11.1), aneurysms
may be found in the proximal parts of the arteries. Other
206 I Sudden unexpected death in infancy

Table 11.2 Pathalagical findings in SUDI in different

circumstances af death (after ref 117)

Instantaneaus death
Natural causes
Cardiac malformation
Coronary artery anomalies
Coronary arteritis
Myocarditis
Cardiomyopathy
Dysrhythmia
Acute laryngitis
Laryngeal cysts
Laryngeal papilloma
Laryngeal/tracheal developmental anomalies
Acute intestinal obstruction, e.g. volvulus
Non-natural causes
Accident
Non-accidental injury
Foreign body in airway

Rapid death, illness recagnized

Respiratory infections, both upper and lower Figure 11.1 Heart of an infant with Kawasaki disease; the
Men ingitis/men ingoencephal itis coronary arteries are prominent and there is aneurysmal
Septicaemia dilatation of the proximal segment of the left coronary artery.
Gastroenteritis
Other infections (including viral)
Reye/Reye-like syndrome
Intracranial haemorrhage
Congenital adrenal hyperplasia

Faund dead in cat

Natural
Congenital heart disease
Respiratory infection
Men ingitis/men ingoencepha litis
Septicaemia
Seizure disorders
Unexplained (up to 80% of deaths in this circumstance)
Non-natural
Suffocation
Other non-accidental injury
Intoxication

Figure 11.2 Histiocytoid cardiomyopathy.

coronary artery pathology, such as anomalous OrIgIn and

bridging of arteries, is also seen l18 and is described and
illustrated in Chapter 12. Cardiomegaly may alert the
pathologist to a cardiovascular cause of death. Ischaemic
myocardial damage is seen in about 50 per cent of cases.
Coronary arterial embolization is rare in early life, but para­
doxical embolism with myocardial infarction is described in
the neonatal period 119,120 and in older infants, when it is
most commonly a complication of endocarditis. 121
Several cardiomyopathies can be fatal during infancy. 122
Hypertrophic obstructive cardiomyopathy can present as
sudden infant death 123 (see Figs 8.12, p. ]56 and 12.2, p.
228). There is disproportionate hypertrophy of the septum,
and in fatal infant cases hypertrophy is severe. 124 The diag­
nosis is an important one in view of the findings that most
cases presenting in early life are familial. 125 Histiocytoid car­
diomyopathy, which may be visible through the ventricular
endocardium as irregular, cream or yellowish patches and as
pallor in the underlying myocardium, always requires histo­
logical confinnation (Fig. 11.2). It has been found in both
anatomically normal and malformed hearts. 126,127

Figure 11.3 Subend ocardial fibroelas tosis affecting the left
ventricle. Sudden death. No evidence of a more specific disorder.
A number of cardiomyopathies that are fatal in early life
exhibit subendocardial Iibroelastosis (sub-EFE), usually in
the left ventricle (Fig. 11.3), but sometimes affecting all
chambers. The majority of these are plimary, 113 but sub -EFE
may be a manifestation of a number of conditions includ­
in g glycogen storage disease type II (Pompe) and mitochon­
drial myopathies. It is important to culture fibroblasts and
reserve snap-frozen samples of fascia lata or pericardium so
that the primary disorder can be elucidated. The dysrhyth­
mias produced are va rious, but co mplete heart block is com­
mon. The mothers of some of these infants have connective
tissue disorders; investigation of the mother and close sur­
veillance of subsequent pregnancies may be important.
A variety of dysrhythmias are associated with sudden
deaths - long QT synd rome is the best recognized. 128 Both
dominant and recessively inherited forms have been
described and gene abnormalities in at least six different
loci have been iden tified. 129 Schwartz et al 130 recorded
electrocardiograms (ECG s) in the first week of life on more
than 34 000 babies. Twenty-four infants died suddenly in
infancy, of whom 12 had prolongation of the QT interva l.
No abnormalities were seen in the ECGs of 10 explained
deaths. Arnestad et al 131 found molecular abnormalities
likely to be associated with sudden death in 19 of 201 (9.5
per cent) cases of sudden infant death in a retrospective
study. Whilst mutations can be defined in formalin-fixed,
paraffin- embedded t issue, frozen sampl es are much easier
to work with.132 Most do not have structu ral abnormalities
in the cardiac cond uction system. Sometimes dysrhythmias
occur in malformed hearts, particularly those with large
ventricular septa l defect l1 4 or conotruncal problems.
Pathological findings in SUDI I 207
Figure 11.4 Larynx opened posteriorly; there is a cyst in the
anterior wall at the narrowest po int of the airway.
Mucous cysts or papillomata in the laryn x can achieve
sufficient size to obstruct the airway (Fi g. 11.4). A lingual
thyroglossal cyst can also result in ainllJay occlusion. 133, 134
Tracheomalacia can produce intermittent (and fatal) ailways
obstruction. 135 Posterior lary ngeal cleft (see Fig. 10.15, p,
194) predisposes to massive aspiration. In sum, there may
be evidence of prior asp iration in the form of gran ulomata.
It is important to remember that foreign bodies may have
been removed during attempted resuscitation.
Acute intestinal obstruction, such as sma ll intestinal
volvuluS136 (Fig. 11.5), is another cause of very rapid
demise. Sudden death is rep0l1ed complicating large intes­
tinal volvulus l3 7 and intussusception in infants.138,139
RAPID DEATH DURING RECOGNIZED ILLNESS
When death occurs rapidly in an infant with a symptomatic
illn ess, the most likely pathology is infection. It is important
in these circumstances that every effort is made to identify
the organism both by culture and by using available molecu­
lar methods. Few infections have pathological findings suffi­
ciently specific that a precise diagnosis can be reached on
morphological and microscopic findings alone. Histological
evidence of infection can be focal but nevertheless signifi­
cant, for example in the myocardium, where wide sampling
208 I Sudden unexpected death in infancy
Figure 11.5 Female infant, 3 weeks old: sudden death. There is
volvulus of the whole of the small intestine.
is appropriate,140 otherwise the abnormality may be over­
100ked. 141 Some infants dying suddenly with myocarditis are
asymptomatic 142 or exhibit non-specific symptomatology.14J
Evidence of viral aetiology may become more frequent with
the availability of molecular techniques. 144
Another important group of diagnoses in this type of
death are genetic metabolic disorders (GMDs), particularly
the ~-oxidation defects. Profound hypo glycaemia and
hyperammonaemia may be triggered by relatively trivial
infection. Although classically accompanied by a fatty liver,
steatosis may not be apparent to the naked eye. It may also
be missed on haematoxylin and eosin (HE)-stained sections
as the fat is microvesicular (see Fig. 12.21, p. 247). Frozen
sections of liver, kidney and heart are useful for screening.
Losty et al 14S dispute this, but the management of their case,
not a typical SUDI, was probably responsible for lack of
steatosis. The most common of the ~-oxidation defects in
Western Europe is medium-chain acyl-CoA dehydrogenase
(MCAD) deficiency with a heterozygote frequency of 1:40 in
Birmingham 146 but only 1 :74 in the west of Scotland. 147
Other GMDs are individually rare but collectively common;
Fitzpatrick l48 describes their presentation in early life. Most
centres are unlikely to be able to offer the full range of
investigations. 149 Lundemose et aliSO suggest using the
[9,1O- JHJ myristic acid oxidation assay in cultured fibro­
blasts as a method of identifying at least nine disorders of
fatty acid oxidation. They found three cases and two car­
riers of the G985 mutation among 79 sum using this
method. More extensive investigation of GMDs should be
Figure 11.6 Total anomalous pulmonary venous drainage,
sudden collapse at 4 weeks of age. The aberrant pulmonary vein
runs through the diaphragm to join the portal venous system.
contemplated when there is a second sum or parental con­
sanguinity. ~-Oxidation defects of all types may account for
up to five per cent of sum. 151 A test for the most common
mutation (G985) is readily available. A toxicological screen
is warranted when hepatic steatosis is found.
Congenital heart disease is less common in this group but
must be carefully sought. Anomalous pulmonary venous
drainage may present with non-specific signs of illness. IS2 It
is worth remembering that this is the structural cardiovascu­
lar anomaly most often missed by the pathologist. It is not
difficult to demonstrate (Fig. 11.6), but needs to be kept in
mind. 1SJ Idiopathic coronary arterial calcification has also
been found in symptomatic infants dying suddenly. 154
Sudden death in the infant age group during the course
of recognized, non-sudden illness, may, as in the older
child, result from pulmonary thromboembolism. 155 It is a
diagnosis rarely considered in early life and likely to be
missed without sampling of the lungs for histological
examination.
FOUND DEAD IN COT
Despite a high probability of death remaining unexplained
when an infant's death is discovered on household waken­
ing, nevertheless a full investigation is warranted. Some
deaths will be explained; both congenital heart disease and
infections are potential causes of death in this situation.
Whilst seizure disorders are another recognized cause of
death in this group, diagnostic or even supportive patho­
logical findings are not common.

Table 11.3 Causes of sudden neonatal death
Congenital heart di sease - particularly ductus-dependent
lesions
Infection - both bacterial and viral
Birth-related causes - both hypoxic and traumatic
Acute intestinal obstruction
Genetic metabolic disease
Figure 11 .7 Herpes type 1 infection . The mother had oral herpes
at delivery. The infant displayed non-specific symptoms followed
by sudd en deterioration at 2 weeks of age. Slice of liver shows
multiple cream, necrotic foci with a hyperaemic border.
SUDDEN DEATH IN NEONATES
Sudden death in the neonatal period (birth to 28 days of age)
is less common than SUDI but is usually explained (Table
11.3). It is particularly important that the full range of inves­
tigations is completed. Although structural cardiac disorders
can be easily demonstrated and there is usually some, albeit
rarely specifIc, naked-eye abnormality in cardiomyop athy,
the other common causes of sudden neonatal death may
show little macroscopic abnormality. Group B streptococcal
infection, both septicaemia and meningitis, is an important
cause of death in this age gro up. Acute viral infections are
important too. Herpesvirus infection has characteristic gross
(Fig. 11.7) and microscop ic fIndings even when it has eluded
pre-mortem diagnosis. Echoviral infection , particularly
serotype 11, results in rapidly progressive and often fa tal ill­
ness. Pathological changes are less specifIc, but haemorrhagic
necrosis of liver and adrenals is a common finding. 156 ,1 57 The
infection is more likely to be fatal when acquired from the
mother. 158 Other virus serotypes are sometimes involved.
Fatal group B Coxsackievirus infection is also usually
acquired from the mother. 159
Because the effects of perinatal hypoxia may be appreci­
ated only at microsco pic level, it is important that a formal
neuropathological examin ation is made of the fixed brain.
Pathological findings in SUDI I 209
Table 11.4 Typical post-mortem findings in SIDS (after ref. 711)
External
Well nourished , normally developed baby
Frothy blood-tinged fluid around the nose (500f0)
Hypostatic staining often anterior (indicates face-down position)
Cyanosis of lips and nailbeds
Internal
'Large' th ym us with petechiae (800f0)
Petechiae beneath viscera l pleura
Epi ca rdial petechiae
Full expansion of the lung s
Liquid heart blood (800f0)
Prominent mesenteric lymph nodes and Peyer's patches
Empty bladder (>500/0)
Microscopic
Thickened laryngeal basement membrane
Pulmonary congestion and oedema
Mild inflammation of the upper respiratory tract
Focal fibrinoid necrosis of the voca l folds
Persistent haemopoiesis in the liver
Failure to do this will result in loss of important info rmation,
particularly in respect of timing of ins ult (and may protect,
rather than condemn , your obstetric colleagues). Although
the majority of in fants with ~ -oxi dation defects will present
during late infancy, a few will be ac utely, and perhaps cata­
strophically, symptomatic in the neonatal period. 16o,161 It is
important that a fatty liver is not written off as due to peri­
mortem hypoxic change. As in later infa ncy, infection may
be a predisposing factor - we have encountered such a death
complicating sta phylococcal in fection. It is important to
consider the possibility of GMDs if only to ensure that
appropriate sa mpl es are reserved. 148
Which Post-mortem Findings are Compatible
With the Conclusion of SIDS?
MACROSCOPIC
Although a conclusion of SlDS is reached after excluding
explained causes of SUDl, nevertheless there are some
fIndings which, although not universally present nor spe­
cifIc, might almost be considered reassuring (Table 11.4).
Several stud ies have identified an increase in dysmorphic
features or minor malformations in SlDS.162-164 Although
many SlDS babies appear well nourished at death, poor
weight ga in has been identifIed in unexplained infant
deaths. 165 A fall in centiles from birth weight to death weight
might thus be anticipated, although differences between pre­
and post-mortem weights should be cautiously interp reted.
However, a weight less than the 10th centile at death,
particularly if accompanied by signs of dehydration or of
210 I Sudden unexpected death in infancy
Figure 11.8 Sudden infant death; large numbers of petechial
haemorrhages are present in a normal-sized thymus, while the
lungs are bulky and overlie the heart.
suboptimal care, for example extensive napkin delmatitis,
demands careful consideration of the overall care of the
infant.
Frothy secretions may commonly be visible at the nares.
They are pale in colour but may be blood tinged. Fixed
dependent lividity over the face and front of the trunk
indicates death in the prone position, but if the body is
moved sholily after death then this very useful pOinter may
disappear.
The appearance of the thymus deserves paliicular atten­
tion. It is usually of normal size, previously often con­
sidered large. Reduction in thymic weight correlates with
stress, often in the form of infection and indicates that the
infant had been unwell for more than 24 hours.166 The cer­
vical extensions are usually pale in comparison with the
thoracic part, which appears relatively congested.
Thymic petechiae, particularly when found in large num­
bers in the thoracic portion (Fig. 11.8), are one of the most
consistent findings in SIDS. They are often most numerous
on the posterior aspect. Beckwith 167 found them in more than
80 per cent of his cases and considered them the result of
respiratOlY obstruction with negative intrathoracic pressure.
They were present in large numbers in 49 per cent of SIDS
in another study, with a further 28 per cent having a few
petechiae. IG8
In SIDS, petechial haemorrhages are frequently promi­
nent along the course of coronary arteries, seen in 48 per
cent by Kleeman et al 1995. 168 Petechiae may be present
subpleurally. They are typically small and multiple. They
were present in 61 per cent of Kleeman's SIDS cases. Isaksen
and Helweg-Larsen 169 found subpleural pulmonary
petechiae more frequently in infants in whom resuscitation
had been attempted. In SIDS, when the baby has not been
subjected to attempts to resuscitate with oxygen, the lungs
are typically fully expanded and fill the chest. They do not
collapse on slicing.
SUDI is more common among pre-term babies and those
who have been admitted to special care baby units. Evidence
of residual bronchopulmonalY dysplasia in the form of scar­
ring or fissUling of the lung is not unusual. These cases have
related pulmonalY histological abnormalities. 170 It could be
argued that these deaths are, in fact, explained by circula­
tory decompensation triggered by minor infection.
Frequently a small amount of straw-coloured fluid is seen
within the pericardial sac. There is usually fluid blood within
the heart; it is the result of post-mortem fibrinolysis said to
be associated with the suddenness of the death. The foramen
ovale is often valvular. The significance of an atrial septal
defect has been debated, although Dancea et al ll3 record two
cases with signs of right heart overload among their sudden
cardiac deaths.
Cervical and thoracic lymphadenopathy are unusual
findings in SIDS, but mesenteric lymph nodes are frequently
enlarged, ref1ecting the level of environmental antigenic
stimulation in this age group. Peyers' patches are often
prominent and may be hyperaemic.
The stomach is often distended, particularly when resus­
citation has been attempted. Milk curd in the stomach is
a common finding and does not necessarily mean that
death occurred shortly after feeding. However, an empty
stomach may indicate survival for some time after the last
feed, but vomiting and gastric emptying during resuscita­
tion must be ruled out. l7l When the small intestine is also
empty, the possibility of gastrointestinal infection and sub­
optimal care should be considered. Any gastric contents
should be retained for analysis.
The urinary bladder is frequently empty in SIDS.172 It is
such a frequent finding in unexplained deaths that the
presence of a full bladder raises my expectation of signifi­
cant pathology.
HISTOLOGICAL
Thorough histological evaluation is important in SUD!. It
was the investigation most likely to yield significant infor­
mation in a UK multicentre study.4
Pul monary oedema and congestion are frequent findings
in SIDS. Minor focal alveolar haemorrhage may accompany
the oedema. This is the origin of the frothy fluid seen at the
nares. There is often an increase in rOLlnd cell infiltration of
alveolar walls. Peribronchial lymphoid aggregates are often
present, but these are seen in most cases of infant death and
are a marker of prior respiratory infection. 17)
The significance of minor inf1ammation is a recurrent
problem in SUDI. 32 Focal acute inflammation is frequently
~ ­
 Pathological findings in SUDI I 211

found (60 per cent of cases) in both the upper and lower
respiratory tract in SIDS. The former is not likely to be a
primary cause of death J74 unless it results in major narrow­
ing of the airway. The latter provokes the question 'How
much pathology is required to equate with a cause of
death?'. This is a difficult question and one that has been
sidestepped on occasion. 16 Consolidation throughout sec­
tions from more than two lung lobes is a yardstick I have
used. Gregersen et al 175 consider the extent of pulmonary
disease a critical factor in deciding whether or not death is
explained. Culture of a single bacterial pathogen from the
organ in question and blood cultures is likely to influence
interpretation of histological findings.
Localized inflammation might acquire significance
because of its localization, for example in the cardiac con­
duction system or brainstem. 32 These authors are uncertain of Figure 11.9 SUDI, frozen section of liver. Microvesicular fat is
the significance of focal inflammation in both the meninges present in zone 1; this amount of lipid is a frequent finding in
and myocardium, although an earlier publication appears to unexplained infant deaths. Oil Red O.
accept that significant myocarditis may be focal. 140
Basement membrane thickening in the vocal cord was a
frequent finding in SIDS in one study.176 However, in a
larger study Krous et al J77 found no difference in laryngeal
basement membrane thickening between SIDS cases and a
control group. More serious damage in the form of fibri­
noid necrosis of the vocal fold is seen in some SIDS
babies. 17s
In the kidney, glomerulosclerosis may be a striking
finding. It is not significant and is seen frequently in
explained death in this age group. J79
Persistence of haemopoiesis in the liver, which normally
disappears soon after birth, is seen in many SIDS cases and
has been promoted as evidence of repeated or chronic
hypoxia. ISO, lSI
Cytomegalovirus (CIvrv) inclusions have been recorded
in SIDS babies for many years. Their incidence in salivary
Figure 11,10 Reperfusion injury in an infant who was on life
gland or mucous glands in the upper respiratory tract is no
support for hours after collapse at home. There is a small, old
more frequent in SIDS than in explained SUDI. IS2 However,
calcified lesion in the periventricular white matter on the right
their association with microglial nodules in the brain stem
indicative of a pre- or perinatal ischaemic lesion.
has provoked the suggestion that, in this circumstance, the
effect of crvrv infection is likely to be significant. ls3
Evidence of minor old periventricular leucomalacia is
sometimes seen in SIDS babies (Fig. 11.9), present in one­
fifth of cases in a Canadian study.184 This is perhaps not A number of changes found at a histological level in SIDS
surprising given the association with SIDS of low birth have come to be regarded as markers of 'stress', i.e. non­
weight, pre-term delivery and neonatal morbidity. A more system-specific markers of illness. These changes are seen in
frequent finding is disruption of the ependymal lining of many SIDS deaths, in th e thymus, liver, adrenals and the
the lateral ventricles, which, when recent, results in atten­ osteochondral interface (most conveniently accessed at the
uation of the columnar epithelial lining and, when longer anterior ends of the ribs).
standing, results in ependymal rosettes and attendant glio­ A starry-sky appearance in the thymus, perhaps accom­
sis. Brainstem gliosis is a relatively common finding. ISS, 186 panied by minor cortical lymphocyte depletion, is compatible
A variety of abnormalities of brainstem serotonergic sys­ with a SIDS designation, but major cortical lymphocyte
tems have been described in SIDS following detailed mor­ depletion or reversal of cellularity betvveen cortex and
phometric and immunohistochemical studies, including medulla indicates a longer (and probably more severe)
hypoplasia of the arcuate nucleus and subtle gliosis of sev­ episode of stress, 166
eral brainstem nuclei. These are reviewed by Kinney and Minor lipid accumulation in the liver (Fig. 11.10) is a
Filiano ls7 and Kinney.18s marker of minor stress. It is microvesicular and usually more
212 I Sudden unexpected death in infancy

marked around central veins. Should fatty change be exten­

sive, the possibility of an inherited metabolic di sease as an
explanation for death should be pursued. When there is pan­
lobular fatty change, particularly when macrovesicular
steatosis is also seen (see Fig. 12.21 , p. 247), then the possibil ­
ity of a GMD is high. Concomitant steatosis should be sought
in the myocardium and proximal renal tubules and specific
investigations for GMDs carried out. It is important to appre­
ciate the importance of severity when assessing fatty change.
r cannot be the only pathologist to have conducted an inves­
tigation into the death of a second infant from the same fam­
ily who has been told by the pathologist who investigated the
earlier one that he had noted gross hepatic steatosis but
thoug ht that a fatty liver was a common finding in SIDS.
Lipid depletion of the adrenal cOliex is another non­
specific marker of stress. 189 It is perhaps worth remembeling
that lipid accu mulation in the adrenal cortex is, in fetal life, a
Figure 11.11 Horizonta l slice through the interve ntricul ar
marker of chronic anaemia, probably mediated via hypoxia. 190
septum of an infant; there is extensive haemorrhagic infarction
Loss of regular transition at the osteochondral junction
caused by ventilation for about 12 hours post collapse.
is a fre quent finding in SIDS, but more serious disruption is
a marker of chronic growth disturbance. 191

not preferentially along the antimese nteric border, as is often

Changes Induced by Attempted Resuscitation seen in the early lesions of NEe.
Cardiopulmonary resuscitation (CPR) is an infrequent
cause of rib fractures in infants because their ribs are very
Minor excoriation around the mouth and on the tongue and
pliable. Only 1 infant out of 2 11 who unde rwe nt resuscita ­
pharynx may occasionally be seen. Resuscitation makes the
tion aro und the time of deat h susta ined rib fractures in a
interpretation of recent bruising aro und the nose, mouth
North American study.1 9J Bilateral fractures at the sterno ­
a nd lower jawline problematic. Kaplan and Fossum 192
chondral junction were found. We have seen similar but
relate minor injuries in th is a rea to specific resuscitation
unilateral fractures in a single infant - an explained natu­
manoeuvres. Gastric dil atation due to assisted ventilation is
ral death following CPR (Fi g. 11.12) - and undisplaced lat­
a common finding when resuscitation has been attempted.
eral rib fractures in a further case (Fig. 11.13). Feldman a nd
Ventilation with oxygen, which is subsequently resorbed,
Brewer 194 and Spevak et al 195 found no rib fractures in
can result in pulmonary coJi apse even when carried out for
infants w ho had undergone CPR in their cases. Contrary to
a brief period. It is important to be aware of this phenom­
popular belief, professionals are more likely to cause injury
enon as it obliterates a typ ica l SIDS findin g, one w hich
during resuscitation than are amateurs, irrespective of
offers potential distinction between SIDS and hypox ia .
physical mass or the use of in appropriate resuscitation
Prolonged (greater than 12 hours) ventilation, usually
techniques. On those infrequent occasions when resuscita­
instituted when cardiac electrical activity is elicited during
tion does resul t in rib fractures, bleeding is min imal. Vis­
resuscitation in hospita l, can produce a number of changes,
ceral injuries are similarly uncommon resuscitation
particularly reperfusion injury. The general level of tissue
injuries. 193 ,1 96
preservation is very poor, unlike the usual situ ation in
Focal disrupti o n of cortical bone of the anterio r tib ia
infant deaths, when histological examination is usually
during attempted intraosseous transfusion can produce a
worthw hile. Not surprisingly, the brain is swollen and
confusing radiological ap pearan ce and give rise to co ncern
extremely soft. Reperfusion cerebral haemorrhagic infarc­
about the possibility of fracture.
tion (Fig. 11.9) ca n occur in this situation. It is impOltant
not to interpret this as an indicat ion of cerebral hypoxia
occurring prior to the term ina l event. Similar changes may Pathological Findings which Raise Concern
be present in the cardiac intraventricular septum (Fig. About the Possibility of Non-Accidental
1l.11). Pulmonary alveo lar and interstitial haemorrhage is Injury (NAI) (Table 11.5)
also common.
Focal haemorrhagic infarction may be seen throughout
EXTERNAL FINDINGS
the intestine but may be confined to caecum and colon.
Unlike necrotizing enterocolitis (NEC), seen particularly in Petechial haemorrhages in the face or neck are unusual
the pre-term infant, all of the lesions are of similar age and findin gs in SIDS and in healthy babies. In the latter gro up,
have a diffuse margi n. They are often circumferential and most occurred below the nipple line. Only 2.5 per cent had
 Pathological findings in SUDI I 213

Figure 11.12 (a) A rib cage with recent fractures at the anterior ends of three adjacent ribs; there is little associated haemorrhage.
(b) Radiograph of affected ribs. The fractures are accompanied by irregularity of outline at their anterior ends. (c) Internal aspect of the
ribs. Fractures are visible close to the osteochondral junction. (d) Histological examination reveals a fracture with haemorrhage but no
reactive changes. Sudden collapse, attempted resuscitation, anomalous pulmonary venous drainage found at necropsy.

petechiae in the head and neck, and only 2.6 per cent had
three or more petechiae in total. 197 Their presence raises the
possibility of deliberate asphyxia (Fig. 11.14). When pre­
sent over the face, neck and upper chest, petechiae are sug­
gestive of chest compression.lg8-200 The infant should be
carefully examined, searching for bruising around the face
and circumoral pallor, tears at the frenulum (recent or
healed) and abrasions or bruises on the inside of the mouth
and lips corresponding to teeth. Collins 98 found little
evidence of contusions, petechiae or intraoral trauma in
overlaying and considered that their presence should
incline towards inflicted injurY.
Conjunctival petechiae should be very carefully sought.
Kleeman et al 168 found small numbers of conjunctival
petechiae in 2.4 per cent of their SIDS cases. This has not
been confirmed by others. Larger numbers of petechiae
were seen in 21.9 per cent of babies dying as a result of
trauma, including in five out of six babies dying from
214 I Sudden unexpected death in infancy
Figure 11.13 Post-mortem rib fractures mid shaft in adjacent
ribs. There is only very minor subperiosteal haemorrhage at the
fracture lines.
Table 11.5 Pathological findings causing concern in SUDI
Petechial haemorrhages face/neck
Pallor around nose/mouth
Torn frenulum
Bru ises - even one is serious
Heavily blood-stained secretions mouth/nose
Blood in pharynx (seen by a doctor)
Rib fractures - recent or old
Any other injury
Blotchy haemorrhages on lung
Alveolar haemorrhage> 10% alveoli
Siderophages in lung
strangulation. A combination of conjunctival petechiae
and acute pulmonary emphysema was found in all of seven
babies with asphyxia or strangulation, but not in cases of
SIDS, other natural deaths or severe head injury in a study
by Betz et al. 201 They considered the combination of two
findings useful in distinguishing asphyxial death from
SIDS.
Development of petechial haemorrhages requires a com­
bination of hypoxic damage to small blood vessels, a func­
tioning circulation and an increase in venous pressure, not
necessarily in that order. 202 It is unlikely, then, that resus­
citation will produce petechiae unless circulation and
Figure 11.14 Petechial haemorrhages on the neck and front of
chest are not a usual finding in unexplained SUDI.
blood pressure have been successfully restored during the
process. 203
Blood or heavily blood-stained fluid issuing from
mouth or nose or visualized in the pharynx is also a cause
for concern. Nasal haemorrhage was reported in 15 per
cent of SIDS cases in a study by Becroft et al,61 based on
information extracted from a structured questionnaire.
Babies with nasal haemorrhage shared epidemiological
characteristics with babies who had pulmonary haemor­
rhage but not with intrathoracic petechiae. It was found (29
per cent) in babies subjected to imposed airways obstruc­
tion but not following explained ALTE.55 In a study of
58059 infants under 2 years of age referred to the accident
and emergency department from a children's hospital 204
only 16 cases were found, clearly a rare event. Eight of
these infants had visible trauma, four had thrombocytope­
nia and tYro cases were associated with ALTE; retrospective
review drew concerns about child protection issues in 44
per cent of cases. Oronasal haemorrhage should be distin­
guished from the blood-tinged frothy fluid at the nares
commonly seen in SIDS. When present, bloody secretions
raise the possibility of asphyxia. Krous et al 205 suggest an
origin from oronasal mucous membranes.
Cutaneous bruises should always cause concern in SUDI.
Even a single bruise in a non-mobile infant demands a clear
explanation. Such 'minor' injuries must always be taken
seriously. 171 Older infants who have some degree of mobility
may exhibit an occasional bruise. It is important with any
infant injury to consider carefully both the physical abilities
of the individual and the explanation proffered by carers.
 Pathological findings In SUDI I 215

Accidental bruises are usually over bony prominences. 206 Pulmonary congestion and oedema, although seen in
Bruises of different ages are most concerning. histological sections, is not a prominent naked-eye finding
in SIDS.

FRACTURES
PULMONARY HAEMORRHAGE/SIDEROPHAGES
Fractures are not an expected finding in SIDS and must be
Histological findings causing concern in SUD! are largely
clearly explained. Fractures with callus always predate ter­
confined to the lungs. The abnormalities that have proved
minal events, probably by a minimum of 2 weeks, and are
most contentious are the presence of widespread alveolar
incompatible with the use of SIDS on the death certificate.
haemorrhage and the presence of haemosiderin within the
Rib fractures are sometimes seen in extremely pre-term
lung (Fig. 11.15).214 Minor (focal) pulmonary haemorrhage
babies. They are accompanied by severe growth distur­
is common in the lungs of deceased infants and may be
bance at the osteochondral junction (rickets). In most units,
exacerbated by a long post-mortem interval. 215 Coffin et
they are encountered less frequently than 20 years
al 216 found that babies with pulmonary haemorrhage had a
ago. 207,2oa However, rickets was described more recently
high frequency of obstetric and neonatal problems; there
among 39 per cent of very low birth weight babies in one
was a significant association with hyaline membrane dis­
unit, with fractures, most commonly of ribs, occurring in
ease and haemorrhage elsewhere.
one-quarter of those affected.209 Fractures are unlikely to
There are numerous causes of alveolar haemorrhage in
occur after the baby is weI] enough to leave hospital.
neonates, including hypoxic stress, sepsis and haemostatic
Should birth injury be offered as an explanation for a frac­
problems,216-218 not all of which are easily excluded by
ture, the age of the infant and details of mode of delivery
necropsy. In older infants, sepsis, heart failure (often sec­
are important factors. It is worth remembering that the
ondary to CHD) and chronic pulmonary infection, as well as
most common birth-related fracture is clavicular, and a
idiopathic pulmonary haemosiderosis (IPH), can give rise to
history of difficul ty in delivery of the shoulders should be
haemorrhage. It is more prominent in babies who have been
anticipated.
resuscitated. 2i s
Femoral fractures (usually mid-shaft, sometimes bilat­
Yukawa et al 2J9 looked at the presence and extent of
eral) can complicate difficult breech delivery but do not
intra-alveolar haemorrhage in a consecutive series of SUD!
occur during vertex presentation. The most common skull
investigated by a forensic pathology department. They
fracture is parietal and linear and runs from the mid-part of
concluded that alveolar haemorrhage involving more than
sagittal margin of the parietal bone down towards the ear
five per cent of alveolar area in histological sections was
and usually follows instrumental delivery.
probably a marker of airways obstruction. It was present in
Rib fractures are exceedingly rare birth injuries and sug­
73 per cent of babies thought to have been subjected to
gestion of this mode of causation should be viewed most
involuntary overlaying (accidental asphyxia) and in 45 per
circumspectly. Only nine cases have been reported world­
cent of cases in which there were suspicions of deliberate
wide. Taken together, a pattern emerges of high birth weight,
airways obstruction.
shoulder dystocia, delayed second stage of labour, instru­
Every consideration must be given to possible natural
mental delivery and symptoms in the first day or two of life.
causes of pulmonary haemorrhage before asclibing the find­
Symptoms include crepitus of the chest wall, tachypnoea
ings to imposed airways obstruction. Concerns expressed
and grunting respiration. That the fractures were unilateral is
about the more general application ofYukawa et al's219 find­
a useful distinguishing point.210-213
ings are the paucity of typical SIDS cases studied and no sig­
nificant difference between number of allegedly asphyxiated
babies and SIDS cases having greater than five per cent
VISCERAL PATHOLOGY
haemorrhage.214 The finding that intra-alveolar haemor­
Injuries and foreign bodies in the mouth, pharynx or stom­ rhage was most common in those babies in whom over-laying
ach, particularly in premobile infants, should arouse suspi­ was a possible factor in their death invokes the possibility
cion of NAI. 10 Intrathoracic petechiae are seen less frequently that repeated non-fatal hypoxia played an important role in
in traumatic deaths 168 and among bed-shaling infants 10) than the haemorrhage, drawing a parallel with the development
in SIDS cases in all of the three expected sites. In contrast, of petechial haemorrhages. The role of chest compression is
small numbers of larger haemorrhagic subpleural haemor­ unclear. 21 9 This group of babies was younger than the rest,214
rhages are seen in some asphyxial deaths. When the asphyx­ a factor associated with more frequent, but scanty, pul­
ial insult is very acute, for example a foreign body in the monary siderophages in another study.220 As an isolated
larynx, very few petechiae may be present as death by finding, intra-alveolar haemorrhage is not a clear marker for
parasympathetic stimulation may have preceded the onset of incontrovertible imposed upper airways obstruction but
mechanical asphyxia. In rabbits, pleural petechiae were read­ should initiate a critical review of the whole case. 214
ily produced by three episodes of sublethal airways occlusion Stewart and Fawcett,221 in an uncontrolled study, found
but not by a single apnoeic episode.2J3 pulmonary interstitial haemosiderin in one-half of 24
216 I Sudden unexpected death in infancy
Figure 11.15 (a) Siderophages are present within the alveoli.
(b) Low-power photomicrograph showing the extent of
siderophage formation follo wi ng alveolar haemorrage.
consecutive SlDS deaths. Babies with siderophages were
usually younger than those without. The autho rs postulated
that haemosiderin might be a marker for 'near miss' events
(ALTEs). Byard et al222 found more pulmona ry interstitial
haemosiderin in SID S infants with prior ALTEs (33.3 per
cent) than in non-traumatized control infants (five per cent).
Haemosiderin was found in 18 per cent of SIDS infants
without ALTEs, a statistical ly non-significant difference.
The authors concluded it was an unreliable marker of a pre­
vious ALTE. Becroft et al223 consider pulmona ry interstitia l
haemosiderin a consequence of normal labour.
Becroft and Lockett 224 found large numbers of intra­
alveolar siderophages in babi es whose deaths were caused by
airways obstruction. They proposed that their presence
should give rise to suspicion of imposed airways obstruction
and that the lungs of all SUD I victims should be stained for
iron. Pulmonary siderophages have been demonstrated in
repeated airways obstruction. 225 Dorandeu et a1 226 found pul­
monary siderophages in 11 of 15 infants with non-accidental
injury; all had intracrania l haemorrhage and 12 had other
injuries as well.
Alveolar siderophages are infrequent in SIDS
cases. 21B,220.227-229 Krous et al230 found wide variati on in
the number of intra-alveolar siderophages in a large group
of unexplained infant deaths and were not able to distin­
guish such infants from babies dying from suffocation.
Pulmonary siderophages have been described in
leukaem ia.217 They have also been found post-operatively, in
serio us sepsis and followin g ventilation, situations which
might have caused hypoxic pulmonary damage. 2IB They
have been see n in infants from homes with preceding water
damage in substandard housing in Cleveland, Ohio,231 sug­
gesting a role for air-borne contaminants as pulmonary
irritants. IPH has been found to occur in infants as well as in
older children 232 and appears always to be symptomatic.
It has been suggested that IPH can present as SUDI,233 but it
is not clea r how confidently imposed airways obstruction
was excluded in either study.
Haemoglobin breaks down quickly in extravasa ted red
cells, which are quickly pha gocytosed. Finely dispersed
haemosiderin can be found after 24- 36 hours. Haemosiderin
is ab undant at 5 days, and phagocytosed erythrocytes are
recognizable. 234 Red cells have disappeared after abo ut 7
days. Inform ation about the persistence of siderophages
from the lung is scanty. They are rapidly cleared from large
airways (within 2 weeks) and from more peripheral parts of
the airway within 4 weeks 235 in infants. Esterley and Oppen­
heimer 236 found that siderophages had disappeared from
the alveoli after massive perinatal pulmonary haemorrhage
within 2 weeks. It is reasonable to suppose that, whilst most
alveolar siderophages are removed from the lung via the air­
ways, some macrophages will migrate to septal and pleural
connective tissue and might remain there for several weeks
after the initiating haemorrhage.
ASPIRATION OF FOREIGN MATERIAL
It seems likely that many babies respond to an asphyxial
insult by vomiting. Evid ence of major aspiration is a cause
for concern. It shou ld be distinguished from the localized
 Pathological findings in SUDI I 21 7

Figure 11.17 Lipid-containing macrophages are present in

alveoli at the periphery of a lung lobule.

Findings of Questionable Significance

Figure 11.16 Massive inhalati on in a baby found to have a
diffuse neuronal migration disorder. There is brown discoloration
EXTERNAL
of the lungs because of acid lysis of red blood cells.
Cutaneous changes are sometimes overinterpreted in infant
deaths. Dribbling can result in erythema and abrasions
around the mouth or on the upp er chest and, when it leads
to excoriation, might be interpreted as an injury. Sometimes
and usually peripheral foreign body-type granulomata,
excoriation takes place after death when the infant's cheek
which are the result of repeated, minor aspiration, often
is bathed in acidic fluid from regurgitated gastric contents;
seen in young babies with oesophageal or lary ngea l prob­
a linear margin points to the artefact.
lems. It is most unusual in the absence of a predisposing
Healing napkin dermatitis (nappy rash) can be pigmented
problem. This may be gastroenteritis, pyloric stenosis,
and give rise to suspicion of bruising. A Mongolian blue
diverse causes of intestinal obstruction, cerebra l palsy or
spot, seen normally, but not always, in infants with pig­
other neurological problem. In a recently investigated case.
mented skin, can simulate bruising; its localization around
brown discoloration in lung fissures (Fig. 11.16) due to acid
the natal cleft shou ld suggest the true nature of the lesion
lysis, extensive lipid in airways and terminal air spaces
(see Fig. 8.7, p. 152). Vascular malformations can sometimes
within the lung (Fig. 11.17) and a diffuse neuronal migra­
mimic injuries. If doubt persists after careful scrutiny, the
tion disorder were observed. The only evidence of cerebral
diagnosis can be resolved by histological examination.
abnormality at necropsy was localized polymicrogyria at
Anal dilatation is common post mortem and should not,
one occipital pole.
as an isolated finding, promote suspicion of abuse. McCann
et al 238 have made a careful study of the post-mortem
appearance of the anus in infants and children. They counsel
HEPATIC SIDEROPHAGES
against overinterpretation of an exposed pectinate line as
Dorandeu et al 226 found increased haemosiderin in the tears or fissures. They stress the importance of tears and peri­
liver, predominantly in periportal Kupffer cells, in abused anal bruising to a conclusion that sodomy has been commit­
infants compa red with control infants who were victims of ted; swabs for DNA studies should be taken nonetheless.
SIDS or accidental or explained death. They asc ribed this to
a chronic increase in red cell breakdown . Increased iron in
INTERNAL
the liver is not unusual in infants with infection (usually
within hepatocytes) and those with inherited erythrocyte The presence of gastric contents in the oesophagus, phar­
a b no rmali ties. 237 ynx, mouth and nose is a common necropsy finding in the
218 I Sudden unexpected death in infancy
Figure 11.18 Infant aged 4 months. There is brown staining of
the dura of tentorium and posterior falx owing to organization of
focal subdural haemorrhage sustained at birth.
infant period. Sphincters relax after death, and moving the
baby in the course of the removal to the mortuary will read­
ily disperse stomach con tents into the pharynx and beyond.
When resuscitation is attempted, the presence of milk curd
in the pharynx is frequently recorded. Attempted resuscita­
t ion will disperse this further, and milk curd in the upper
airways and focally in large intrapulmonalY bronchi should
not be interpreted as being a result of inhalation unless it is
massive 31 (Fig. 11.16) and there is histological evidence of
reaction in the form of extensive (frequently haemorrhagic)
oedema and inflammatory cells within the lungs. Signifi­
cant gastric aspiration in infants will usually be accompan­
ied by evidence of a predisposing problem.
Healing birth injury may give rise to suspicion of NAl.
The most commonly encountered injUly is minor subdural
haemorrhage, whose appeara nce should be commensurate
with the age of the infant, i.e. it wi ll be red-brown and rec­
ognizable as a haemato ma, prob ab ly 1-2 mm thick, for 2,
possibly up to 4, weeks post partum. Case et al23 9 estimate
that these small haemorrhages occur in 20-30 per cent of
asymptomatic neonates. Later th a n that, brown staining of
the dura is apparent (Fig. 11.18). This may persist for sev­
eral months. Histological examination of the dura will
reveal abundant coarsely granular haemosiderin on the
surface and within the membrane itself (Fig. ) ).19).
, ', -I t­
• I' ~,, _. . . '_
.. .. - .
Figure 11.19 Histological examination of an area of brown
staining of the dura. Abundant iron is demonstrated by Perls'
Prussian blue reaction. Iron persists in the dura for many months.
An organizing cephalhaematoma is usually apparent as
an irregu lar elevation on the parietal bone, and it may be
bilateral. Most cephalhaematoma ta occur during uncom­
plicated vaginal delivery. A history of a difficult birth is
unusual. Cephalhaematomata may be noticed only once
bone is laid do wn beneath the elevated periosteum. Its
appearance should correlate with post-natal age. Evidence
of this new bone formation may persist for many years.
The most common fracture susta ined as a birth injury is
a mid-shaft fracture of the clavicle. There may be a history
of shoulder dystocia, but most will not have this history.
A lump over the clavicle may have been noted by the mother
or health professionals prior to death. The radiographic an d
histological appearance should be compatible with post­
natal age. Other long-bone fractures are rare birth injuries
and occur in well-recognized circumstances. IS3
MICROSCOPIC
Basement membrane thickening in the larynx is an incon­
sequential finding (see p. 211) . Explosive desquamation of
bronchial and bronchoalveolar epithelium has been
interpreted as evidence of pulmonary infection 240 but is a
post-mortem artefact and may be found even when refriger­
ation of the body has been prompt. Peribronchial lymphoid
aggregates are a normal finding in infants.
Pancreatic isl ets often appear prominent in SlDS. They
are usually sited towards the centre of the lobule and become
more generally dispersed with subsequent acinar develop­
ment. This should not be interpreted as islet hyperplasia.
Examination of the pancreas from an expl ained infant death
is a useful compalison.
DEATH CERTIFICATION
As SIDS is a conclusion reached by exclusion of recognized
causes of death and its definition requires that a thorough
post-mortem be performed, it is not appropriate to use the
term on a dea th certificate unless certification is delayed
until investigations are completed. This period is likely to

be longer than the statutory period for death registration
(within 8 days in the UK).
My practice accords with that of Gilbert-Barness and
Barness B in that, when history, circumstances and naked­
eye findings are typical , I use the terms SUD! or 'sudden
unexpected infant death', together with an indication to the
registrar that additional information may become available.
In recent years, pathologists have become much more
likely to use the terms 'unascertained' or 'undetermined'
on death certificates, although in the UK practice varies
widely.13 This has provoked protest from parents' support
groups. However, it is the pa thologist, not the counsellor,
whose opinion or competence is called into question when
an atypical infant death is reinvestigated after the abuse or
death of another family member. 9 Mitchell et al lio have
highlighted the swing away from SIDS to 'undetermined'. It
is important to consider the potential consequences of the
use of SIDS, in effect stating that death is due to natural
causes, on subsequent infants in the family.214 The use of
'unascertained' in these circumstances ensures that the case
is kept open and might alert medical and support services
to the need for vigilance and an open mind when assessing
problems in subsequent infants. The importance of multi­
disciplinary review of unexpected infant deaths has been
stressed earlier in this chapter.
REFERENCES
Carpenter RG, Shad dick Cw. Role of infection, suffocation
and bottle feeding in cot death. Br ] Pre/} Med 1965; 19: I.
2 Beckwith JB. Introdu ction: discussion of terminology. In
Bergman AB, Beckwith JB, Ray CG (eds) Sudden Infant
Death Syndrome. Washington DC: University of Washington
Press, 1970, p. 18.
3 Bass M, Kravath RE. Glass L. Death-scene investiga tion in
sudden infant death. N Engl] Med 1986 ; 315:100-5.
4 Berry J, Allibone E, McKeever P et al. The parhology study:
the contribution of ancillary pathology tests to the
investigation of unexpected infant death. In Fleming P,
Bacon C, Blair P, Berry PJ (eds) Sudden Une.rpected Deaths
in Infancy th e CESDI SUD! Studies 1993-1996. London:
The StationeJY Office, 2000, pp. 97-112.
5 Willinger M, James LS, Catz C. Defining the sudden infant
death syndrome (SIDS): deliberations of an expert panel
convened by the National Institute of Child Health and
Human Development. Pediatr Pathol 1991; 11 :677-84.
6 Fleming P, Blair P, Ward-Platt M et al. The case-control
study: results and discussion. In Fleming P, Bla ir P,
Bacon C, Berry J (eds) Sudden Une.rpecred Deaths in
lnfa ncy: The CESD! SUD! Studies. London: The Station ery
Office, 2000, pp. 13-96.
7 Bajanowski T, Vennemann M, Rauch E et al. Unnatural
causes of sud den unexpected deaths initially thought to be
sudden infant death syndrome. Int] Legal Med 2005;
119:21J-16.
8 Gilb ert-B arness EF, Barness LA. Sudden infant death
syndrome. Is it a cause of death? Arch Pathol Lab Med
1993; 117: 1246-8.
9 Green MA. Time to put 'cot dearh' to bed 7 BM] 1999 ;
319:697-8.
References I 219
10 Meadow R. UnnalUral sudden infant death. Arch Dis Child
1999; 80:7 -14.
11 Limerick S. Not time to put cot death to bed. BM] 1999;
319 :698-700.
12 Bacon CJ. Cot death after CESDI. Arch Dis Child 1997;
76:171-3.
13 Limerick SR, Bacon CJ. Terminology used by pathologists in
reporting on sudden infant deaths. ] Clin Pathol 2004;
57:309-11.
14 Flemming PJ, Blair PS, Sidebottom P, Hayler T. Investigating
sudden unexpected deaths in infancy and childhood and
caring for bereaved families: an integrated multiagency
approach. Br Med] 2004; 328:331-4.
15 Krous HF, Beckwith JB, Byard RW et al. Sudden infant death
syndrome and unclassifi ed sudden infant deaths:
a definitional an d diagnostic approach. Pediatrics 2004;
114:234-8.
16 Knowelden J, Keeling J, Ni choll JP et a1. A Multicentre Study
of Post-neona ta l Mortality. London: HMSO, 1984.
17 Gibson AAM. Current ep id emiology of SIDS. ] Ciin Pathol
1992; 45(Supp1.):7-10.
18 Gilbeli RE, Fleming PJ, Azaz Y, Rudd PI. Signs of illness
preceding sudden unexpected death in infants. BM] 1990;
300:1237-9.
19 Wigglesworth JS , Keeling JW, Rushton DI, Berry P1.
Technical method. Pathological investigations in causes of
sudden infant death.] Clin Patlzol 1987; 40:1481-3.
20 Royal College of Pathologists. Guidelines for Post-mortem
Reports. London: Royal Co llege of Pathologists, 1993, pp. 8- 10.
21 Valdes-Dapena M, McFeeley PA, Hoffman HJ et al.
Histopathology Atlas fo r the Sudden Infant Death Syndrome.
Washington DC: Armed Forces Instit Pathol, 1993.
22 Krous HF. An international standardized autopsy protocol
for sudden unexpected infant death. In Rognum TO (ed.)
Sudden Infant Death Syndrome. New Trends in the Nineties.
Oslo: University of Sca ndinavia Press, 1995, pp. 75-89.
23 Howatson AG. The autopsy for sudden unexpected death in
infancy. Current Diagnostic Pathology 2006; 12: 173-83.
24 Report of a Working Group: Sudden Unexpected Death in
Infancy. London: Royal College of Pathologists and Royal
College of Paediatrics and Child Health, 2004.
25 Landi K, Gutierrez C, Sampson B et al. Investigation of the
sudden death of infants: a multicenter anal ys is. Pediatr Del;
Patl101 2005; 8:630-8.
26 Sadler DW. The va lue of a thorough protocol in the
investigatio n of sudd en infant deaths.] Ciin Pathol 1998 ;
51 :689-94.
27 Mitchell E, Krous HF, Donald T, Byard RW. An analysis of
the us efuln ess of specific stages in the pathologic
investigation of sudden infant death. Am ] Forensic Med
Pathol 2000; 21 :395-400.
28 Langlois NE, Ellis PS, Little D, Hulewicz B. Toxicologic
analysis in cases of possible sudden infant death syndrome:
a worthwhile exerc ise? Am] Forensic Med Pathol 2002;
23:162-6.
29 Arnestad M, Vege A, Rognum TO. Evaluation of diagnostic
tools applied in the examination of sudd en un expected
deaths in infancy and early childhood. Forensic Sci Internal'
2002; 125:262-8.
30 Keeling JW, Knowles SAS. Sudden death in childhood and
adolescence.] Patho11989; 159:221-4.
3I Arneil GC, Gibson AA, McIntosh H et a1. National post­
perinatal infant mortality and cot death study, Scotland
1981-82 . Lancet 1985; 1:740-3.
32 Byard RW, Krous HF. Minor inf1ammatory lesio ns and
sudden infant death: cause, coincid ence, or ep iphenomena?
Pediatr Pathol Lab Med 1995; 15:649-54.
220 I Sudden unexpected death in infancy
33 Bacon C, Fleming P. Design of the SUDI study. In Fleming P,
Blair P, Bacon C, Berry J (eds) Sudden Unexpected Death In
Infancy. Th e CESD! SUD! Studies. London: The Stationery
Office, 2000, pp. 7-12.
34 Irgens LM. Risk factors for SIDS: Do they exist? In Rognum
TO (ed.) Sudden Infant Death Syndrome. Os lo: Scandinavian
Uni vers ity Press, 1995, pp. 99-105.
35 Blair PS, Fleming P J, Bensley D et al. Smoking and the
sudden infant death syndrome: results of 1993-5 case­
control study for confidential inquiry into still bil1hs and
deaths in inFancy. 8M] 1996; 313:195- 8.
36 Aim B, Milerad J, Wennergren Get al. A case-control study
of smoking and sudden inFant death syndrome in the
Scandinavian countries, 1992 to 199 5. Arch Dis Child
1998 ; 78 :329- 34.
37 Blair PA, Sidebotham P, Berry P J et a l. Major
epidemiolo gica l changes in sudden infant death syndrome: a
20-year population-based study. Lancet 2006; 367:
314-19.
38 Daltveit AK, 0yen N, Skjaerven R, Irgens LM. The epidemic
of SIDS in Norway 1967-93: chan ging effects of risk factors.
Arch Dis Child 1997; 77:23-7.
39 Smith Gc, Wood AM, Pell JP, Dobbi e R. Sudden infant death
syndrome and complications in other pregnancies. Lancet
2005; 366:2107-11.
40 Malloy MH. Sudden infant death syndrome among
extremely pre-term infants: United States 1997-1999.
] Perinatol2004; 24:101-7.
41 Leach CE, Blair PS, Flemming P J et al. Epidem iology of SIDS
and explained sudden infant deaths: CESDI SUDI research
group. Ped,:atrics 1999; 104:e43.
42 Golding J, Peters TJ. What else do SIDS risk prediction
scores predict? Early Hum Develop 1985; 12:247-60.
43 Smith GC, Wood AM, Pell JP et a!. Second-trimester
maternal serum alphaFeto protein levels and the subsequent
risk of sudden inFant death syndrome. N Engl] Med 2004;
351 :978-86.
44 Bass M. The Fallacy of the simultaneous sudden infant death
syndrome in ('wins. Am] Forensic Med Pat/101 19139 ;
10:200-5.
45 Carpenter RG, Gardner A, McWeeny PM , Emery JL.
Multistage scoring system For identifying inFants at risk of
unexpected death. ATCh Dis Child 1977; 52:606-12.
46 Golding J, Limerick S, Macfarlane JA. Sudden Infan/ Death:
Pattems, Puzzles and Problems. London: Op en Books, 1985,
pp. 94-5, 190-5.
47 Ward-Platt M, Blair PS, Fleming PJ et al. A clinical
comparison of SJDS and explained sudden inFant deaths:
ho w healthy and ho w norm a l? Arch Dis Child 2000 ;
82:98-106.
48 Overpeck MD, Brenner RA, Trumble AC et al. Risk factors for
inFant homicide in the United States. N Engl] Med 1998;
339 :1211-16.
49 Stanton AN, Downham MAPS, OakJey JR et al. Terminal
symptoms in children dying suddenly and unexpectedly at
home. Preliminary report of the DHSS multicentre study of
postneonatal mortali ty. 8M) 1978; 2:1249-51.
50 Stanton AN, Oakley JR. Pattern of illnesses beFore cot
deaths. Arch Dis Child 1983; 58:878-81.
51 Ford RPK, Mitchel! EA, Stewart AW et al. SIDS, illn ess, and
acute medical care. Arch Dis Child 1997; 77:54-5.
52 L'Hoir MP, EngelbeJ1s AC, van Well GThJ et al. Sudden
unexpected death in infancy: epidemiologically determined
risk factors related to pathological classification. Acta
Paediatr 1998; 87:1279-87.
53 British Paediatric Association. Report of a Working Party.
Evaluation of Suspected Imposed Upper Airway ObSfillction.
British Paediatric Association. London: Roy Soc Med Press,
1994.
54 Rosen CL, Frost JD, Glaze DG. Child abuse and recurrent
infant apnea.] Pediatr 1986; 109: 1065-7.
55 Southall DP, Plunkett MCB, Banks MW et al. Covert video
recordings of life-threatening child abuse: lessons for child
protection. Pediatrics 1997 ; 100:735-6.
56 Berger D. Child abuse simulating 'near- miss' sudden infan t
death syndrome.] Pediatr 1979; 95:554-6.
57 Ariagno RL, Guilleminault C, Korobkin R et al. 'Near-miss'
for sudden infant death syndrome infants: a clinical
probl em. Pediatrics 1983; 71:726-30.
58 Byard RW, Krous HF. Sudden infant death syndrome:
overview and update. Pediatr Dellelop Patlwl 2003; 6: 1
12-27.
59 Truman TL. Ayoub CC. Consid ering suffocatory abuse and
Munchausen by proxy in the evaluation of children
experiencing apparent life-threatening events and sudden
infant death syndrome. Child Maltreat 2002: 7:138-48.
60 Galvin HK, Newton AW, Vandevan AM. Update on
Munchausen by proxy. Curr Opill Pediatr 2005; 17:252-7.
61 Becroft DMO, Thompson JMD , Mitchell EA. Nasal and
intrapulmonary haemorrhage in sudden infant death
syndrome. Arch Dis Child 2001; 85:116-20.
62 Barber MA, Davis PM. Fits, faints, or fatal fantasy? Fabricated
seizures and child ab use. Arch Dis Child 2002; 86:230-3.
63 Proudfoot A. Is the cause toxicological? Medicine 1999;
27(4):4.
64 Emery JL, Thornton JA. Effects of obstruction to respiration
in infants, with particular reference to mattresses, pillows,
and their coverings. 8M} 1968; 3:209-13.
65 Foundation For the Study of InFant Death s. Reduce the risks
of cot death - an easy guide (Leaflet). FSID, 1996.
66 Foundation For the Study of Infant Deaths. Reduce the ris.~s
of cot death (Leafl et) . FSID, 1991.
67 UNICEF UK. Sharing a bed with your baby. London: UNICEF
UK,2003.
68 Scragg KR, Mitchell EA, Stewart AW et al. Infant room­
sharing and prone sleep position in sudden infant dea th
syndrome. Lancet 1996 ; 347:7-12.
69 Stanton AN. Sudden inFant death. Overh eating and cot
death. Lancet 1984; 2:1199-2 01.
70 Fleming P J, Gilbert R, Azaz Y et a!. Interaction between
bedding and sleeping position in the sudden inFant death
syndrome: a popul ation based case-control study. 8M}
1990; 30:85-9.
71 Wigfield RE, Fleming PJ, Azaz YEZ et al. How much
wrapping do babies need at night? Arch Dis Child 1993;
69:181-6.
72 Beal SM, Finch CF. An overview of retrospective case­
control studies investigating the relationship between prone
sleep ing position and SIDS: } Paediatr Child Health 1991;
27:334-9.
73 Beal SM. Sleeping position and sudden inFant death
synd rom e. Med] Aust 1988; 149:562.
74 de Jonge GA, Engelberts AC, Koomen-Liefting AJM,
Kostense P J. Cot death and prone sleeping position in The
Netherlands. 8M] 1989; 298:722.
75 Mitchell EA, Scragg R, Stewal1 AW et a!. Results From the
first year of the New Zealand cot death study. N Z Med]
1991; 104:71-6.
76 Brooke H, Gibson A, Tappin D, Brown H. Case-control study
of sudden inFant death syndrome in Scotland, 1992- 5. 8M}
1997 ; 314:1516-20.
77 Mitchell EA , Hutchison L, Stewart AW. The continuing
decline in SIDS mortality. Arch Dis Child 2007; 92:
625-6.
 Referen ces I 221

78 Ponsonby A-L, Dwyer T, Couper 0, Cochrane J. Association 102 Rintahaka PJ, Hirvonen J. The epidem iology of sudden
between use of quilt and sudden infant death syn drome: infant death syndrome in Finl and in 1969-1980.] Forensic
case-control study. BM] 1998; 316: 195-6. Sci Intern 1986; 30:2 19-33.
79 Waters KA, Gonzalez A, Jean C et a!. Face-straig ht-down
103 Norvenius SG. Sudden infant death syndrome in Sweden in
and face-near-straight-down posilions in healthy, prone­
1973- 1977 a nd 1979. Acta Paediatr Sco nd Supp/1987;
sleeping infants. ] Pediatr 1996 ; 128:616-25.
333:1-138.
80 Fleming PJ, Blair PS, Bacon C et a!. Environment of infants 104 Rushton Dl. Co-sleeping and sudden infant death syndro me.
during sleep and risk of the sudden infant death syndrome: Lancet 1997; 349:56-7.
results of 1993-5 case-control study for contidential 105 Blair PS, Flemin g PJ, Smith IJ et al. Babies sleeping with
inquiry into still births and deaths in infancy. BM] 1996; parents: case-control study of factors influencing the risk of
3 13:191 -5. the sud den infant death syndrome. BM] 1999; 319:1457-62.
81 Ramanathan R, Chandra 5, Gilbert-Barness E, Fra nciosi R. 106 Tappin 0, Ecob R, Brooke H. Bedsharing, roomsharing and
Sudden infant death syndrome and water beds (Letter to the sudden infant death syn drome in Scotland: a case control
editor). N Eng] Med 1988; 3 18 : 1700. study.] Pediatr 2005; 147:32-7.
82 Bass M. More on sudden infant death syndrom e and water
107 Czegledy-Nagy E, Cutz E, Becker LE. Sudden death in
beds (Letter). N Engl] Med 1988; 3 19:1415.
infants under one year of age. Pedian' Pathol 1993;
83 Gilbert-Barness E, Hegstra nd L, Chandra 5 et a!. Hazards of 13 :671-34.
mattresses, beds and bedding in deaths of infants. Am] lOB Byard RW, Carmichael E, Beal S. How useful is post-mortem
Forensic Med Patl101 1991; 12:27-32. examination in sudd en infant death synd rome? Pediatr
84 Kemp JS, Tha ch BT. Sudden death in infants sleep ing on
Pathol 1994; 14:817-2 2.
polystyrene-tllled cushions. N Engl] Med 1991 ;
109 Parham OM, Savell VH, Kokes CP et a \. Incidence of autopsy
324: 1858-64. findings in unex pected deaths of children and adolescents.
85 Gilbert-Barness E, Emery JL. Deaths of infants on
Pediatr De)} Pathol 2003; 6: 142-55.
polystyrene-filled beanbags. Am ] Forensic IVied Patllol
110 Mitch ell E, Krous HF, Donald T, Byard RW. Changing trends
1996; 17:202-6.
in the diagnosis of sudden infant death. Am] Forensic Med
86 Byard RW, Beal SM. V-shaped pillows and un safe infant
PatllOl 2000; 21 :311-14.
sleep ing.] Paediatr Child Health 1997; 33:171-3.
111 Berry PJ , Keeling J"I/. The in ves tigation of sudden
87 Moo re L, Bourne AJ , BealS e t aJ. Unexpected infant death in unex pected death in infancy. In A nth o ny PP, MacSween
associa tion with suspend ed rocking cradles. Am ] Forel1sic RNM (eds) Recent Advances in Histopath ology 14. London:
Med PatilO11995; 16:177-80. Churchill Livingstone, 1989, pp. 151-279.
88 Bass M. Asphyxial cri b death. N Ellgl] Med 1977; 296:
112 Byard RW, Cohle SD. Sudden Death in Infancy, Childhood
555-6.
and Adolescence. Cambridge: Cambridge Uni versity Press,
89 Smi a lek JE , Smialek PZ, Spitz WU . Accidental bed deaths in 1994.
infants due to un safe sleepin g situatio ns. c/in Pediatr 1977; IJJ Oancea A. Cote A. Roblicek C et al. Cardiac pathology in
11:1031-6. sudd en unexpected infant death.] Pediatr 2002;
90 Byard RW, Beal S, Bourne AJ. Potentially dangerou s s leeping 141 :336-42.
enviro nments and accide ntal asphyxia in infancy and ear ly 114 Cohl e SD, Balraj E, Bell M. Sudden death due to ven tricular
childhood. Arch Dis Child 1994 ; 17:497-500. septal defect. Pediatr Develop Pathol 1999; 2:327-32.
9) Variend 5, Usher A. Broken cots and infant fatali ty. Med Sci 115 Smith NM, Ho SY. Heart block and sudden death associated
Law 1984; 24:111 - 12. with tibrosis of the conduction system at the margin of a
92 Byard RW, Bourne AJ, Beal SM. Mesh -s ided cots: yet ventricul ar septal defect. Pediatr Cardiol 1994; 15:139-42.
anothe r potentially dangerous infant s leeping environme nt. 116 Bajanowski T, Ortmann C, Teige K et a\. Pathological
Forensic Sci Int 1996; 83:105-9. changes of the heart in sudden infant death. In t] Legal Med
93 Cassidy SP. Analysis of crib-related deaths for the years
2003; 11 7: I 93-203.
1989 through 1991 (memorandum). Bethesda, MD:
117 Levi n M, Tizard EJ, Dillon MJ. Kawasaki disease: recent
US Consumer Product Safety Commission, 1994.
advances. Arch Dis Child 1991; 66:1369-74.
94 Foundation for the Study of Infant Deaths. Newsletter FSID 118 Lipsett J, Cohle SD, Berry Pl, Byard RW. Anoma lous
1999; 59:12. cOTOnary arteries: a multicenter pediarric autopsy study.
95 Nakamura 5, Wind M, Danello MA. Review of hazard s
Pediatr Patl101 1994 ; 14: 28 7-300.
associated with children placed in adult beds. Arch Pediatr
119 Berry CL. Myocardial infarcti o n in a neonate. Br Heart ]
AdolescMed 1999; 153:1019-23.
1970; 32:41 2- 15.
96 Moore L, Byard RW. Pathological findings in hanging and
120 Tillett A. Hartl ey B, Simpson l. Parad oxical embQlism
wedging deaths in infants and young chi ldren. Am]
causing fatal myocardial infarcti on in a newborn infant.
Forensic Med Path ol 1993; 14:296- 302.
A rch Dis Child 2001; 35:FI37-B.
97 Bya rd RW, Beal SM, Simpson A et a!. Accidental infant
12 1 Bya rd RW. Vascular conditions. In Byard RW, Cohle SO
death and stroller-prams. Med] Aust 1996; 165:140-1.
(eds) Sudden Death in Infancy, Childhood and Adolescence.
98 Collin s KA. Death by overlaying and wedging. A IS-year
Ca_m bridge: Camb ridge Univers ity Press. 1994,
retrospective study. Am ] Forensic Med Pathol 2001;
pp. 175-261.
22:1 55 - 9.
122 Gilbert -Barn ess E, Barness LA. Nonmalformative
99 Byard RW. Is breast feeding in bed always a sa fe practice ?
cardiovascular pathology in infants and children. Pediatr
] Paedial,. Child Heal th 1998; 34:413-19.
Develop Pathol 1999; 2:499-530.
100 Scragg KR, Mitchell EA. Side sleeping position and bed 123 Fried K. Beer S, Vure E et al. Autosomal recessive sudden
sharing in the sudden infant death syndrom e. Ann Med un expected death in children prob ably caused by a
1998; 30:345-9. cardiomyopathy assoc iated with myopathy. ] Med Genet
JOI Luke lL. Sleeping arrangements of sudden infant death 1979; 16:341-6.
syndrome victims in the District of Columbia: a preliminalY 124 Ed wa rds WO. Cardiomyopathies. HlIm Pathol 1987;
report.] Forensic Sci 1978; 23:379-33. 6:625-35.
222 I Sudden unexpected death in infancy
125 Goodwin Jf. Sudden cardiac death in the young. A family
history of sudd en death need s investigation. BM] 199 7;
314:843.
126 Sua rez V, Fuggle WJ , Ca meron AH et al. Foamy myocardial
transformation of infancy: an inherited disease. ] Ciin Patl70l
1987; 40:3 29 -34.
127 Shehata BM, Patterson K, Thomas JE et aJ. Histiocytoid
ca rdiomyopathy: three new cases and a rev iew of the
literature. Pedialr Devel Patl70l 1998; 1 :56-69.
128 Adgey J, Johnston PW, Math ew TP. Sudden cardi ac death - a
predictable, avoidable and treatable event? Proc R Coli
Physicians Edinb 1998; 28 :350-6.
129 Towbin JA. Molecular genetic basis of sudden cardiac death.
Perliatr Ciin N Am 2004; 51: 1229 -55.
130 Schwartz PJ, Stramba-Badiale M, Segantini A et al.
Prolongation of the QT interval and the sudden infant deat h
syndrome. New Engl] Med 1998; 338:1709-14.
131 Arnestad M, Crolti L, Rognum TO et al. Prevalence of
long-QT sy ndrome ge ne vari a nts in sudden infant death
syndrome. Circulation 2007; 115:361-7.
132 Ackerman MJ, Tester DJ, Driscoll DJ. Molecular autopsy of
sudden unexplained death in the young. Am ] Forensic Med
Patlwl 200 I ; 22: 105- 11.
133 Hanzlick RL. Case report: lingual thyroglossal duct cyst
causin g death in a four-week-old infant. ] Forellsic Sci
1984; 29 :3 45-8.
134 Kanawaku Y, Funayama M, Nata M, Kanetake 1. Sudden
infant death: thyroglossal duct cyst ve rsus environmental
factors. Forensic Sci Int 2006; 156: 158-60.
135 McNamarra VM, Crabbe DC. Tracheomalacia. Paed iatr Resp ir
Rev 2004 ; 5: 147-54.
136 Wong SW, Gardn er V. Sudd en death in children due to
mesenteric defect an d mesenteric cyst. Am ] Fo ren sic Med
Patho11992; 13:2 14-16.
137 Byard RW. Sudden infant death, large intestinal volvu lu s
and a duplication cyst of the terminal ileum. Am ] Forens ic
Med Pat/wi 2000; 21:62-4.
138 Byard RW, Simpson A. Sudden death a nd intussusception in
infancy and childhood-autopsy cons ideration s. Med Sci La w
2001; 41 :41-5.
139 Ng'wa la li PM , Yonemitsu K, Tsuenari S. Fatal
intussusception in infancy: an exp erience in forensic
autop sy. Leg Med (Tokyo) 2003; 5:181 - 4.
140 Smith NM, Bourne J , Clapton WK, Byard RW The spectrum
of presentation at autopsy of myocarditis in infancy and
childh oo d. Pathology 1992; 24: 129-3 I.
141 Shatz A, Hiss J , Arensburg B. Myocarditis misd iagnosed as
sudden infant death syndrome (SlDS). Med Sci Law 1997;
37: 16-18.
142 de Sa DJ. Isol ated myocarditis in the first year. Arch Dis
Child 1985; 60:484-5.
143 Krous HF, Haas E, Chadwi ck AE, Wagner GN. Sudden death
in a neonate with idiopa thic eosin ophilic endomyocarditis.
PerliMr DelJel Pathol 2005; 8:587-92.
144 Dettm eyer RB, Padosch SA, Madea B. Lethal enterovirus­
indu ced myo card itis and pancreatitis in a 4-month-old boy.
Forensic Sci i M 2006; 156:51-4.
145 Losty HC, Lee P, AJfah am M et al. Fatty infiltrat ion in the
liver of medium chain acyl CoA dehydrogenase deficiency.
Arch Dis Child 1991; 66:727-8.
146 Matsubaru Y, Narisawa K, Tada K et al. Preva lence of K329E
mutatio n in medium-chain acyl- CoA dehydro genas e gen e
determ ined from Guthri e ca rds. Lancet 1991; 338:
552-3.
147 Dundar M, Lanyon WG, Con nor JM. Scottish frequency of
the common G985 mutation in the medium-chain Acyl-CoA
dehydroge nase (MCAD) gene and the role of MCAD
deficiency in su dden infa nt death sy ndrome (SlDS).
] inherit Metab Dis 1993; 16:99 1-3.
J 48 Fitzpatrick D. Inborn erro rs of metabolism in the newbo rn:
clinical presentation and investigation. ] R Coli PhYSicians
Erlillb 2006; 36: 147-51.
149 Emery JL, Variend S, Howat AJ, Vawter GF. Investigation of
inborn errors of metabolism in unexpected infant deat h.
Lancet 1988; 2:29-3J.
150 Lundemose JB, Kolvraa S, Gregersen N et al. Fatty acid
oxidation disorders as primary cause of sudden and
unexpected death in infa nts and yo ung children: an
investigation performed on cultured fibrobla s ts from 79
children who died aged between 0- 4 years. ] Clin Patl7ol:
Mol Patho11997; 50:212-17.
151 Boles RG, Buck EA, Blitzer MG et aJ. Retrospective
biochemical sc reening of fatty ac id oxid ation disord ers in
post-mortem livers of 418 cases of sudden death in the first
yea r of life.] Pediatr 1998 ; 132:924-3 3.
152 James CL, Keeling JW, Smith JW, Byard RW Total
anomalous pulmonary venous drainage associated with fat a l
outcome in infancy a nd early ch ildhood - an autopsy study
of 52 cases. Pediatr Parhol 1994; 14:6 65 -78.
153 Keeling JW. The perinatal ne cropsy. In Keeling JW, Khong TY
(eds) Fetal and Neonatal Pathology, 4th edn. London:
Springer-Verlag, 2007, pp. 20-5 3.
154 Byard RW Idiopathic arterial ca lcitlcation and unexpected
infant death. Ped Pathol Lab Med 1996 ; 16:985-94.
155 Champ C, Byard RW Pulmonary thromboembolism and
unexpected death in infancy. ] Paedialr Child Health 1994 ;
30:5 50-1.
156 Berry PJ , Nagington J. Fata l infec tion w ith echovirus II.
Arch Dis Child 1982; 57 :22 - 9.
157 Modlin JF. Perinatal echovirus infection: insights from a
litera ture review of 6 1 cases of serious infection and 16
ou tbreaks in nurseries. Re/J Infect Dis 1986; 8:918-26.
158 Wang J, Atchison RW, Walpusk J, Jaffe R. Echovirus
hepatic failure in infancy: report of four cases with
speculation on the pathogenesis. Pediatr Devel Patl101
2001; 4:4 54 -60.
159 Kaplan MH, Klein SW, McPh ee J, Harp er RG. Group B
coxsackievirus infections in infants you nger than three
months of age: a serious childhood illness. Rev II/fect Dis
1983; 5:1019-32.
160 Wilcken B, Carpenter KH, Hammond J. Neonatal symptoms
in medium chain acyl coenzyme A dehydrogenase
deficiency. Arch Dis Child 1993; 69:292-4.
161 Kirk JM , La ing lA, Smith N, Uttley WS. Neonatal presentation
of med ium-chain acyl-CoA dehydrogenase deficiency in two
families.] Inheri t N/etab Dis 1996; 19:370-1.
162 Vawter GF, Koza kewi ch HPW Aspects of morph o logic
va riation among SJDS-v ictims. In Tild o n JT, Roeder LM ,
Steinschneider A (eds) Sudden Infant Death Syndrome.
New York: Acad em ic Press, J 983, pp . 133-44.
163 Molz G, Hal1mann H. Dysmorphism, dysplasia, and anomaly
in sudden infant death. N Engl ] /vIed 1984; 311 :259.
164 Molz G, Brodzinowski A, Bar W, Vonlanthen B.
Morphologic variations in 180 cases of sudden infant
death a nd 180 controls. Am] Forensic Med Patho11992;
13: 186-90.
165 Blair PA, Nadin P, Cole TJ et al. Wei ght gain and sudd en
infant death syndrome: cha nges in we ig ht z sco res may
identify infants at increased risk. Arch Dis Child 2000;
82:462-9.
J 66 va n Baarlen J , Schuurman H-J, Huber J. Acute thymus
involution in infancy and childhood: a reliable marker
for duration of illness. Hum Pathol 1988; 19:
1155-60.

167 Beckwith JB. Intrathoracic petechial haemorrhages: a clue to
the mechanism of death in sudden infant death syndrome?
Ann Am Acad Sci 1988; 553:37-47.
168 Kleeman W J, Wiechern V, Schuck M, Trager HD.
Intrathoracic and subconjunctival petechiae in sudden infant
death syndrome (SIDS). Forensic Sci Int 1995; 72:49-54.
169 lsakasen CV, Helweg-Larsen K. The impact of attempted
resuscitation in SIDS: post-mortem findings. In Rognum TO
(ed.) Sudden Infant Death Syndrome. Oslo: Scandinavian
Unive rsity Press, 1995, pp. 70-80.
170 Werthammer J , Brown ER, Neff RK, Taeusch HW Jr. Sudden
infant death syndrome in infants with bronchopulmonary
dys plasia. Pediatrics 1982 ; 69:301-3.
171 Berry PJ. Pathological findin gs in SIDS. J Ciin Path 01 1992 ;
45(Suppl.): 11-16.
172 Valdes-Dapena M. The pathologist and the s udde n infant
death syndrome. Am J Pathol 1982; 106: 118-3 1.
173 Emery JL. Dinsdale F. Increased incid ence of Jy mphoreticular
aggregates in lungs of children found une xpectedl y dead.
Arch Dis Child 1974 ; 49:107-11.
174 Krous HF, Nadeau JM, Silva PO, Blackbourne BD. A
comparison of respiratolY symptoms and inflammation in
sudden infant death syndrome and in acc idental or infli cted
infant death. Am J Forensic Med Pathol 2003; 24: 1-8.
175 Gregersen M, Rajs J, Laursen H et al. Pathologic cri teria for
the Nordic study of SIDS. In Rognum TO (ed.) Sudden Infant
Death Syndrome. Oslo: Scandinavian University Press, 1995,
pp. 50-8.
176 Shatz A, Hiss J, Arensburg B. Basement-membrane
thickening of the vocal cords in sudden infant death
syndrome. Laryngoscope 1991; )01 :484-6.
177 Krous FH. Hauck FR, Herman SM et al. Laryngeal basement
membrane thickening is not a reliable post-mortem marker
for SIDS. Results from the Chicago Infant Monality study.
Am J Forensic Med Patho11999; 20:221-7.
178 Adelson L, Kinney ER. Sudden and unexpected death in
infancy and childhood. Pediatrics 1956; 17:663.
179 Valctes-Dapena M, Hoffman HJ , Froelich C, Requeira O.
Glomerulosclerosis in the sudden infant death syndrom e.
Pediatr Pathol 1990; 10:273-9.
180 Naeye RL. Hypoxemia and the sudden infant death
syndrome. Science 1974; 186:837-8.
181 Gilbert-Barness EF, Kenison K, Giulian G, Chandra S.
ExtramedullalY hematopoiesis in the liver in sudden infant
death syndrome. Arch Pathol Lab Med 1991; 115:226-9.
182 Smith NM, Telfer SM, Byard Rw. A compa rison of the
in cidence of cytomegalovirus inclusion bodies in
submand ibul ar and tracheobronchial g lands in SIDS and
non-SID S autopsies. Pediatr PnthoI1992; 12:185-90.
183 Variend S. Infant morta lity, microglial nodules and parotid
CMY-type inclu sio ns. Ea,.ly Hum Deuel 1990; 21 :31- 40 .
184 Takashima S, Armstrong 0 , Becker LE, Huber J. Cerebral
white maner lesions in sudde n infant death syndrome.
Pediatrics 1978; 62: 155-9.
185 Takas hima S, Armstrong 0 , Becker L, Bryan C. Cerebral
hypo perfusion in the sudde n infa nt death syndrome?
Brainstem gliosis and vasc ulature. Ann Neuro l 1978;
4: 257-62.
186 Kinney H, Burger P, Harrell F, Hudso n RP. 'Reactive gliosis'
in the medulla oblongata of v ictims of the Sudden Infant
Deat h Syndrome. Pediatrics 1983; 72: 181-7.
187 Kinn ey HC, Filiano JJ. Brain research in SlDS. [n Byard RW,
KrOllS HF (eds) Suddell Infant D~ath Syndrome. Pmblems,
Progress and Possibilities. London: Arnold, 2001, pp. 118- 37.
188 Kinney He. Abnormalities of the brainstem serotonergic
system in the sudde n infant death syndrome: a review.
Pediatr Deuel Pathol 2005; 8:507-24.
References I 223
189 Perez-Platz U, Saeger W, Dhom G, Bajanowski T. The
pathology of the adrena l glands in sudden infant death
syndrome (SJDS). Inr J Legal Med 1994; 106:244-8.
190 Becker AE, Becker MJ. Fat distribution in the adrenal cortex
as an indication of the mode of intra-uterine death. Human
Pathol 1976; 7 :495-504.
191 Sinclair-Smith C, Dinsdale F, Emery J. Evidence of duration
and type of illness in children found unexpectedly dead.
Arch Dis Child 1976; 51 :424-9.
192 Kaplan JA, Fossum RM. Patterns of facial resuscitation
injury in infan cy. Am J Foremic Med Pathol 1994;
15:187-91.
193 Bu s h CM, Jon es JS, Co hle SO, Johnson H. Pediatric injuries
from ca rdiopulmonary resuscitation . Anll Emerg Med 1996;
28:40-4.
194 Feldman KW, Breweer OK. Child abuse, cardiopulmonary
resuscitation, and rib fractures. Pedia trics 1984 ; 73:339-42.
195 Spevak MR, Kleinman PK, Belanger PL et al.
Ca rdiopulm o nary resuscitation and rib fractures in infants: a
post-mortem radiologic-pathologic study. JAMA 1994;
272:617-18.
196 Price EA , Rush LR, Perper JA, Bell MD. Cardiopulmonary
resuscitat ion- related injuries and homicidal blunt abdominal
trauma in children. Am J Forensic Med Patho12000;
21 :307- 10.
197 Dow nes AJ, Crossland OS, Mellon AF. Preva lence and
distribution of petechiae in well babies. Arch Dis Child 2002;
86:291-2.
198 Perrot U. Masque ecchymotique. Specific or nonspecific
indicator for abuse. Am J Forensic Med Patlrol 1989;
10:95-7.
199 Byard RW, Krous HF. Petechial hemorrhages and unexpected
infant death. Leg Med (Tokyo) 1999; 1 :193-7.
200 Oehmichen M, Gerling I, Meiilner e. Petechiae of the baby's
skin as differentiation symptom of infanticide versus SIDS.
J Forensic Sci 2000; 45:602-7.
201 Betz P, Hausmann R, Eisenmenger W. A contribution to a
possible differentiation between SIDS and asphyxiation.
Forensic Sci Jnt 1998; 91:147-52.
202 Rao V J, Wetli CY. The forensic significance of conjunctival
petechiae. Am J Forensic Med Pathol 1988; 9:32-4.
203 Hood I, Ryan D, Spitz WU. Resuscitation and petechiae.
Am J Forensic Med Patho11988; 9:35-7.
204 McIntosh N, Mok JYQ, Margerison A. Epid emio logy of
oronasal hemorrhage in the first two years of life:
implications for child protection. Pediatrics 2007;
120: 1074-8.
205 Krous HF, Nadeau JM, Byard RW, Blackbolll"ne BD. Oronasal
blood in Sudden Infant Death. Am J Forensic Med Parhol
2001 ; 23:346-51.
206 Sugar NF, Taylor JA, Feldman KW, Puget Sound Pediatric
Research Network. Bruises in infants and toddlers. Those
who don't cruise rarely bruise. Arch Pediatr Adolesc Med
1999; 153 :399-403.
207 Amir J, Katz K, Grunebaum M et al. Fractures in premature
infants. J Pediatr Orthop 1998 ; 8:41-4.
208 Koo WW, Sherman R, Succop Pet al. Fracture and rickets in
very lo w birth weight infants: co nservative management and
outcome. J Pediatr Orthop 1989 ; 9:326-30.
209 Dabezies EJ , Warren PD. Fractures in very low birth
weig ht infants w ith rickets. Ciin Orthop Relat Res 1997;
335:233-9.
210 Barry PW, Hocking MD. Infant rib fracture - birth trauma or
non-a ccidental injury? Arch Dis Child 1993; 68:250.
211 Bullock B, Schu bert CJ, Brophy PO et al. Cause and clinical
characteristics of rib fractures in infants. Pediatrics 2000;
105:e48.
224 I Sudden unexpected death in infancy
212 Durani Y, Depiero AD, Rib fractu re and birth trau ma, Ann
Emerg Med 2006; 47 :2 10,2 15,
213 Campbell CJ, Read DJ(, Lung petechiae - their
pathophysiology and diagnostic sig nificance in co t death,
Aust Physiol Pharmacol Soc 197 8; 9 : 143,
214 Berry PJ, Intra-al veo lar haemorrh age in sudd e n infant death
syndrome: a cause for concern? ) Ciin Pa thol 1999;
52 :553-4,
215 Hanzlick R, Pu lmon ary hemorrhage in deceased infants:
ba selin e d ata for fu rther study of infant mo rtali ty,
Am) Forensic Med Path 01 2001 ; 22: 188-92,
216 Coffin CM , Schechtman K, Col e FS, Dehner LP, Neonata l and
in fantil e pulmo na ry hemorrhage: an autopsy study with
clinica l co rrelation , Pediatr Pathol 1993 ; 13:583-9,
21 7 Golde DW, Drew Wl, KJei n HZ et a l. Occult pu lmona ry
haemorrh age in leukae mia, BM) 1975; 2: J 66 -8,
218 Keeling JW, Maxwell RS, Busuttil A. Haem osideri n in infant
lun gs, ) Pa thol 1998 ; 184:29A,
219 YlIkawa N, Carter N,RlItty G, Green MA. Intra-alveo lar
haemorrh age in sudden infant death syndrom e : a cau se fo r
concern?) Ciin Pathol 1999 ; 52:581-7,
220 Berry PJ, Lee J, Alveolar s id erop hages a nd the sudd en infant
death synd rome,) Pathol 1998 ; 184:27A.
22 1 Stewart S, Fawcett J. Interstitia l haemosid erin in the lungs of
sudden infan t death sy ndrome: a histol ogica l hallmark of
'near-m iss' episodes? ) Patlwl 1985 ; 145 :53 -8,
222 Byard RW, Stewart WA, Tel fer S, Beal SM, Assessment of
pulmon ary and int ra thymic hemosiderin deposition in
sudd en infant death sy ndrom e, Ped iatr Pathol Lab Med
1997 ; 17:275- 82.
223 Becroft DMO, Thompso n JMD, Mi tchell EA. Pulmonary
interstiti a l haemosiderin in in fa ncy: a common con seq uence
of normal labour. Pedia tr DelJel PatllOl 2005; 8 :448- 52,
224 Becroft DM, Lockett BK, Intra-a lveo lar pulmona ry
sid erophages in sudden infant death: a marker for previous
impose d suffocation, Pathology 1997; 29 :60-3,
225 Milroy CM. Munchausen syndro me by proxy and in tra ­
al veolar haemosid erin. Int) Legal Med 1999; 1112 :
309-12.
226 Dorandeu A, Peri e G, Jouan H et al. Histolog ical
demonstration of haemosiderin deposits in lungs and liver
from victims of chronic physica l child ab use. Int) Legal
Med 1999; 11 2:280-6 ,
227 Fagan DG, Ha emosiderin in pulmonary macrophages from
SUD , SIDS and deaths in an fVTVA.) Patho11997; 182:49A.
228 Han zli ck R, Del aney K. Pulm onary haem os iderin in dece ased
infants: baselin e data for furth e r study of infant mortality.
Am) Forensic lVled Pat/wi 2000; 21:319-22,
229 Schlu ckebier DA. Coo l CD, Henry TE et al. Pulmon ary
sid erop hages and unexpected infant death , Am) Forens ic
Med Pat/wi 2002; 23:360-3.
230 Krous HF, Wixom C, Chadwick AE et aL Pulmona ry intra­
alveo lar siderophages in SIDS and suffocation: a San Diego
SID S/SU DC project report, Pediatr Deliel Pathol200G;
9:103-J4.
231 Monta na E, Etzel RA, Allan T et aL En vironmenta l risk
factors associ ated with pe diatric idiop athic pulmon ary
hemorrhage and hemosid eros is in a Cleveland community.
Pediatrics 1997 ; 99:3 1-8,
232 Pappas fVlD , Sa rn a ik AP, Meert KL et aL Idiopathic
pulmonary hemorrhage in infa ncy. Clinical features and
management with high frequency venti lation, Chest 1996;
110 :553 -5.
233 Cutz E, Perrin DG, Vujani c GM, Ackerly C. Idiopathic
pulmonary haemos iderosis (IPH) presenting as sudd en
unex pected de ath SUD in infancy (Ab stract), Anales
Espanoles de Pediatria 1999; 92(Suppl.):52,
234 Mu ir R, Niven J SF. The loca l form ation of bl ood pigments,
) Pa rhol Bacteriol1935; 41:183-97.
235 Sherman JM , Winnie G, Thomassen MJ et al. Time course of
hemos iderin production a nd clearance by huma n pulmonary
macrophages, Chest 1984 ; 86:409- 1I.
236 Este rley JR, Oppe nheimer EH. Massive pulmonary
haemorrhage in the newbo rn, Pathol og ic considerat ions.
) Pediatr 1966 ; 69:3-11.
237 Rushton DI. Liver and gallbl adder. In Kee lin g.JW (ed.) Fetal
And Neona tal Pathology, 3 rd edn, Lond on: Springer-Verlag,
2001, pp, 409-39,
238 McCann J , Reay 0 , Siebert J et al. Postmortem perian al
findings in children, Am ) Foren sic lVled Path ol 1996;
17:289-98 ,
239 Case ME, Graham MA, Handy TC et al. The National
Association of Medical Examiners Ad Hoc Committee on
Shaken Baby Synd rome, Position paper on fatal abusive
head injuries in infan ts and young children, Am ) Foren sic
Med Path o12001; 22:112-22,
240 Bodian M, Hes lop B, Sudde n infan t death syndrome , 1n
Siim J-C (ed,) Proceedings of the Eighth 11Iternational
Co ngress of Paediatrics, Basel, 1956. Copenhagen:
Williams 8: Wilkins, 1960, p. 91.

SUDDEN NATURAL DEATH IN
INFANTS AND CHILDREN
Dick Variend
Introduction 225
Cardiovascular causes of sudden death 226
X-linked hypohidrotic (anhidrotic) ectodermal
dysplasia 235
Intracranial haemorrhage, neoplasms and
malformations 236
Gastrointestinal causes 239
Fatal anaphylaxis 239
Sickle cell disease 240
Haemorrhage as a cause of sudden death 240
Respiratory causes of sudden death 240
Epilepsy and sudden death 242
INTRODUCTION
heart disease), whereas deaths associated with asthma ,
Sudden death refers to the instantaneous death of an appar­
ently healthy person, but in practice the definition is
extended to include all deaths within 24 hours from the onset
of symptoms. The period between the onset of symptoms and
the time of death is called the 'terminal event'. The term 'sud­
den' generally describes the rapidity of the death and the term
'unexpected' describes the surprising nature of the death .
There is a wide range in the degree to which death may
be expected. Some conditions, most notably cardiovascular
in origin, are more prone to cause sudden death, and preced­
ing knowledge of t heir existence tends to redu ce the level of
surprise. By their nature, sudden deaths tend to occur at
home, in the ambulance or soon after arrival at hospital. As
might be expected, sudden deaths have enormous impact
on the family, the community and medical attendants. They
are relatively uncommon occurrences but, when they occur,
tend to foster suspicion and, in the United Kingdom, are
genera lly dealt with by the coroner or procurator fiscal.
Many causes of sudden death are age related. Explained
sudden deaths in the first year of life are more likely to be due
to an infection or congenital abnormality (e.g. congenital
I CHAPTER 12 I
Deaths from acute asthma 243
Diabetes mellitus 243
Genetic metabolic disorders 244
Other bacterial infections 247
Deaths related to obstetric events and
premature birth 248
Miscellaneous causes of sudden natural death 248
Sudden unexplained death in older children 249
Sudden natural death in the early neonatal period 249
Sudden death associated with 'intermediate' pathology 249
References 250
epilepsy and diabetes mellitus are more likely to affect the
ado lescent years.
The terminal event may be non-existent, non-specific or
insignificant. Some patients (or their carers) may not fully
perceive the severity of the symptoms that precede death,
and this may delay summoning medical assistance. A lack of
self-perception has particular relevance in conditions such
as asthma and hypoglycaemia. Symptoms may be modified
by medic atio n. Th e underlying disease mayor may not be
symp tom atic. vVhen symptoms are present the terminal
event can be variable in duration, depending on the under­
lying disease. Those dyi ng fro m hypertrophic cardiomyopa­
thy or congenital aortic stenosis can be expected to have a
short terminal event. It might be argued that distinction
should be made between sudden death in apparently healthy
subjects and in those with recognized preceding illness
(e.g. achondroplasia, cardiovascular disease). It is debatable,
however, whether the latte r group should be included in the
definition of sudden death.
The inability of the very young to adequately communi­
cate symptoms may partly account for their increased vul­
nerability to sudden death. The discrepancy sometimes
226 I Sudden natural death in infants and ch ildren
encountered between the account given and the gravity of
the post-mortem findings may be explained by failure of the
patient (or attend ant) to fully appreciate the seriousness of
symptoms . Sudden death is often accompanied by emotional
turmoil in the fam ily, and this may affect fam ily members'
ability to give a clear account of events leading to death.
Relevant information may come to light only after the
autopsy has been completed. On the other hand, family
members (or med ical attendants) may accurately perceive
symptoms but misinterpret their significance. Fo r instance,
the parents of a young child may mistake melaena for diar­
rhoea and not feel the need to ask for urgent medical advice.
Severe haemorrhage may thus go unnoticed.
A history of chronic disease (e.g. diabetes mellitus, aller­
gic asthma and epilepsy), w hile clearly very useful, does not
always clarify the mechanism of sudden death. The cause of
death in such a n individual as determined at post-mortem
may be entirely unrelated to the clinically diag nosed entity.
Sudden death is well known to be associated with certain
syndromic entities (e.g. achondroplasia, La rsen 's syndrome),
and knowledge of the relevant pathop hysiology may point
the pathologist to the system affected. Deaths occurring
during exercise are more likely to be associated with the
cardiovascular system and , in such cases, a positive family
history or evidence of preceding cardiac dys rhythmia may
be helpful.
The cause of sudden death at the time of autopsy may be
obvious to the naked eye or on microscopic analysis; those
without discernible findings require further diligent search
or special investigations. After available investigation has
been exhausted, those without morphological substrate are
suitably referred to as 'physiolo gical deaths' which theo ret­
ically may cover causes such as cardi ac arrhythmia, con­
vulsion, laly ngeal spasm or unstable respiratory control.
They cover a wide age spa n and may affect the velY yo un g
or o ld er child.
Medical intervention frequen tly prolongs life following
a sudden loss of consciousness or cardiovascular collapse,
with death following a period of intensive care. The circu­
latory collapse and the medical intervention that foUow
often produce changes at post-mortem that are likely to
mask the picture produced by the initial event. Separation
of such ischaemic or reperfusion injuries from changes
directly caused by the initial event is important in deter­
mining the underly ing cause. The results of biochemica l
and enzymological studies may be difficult to interpret in
such 'lingering' deaths. These cases are generally in cl uded
in the definition of sudden death.
The distinction between explained or unexplained sudden
death often depends on the thoroughness of the autopsy
investigation, the experience of the pathologist and avail ­
ab ility of laboratory resources. When facilities are ava il ab le
for t he investigation of genetic metabolic disease, the num­
ber of explained sudden deaths is more likely to increase. In
some cases the relevance and significance of certain lesions
found at autopsy may be difficult to determine. Pathologists
may, for instance, vary in their interpretation of the severity
and/or exten t of certain diseases (e.g. pneumonia or hydro­
cephalus) and this may have a sig nificant bearing in dete r­
mining the cause the death in any particular individual.
Such t1ndings, on the other hand, may be regarded as coin­
cidental.
Death ma y also be accelerated when infection affects a
pa tient with congenital or acqu ired immune deficiency or
when the disease is caused by a part icularly vilUlent infec­
t ive agent.
CARDIOVASCULAR CAUSES OF SUDDEN DEATH
Sudden cardiac death in children is far less frequent than in
the adult population, and condi t ions that predispose to such
deaths in the young are quite different from those preva lent
in later years. The preva lence of cardiovascular disease as a
cause of sudden death is difficult to determine because of
variation in definition of sudde n death and inconsistency in
the examination of the conduction system of the heart.
In the series of Lambert et al,] the following heart con­
ditions accounted for 52 per cent of the cases of sudd en
death: congenital aortic ste nosis, Eisenmenger's syndrome,
cya notic congenital heart disease with pulmonary stenosis
and hy pel1rophic obstructive ca rdiomyo pathy. Arrhythmia,
hypo xia and coronary insufficiency are recog nized modes
of sudden card iac death but, in a substantial proporti o n of
the cases, the mechanism is ill-defined. I
Cardiac Malformation
Unrecog nized congenital ca rdiac malformation is an impor­
tant cause of morta li ty in the first year of life. 2 Of 185 cases
of infant death from congenital heart disease investigated,
70 per cent had previously diagnosed heart disease; in the
remaining 30 per cent, the heart disease was unsuspected or
unconfirmed. Of those with unsuspected heart disease about
one-half had other severe abnormalities, mainly related to
trisomy 13, 18 and 21. Two babies with severe aortic steno­
sis (both aged 2 months) and interlUption of the aortic arch
(aged 6 days) died sud den ly at home. The authors speculated
that 200 babi es died each yea r in the UK from unsuspected
congenital heart disease. Of these, one-half of the heart
defects were unassociated with other malformations.
After infancy, death from congenital heart disease is more
likely to be due to abnormalities of the coronary arteries,
post-operative congenital heart disease, aortic valve stenosis,
conduction system abno rmalities and dissection of the aorta
complicating Marfan's syndrome, aortic isthmic coarctation
or isolated bicuspid aortic valve. 3 The mechanisms of death in
this group include co ronary thromboemboli c phenomena,4
fatal cardiac alThyth mias and severe pulmonary vascular dis­
ease.] Post-operative congenital heart disease is an imp0l1ant
cause of cardiac dysrhythmia leading to sudden death;5 the
 Cardiovascular causes of sudden death I 227

defects mostly affected are tetralogy of Fallot, tra nsposition deaths occur in winter months.17 Of 207 cases of sudden
of the great arteries and double-outle t right ventricle. An death among individuals aged between 1 and 21 years,
older age at the time of operation appears to be a significant Neuspiel and Kuller 20 found myocarditiS to be the predomi­
risk factor for sudden death. nant cause of cardiac death. Human myocarditis has been
Of 20 patients reported by Steinberger et al 6 who died associated with a number of viruses, most commonly of the
suddenly at the age of less than 12 months, 13 (65 per cent) coxsa ckie group.21 Presenting symptoms are related to the
had a cardiac abnormality, 10 of which were anomalies of age of the patient, and older subjects may complain of chest
the coronary aliery and, of these, 5 were initially consid­ pain, but symptoms are more often non-specific, especially in
ered to be sudden infant death syndrome. infants.22 Myocarditis may also present with cardiac failure.
Congenital aortic stenosis is a rare but well-recognized Pathological findin gs include cardiac dilatation, mottling
cause of sudden death,l and preceding symptoms such as of the myocard ium and variable opacification of the endo­
syncope, fatigue, shortness of breath a nd dyspnoea on cardium. There is usu ally a diffuse interstitial mononuclear
exertion are often present. Dea ths tend to predominate in inflammatory infiltrate with a predominance of lymphocytes
the second decade of life and are often precipitated by on microscopic examination (Fig. 12.1).12,16 Interstitial
exertional exercise. B oedema and scattered foci of necrotic muscle fibres are usu­
Intimal tears and aortic dissection leading to intraperi­ ally present. Both ve ntricles and atria may be involved. Viral
cardi al rupture and cardiac tamponade is a recognized com­ inclusions are usually absent. The lesion may be foc al or dif­
plicat ion of undiagnosed aortic coarctation.] Patients with fuse and may show a predilection for the endocardium or
Eisenmenger's syndrome (pulmonary obstructive vascular subepicardial region. Specialized areas such as the conduc­
disease secondary to communication between the systemic tion system may be involved. A subepicardial distribution is
and pulmonary circulations) have a moderately high risk of often ass ociated with pericarditis. The heart weight is often
sudden death, with many of the deaths occurring in the sec­ mildly to moderately increased, suggesting a latent phase of
ond decade. I myocarditis despite a short clinical history. 12 However, when
A ventricular septal defect is a common cardiac anomaly, involvement is confined to areas such as the conduction sys­
and the releva nce of its discovery at autopsy in the sudden tem , the heart weight may be normal. 19
death of an infant or young child is often difficult to
assess. 9.10 Marked cardiac hypertrophy, pleural and intraperi­
toneal effusions , pulmonary oedema and 'heart failure' cells
in the lungs are features of decompensation and implicate
the septal defe ct in the cause of death. 9 A multinational
study of 186 cases attributed sudden death to ventricular
septal defect in two cases (one per cent), suggesting its rarity
as a cause. I A ventricular septal defect may also form the
basis of Eisenmenger's syndrome, resulting in sudden death
much later in childhood or in the second decade of life. I
Williams' syndrome is a malformation complex charac­
terized by typical facies, mental retardation, mild growth
deficiency and cardiovascular disease. Patients are at
increased risk of sudden death from cardiac deformities
that comprise supravalvar aorti c stenosis, bilateral outflow
tract obstruction and coronary al1elY stenosis . Myocardial
ischaemia, with or without cardiac arrhythmia, from coro­
nary aliery stenosis is the most likely mechanism of
death, II but severe bilateral outflow obstruction may also
cause sudden de ath . Abnormal elastin is thought to be
responsible for the cardiovascular disease, and the condi­
tion is linked to the elastin gene on chromosome seven. ll

Myocarditis

Myocarditis is an important cause of mortality in infants l2

and older children,J3-15 and sudden death is a well-recognized Figure 12.1 Microscopic view of the myocardium from a
presentation. 16,17 This may be related to ventricular asystole, 7-year-old girl who collapsed at home and died soon after arrival
ventricular fibrillation or conduction defects. IB ,19 Infants are in hospital. Extensive mononuclear cellular infiltrat ion of the
more commonly affected than older children, and more of the myocardium is seen.
228 I Sudden natural death in infants and children

The heart may be specifically affected ('isolated'
myocarditis) or cardiac involvement may be part of a gen­
eralized involvement ('incidental' myocarditis). 12, 17 A prob­
lem often faced by the pathologist, and for which there is
no easy answer, is how many foci of inflammatory cells in
the myocardium are sufficient to cause death. 16 Noren
et al 17 demonstrated the presence of viral myocarditis as a
coincidental finding in two children who suffered violent
deaths. This calls for caution in attributing death to viral
myocarditis in children who die unexpectedly.
It is important to save samples of heart muscle in aJl cases
of sudden death for virological investigatio n. Diagnostic
methods advocated for clinical investigation are also appli­
cable at autopsy. These include serology for specific viral
antibody, viral culture using tissue or fluid and morphologi­
cal determination of the vil1Js in infected cells; the last may
be au gmented by specific probes for viral antigens or nucleic
acid sequences using in situ techniques for probe visualiza­
tion. 22 Molecular techniques such as polymerase chain reac­
tion (PCR) have proved valuable in identifying viral genome
in formalin-fixed paraffin-embedded tissue.2J - 25
Contraction band necrosis in 'neuro genic cardiomyopa­
thy' and catecholamine-induced cardiomyopathy may cause
minor inflammatory changes that should not be interpreted
as myocarditis. 26 ,27 Myocarditis can also occur with any bac­
terial infection (meningococcus, diphtheria, Staphylococcus
aureus, pneumococcus, gonococcus and Haemophilus
injlurnzae).28 With the notable exception of diphtheria, car­
diac involvement in these cases is likely to be incidental.
In acute rheumatic fever, pancarditis occurs in 50-75 per
cent of children and acute rheumatic carditis may present as
sudden death.29 Sudden death in a 6-year-old girl with acute
rheu matic carditis complicating by thromboembolic occlu­
sion of the left anterior descending coronary artery has been
described. 4 Acute myocard itis has also been reported in
patients with dermatomyositis 30 and Kawasaki's disease.31
Over 50 per cent of cases are inherited as an autosomal
dominant trait (familial hypertrophic cardiomyop athy).33,35
The disease may commence in infancy, and death has been
recorded in infants as young as 1 year of age. 3 Studies in
affected families, however, do not show Signifi cant num­
bers of infant deaths.32 Genes on five loci on separate chro­
mosomes are now known to be responsible for the familial
disease.32 Of the three genes that have been identified, the
best characterized so far is the one that encodes for ~ ­
heavy chain myosin, which is found in abou t 50 per cent of
affected families. At least 30 different point mutations are
kno wn for this gene.
Characteristic of this condition is an increase in ventricu­
lar muscle mass, with histology marked by myocyte disarray
(disorganization), broad and misshap en individual myocytes
and circular alignment of myocytes around central foci of
connective ti ssue 32 (Fig. 12,2). Interstitial and replacement
myocardial fibrosis and acute or subacute myocardial necro­
sis may be present. 3 The histological changes are maximal in
areas of macroscopically thickened wall.
The exact mechanism of sudden death in this condition is
still unclear, but theories include inappropriate circulatory
reflexes leading to haemodynamic collapse,36 arrhythmia an d
conduction abnormalities. 37 Cases with the histological char­
acteristics of HCM may occur in the absence of hypertro­
phy,16,36 Consequently, extensive histological examination

Cardiomyopathy

HYPE RTRO PHIC CARDIOMYOPATHY AND VARIA NTS,

Hypertrophic cardiomyopathy (HCM) is a primary myocar­

dial disease of unknown cause characterized by a hypertro­
phied, non-dilated left ventricle in the absence of another
cardiac or systemic disease,3 2 Symmetrical and asymmetrical
forms exist. It is essentially a disease of young subjects, who
are often asymptomatic. Clinical findings are ventricular
hypertrophy, predominantly involving the ventricular sep­
tum, with dynamic obstl1Jction to left ventricular outflow,
Because of variable phenotypic expression, the incidence of
the condition is difficult to assess,33 It is the most common
cause of sudden exertional death in yo un g persons, and
males are more often affected than females. A combination
of young age, a family his tory of sudden death owing to Figure 12.2 Microscopic view of the myocardium showing
HCM and unsustained ventricu lar tachycardia identifies a myofibre disarray and misshapen myofibres in an 8-year-old girl
subset that is esp ecially prone to sudden dea th. 34 who died of hypertrophic cardiomyopathy.

may be required in cases of sudden death to confinn or
exclude the di agnosis.
A spectrum of disorders is responsible for other cases of
HCM that are increasingly attributed to metabolic, genetic
and molecular disease (e.g. fatty acid oxidation defects,
mitochondrial disease, glycogen storage disease types 2 and 3,
Gaucher's disease, GM 1 and GM2 ga ngliosidosis, sialidosis
and mannosidosis). Cardiomyopathy occurs in about 25 per
cent of patients with Noonan's syndrome J5 and the
myocardium in affected patients is histologically indistin­
guishable from familial HCM.J8 Other associations include
Friedreich's ataxia, Turner's syndrome and some disord ers of
neural crest tissue. 16 Catecholamine-induced cardiomyop a­
thy has been reported in patients with phaeochromocytoma.
Neonatal cardiomyopathy may follow poorly controlled
maternal diabetes. J9 Affected infants are usu ally macrosomic
reflecting the growth hormone effect of insulin. An underly­
ing anatomica l malformation of the heart must be excluded.
Ventricular hypertrophy may persist for up to 2 years when
such patients are followed by echocardiography.J2 Some
vacuolar and hydropic changes of the myocardium are seen
microscopically, but the myocardial fibre disarray and
inflammatory infiltration that characterize classic HCM do
not occur. Cardiac dilatation, which is also sometimes seen,
may be secondary to hypo glycaemia and acidosis?9
Sudden death can occur from severe left ventricular
hypertrop hy from any cause.J2
DI LATED CARDIOMYOPATHY
Dilated cardiomyopathy, an uncommon disorder in child ren,
is characterized clinically by biventricular dilatation and
reduced myocardial contractility.4o It acco unts for a dispro­
portionate number of deaths in paediatric practice, and the
condition has been associated with sudden death,5,41
altho ugh the causative mechanism is poorly understood.
Macroscopic pathological features are an increased left
ventricular mass and decreased left ventricular wall thick­
ness.J2 Myocyte hypertrophy, myocyte myofibrillaJY loss,
interstitial fibrosis a nd increased numbers of interstitial lym­
phocytes co nstitute the histological picture. Interstitial fibrosis,
myocyte hypertrophy and absence of myocyte damage per­
mits the condition to be distinguished from acute myocarditis.
As in hypertrophic cardiomyop athy investigation
should be directed at an underlyin g cause,40 but in the
majority of children dilated cardiomyopathy is idiopathic.
A favoured view is that some cases are a sequel to viral
myocarditis. 16,22,42
ENDOCARDIAL FIBROELASTOSIS
Endocardial fibroelastosis is customarily divided into pri­
mary and secondary types. 29 Most infants with the primary
form of the disease die in the first year of li fe and sudden
infant death is a common presentation. The affected ventri­
cle may be constricted or dilated and fibro elas tic thickening
Cardiovascular causes of sudden death I 229
of the endocardium is seen macroscopically. Twelve cases
(seven per cent) of endocardial fibroelastosi s were included
in the 186 cases of sudden cardiac death in children sur­
veyed by Lambert et al. 1 Secondary endocardial fibroelasto­
sis may be superimposed on cardiac malformation or
myocardial disease, and sudden death may be related to the
underlying disease. 4J
ARRHYTHMOGENIC RIGHT VENTRICULAR DYSPLASIA
(RIGHT VENTRICULAR CARD IOMYOPATHY)
Arrthythmogenic right ventricular dysplasia (ARVD) or car­
diomyopathy is a diso rd er of heart muscle 44 ,45 of unknown
prevalence. 46 ,47 The disease is often familial (about 30 per
cent of cases) with an autosomal dominant inheritance 46
and appears to be related to abnormal developmen t of the
myocardium. 47 A gene for ARVD has been mapped to chro­
mosome 3p23. 48 Sudden death or a rrhythmia of ri ght ven ­
tricular origin is the usua l mode of presentation. Symptoms
ra rely appear before 20 years of age and death often occurs
during exertion. 44 ,45 The rate of sudden death is about one
per cent of cases per year. 47
The usual macroscopic appearance at autopsy is marked
dilatation of the right ventricle with replacement of its free
waJl by adipose tissue;49 microscopy confinns the presence of
adipose tissue or fibrofatty tissue (Fig. 12.3). Inflammatory
\.. . ,
"\ -'4
""\ '
i-
. ... ~
,~
'­
(' " 1
'~
Figure 12.3 The fre e wall of the right ventricular myocard ium is
extensively replaced by adipose tissue consistent with a diagnosis
of arrhythm ogenic right ventricular dysplasia.
230 I Sudden natural death in infants and chi ldren
cells are seen in about 25 per cent of cases, raising the possi­
bility of an infectious or a genetical ly determined immune
process in some of the cases. Heart weight is normal or only
moderately increased.45 Similar histological features are
rarely seen to involve the left ventricle.47 As some degree of
adipose tissue is often seen in the free wall of normal right
ventricle, the amount of adipose tissue required for the diag­
nosis of ARVD is sometimes unclear. 49 ,50
ARVD and Uhl 's anoma ly are now considered to be two
distinct morphological entities. 49 Uhl 's anomaly refers to
absence of the right ventricu lar parietal myocardium, the
endocardium and epicardium being directly apposed; the
condition usually presents in neonates and infants with
congestive cardiac failure .
HISTIOCYTOID CARDIOMYOPATHY
Histiocytoid card iomyopathy (or oncocytic cardiomyopa­
thy), is a rare condition beset with a multitude of syn­
onyms, The entity is sometimes regarded as a tumour of the
myocardium (see below), 51 Females less than 24 months of
age are chiefly affected. 52 The clinical presentation is vari­
ab le, with congestive cardiac failure, arrhythmias or sud­
den death all being encountered. In a large revlew,52 68 per
cent of the patients were reported to experience cardiac
arrhythm ia before their death, Sudden death occurred in
22 per cent of the cases.
The aetiology is unknown, but the characteristic
myocardia l changes may represent the final common path­
way of a number of different aetiologies,53 Familial occur­
rence has been reported S3 At auto psy the heart is
hypertrophied with left ventric ular prominence. The
myocardium is pale and may show tan -yellow nodules. The
endocardium may be thickened. Light microscopy shows
groups of large polygonal cells with foamy to granular
cytoplasm scattered within the myocardium (see Fig, 11.2,
p. 206) . Their distribution may be diffuse, focal or multifo­
cal. Any part of the heart can be involved and the conduc­
tion system may be affected predominantly.54
Ultrastnrctura ll y, the swollen cells contain numerous
enlarged mitoc hondria and myofibrils and occasional lipid
vacuoles that are displaced towards the cell periphery.
Sma ll dense bodies may be found within the mitochondria.
Accompanying cardiac malformation has been described,
and the patient reported by Koponen and Siegel (1996)54
had Peter's anomaly and congenital glaucoma. A defect in
complex III of the respiratory chain has been demonstrated
in one case, 52 However, the condition differs from mito- ..
chondrial cardiomyopathy in several respects:
• In mitochondrial cardiomyopathy, all myocytes are
affected; whereas in histiocytoid cardiomyopathy,
involvement is focal.
• In mitochondrial cardiomyopathy, the mitochondria
are consistently abnormal in shape. In histiocytoid
cardiomyopathy, they are stnrctura lly normal or
abnormal but are numerically increased. 52
(a) (b) (c)
Figure 12.4 Different abnorma l ang les of origin of the coronary
artery wh ich may be associated with sudden death. (al. normal
ang le; (b), angle of origin less than 30°; (c), coronary artery is
invested in the aortic adventitia as it departs at an acute angle.
Anomalous Coronary Arteries
Sudden death associated with anomalous coronary arteries
covers a wide age range (neonates, infants, children and
young adu lts). Death is commonly associated with exertion
and the anomaly is an important cause of sudden death
among athletes. 41 Sudden infant death syndrome may be
mimicked in the very young,55.56 indicating that death is not
always related to exercise. In one autopsy series the inci­
dence of anomalous coronalY arteries as a cause of sudden
infant death was estimated to be 0.4 per cent. 56 One patient
reported by these authors had coexisting vascular abnorma l­
ities. A wide range of anomalies were described: 3,6,41.57,58
(I) an ectopic origin from the wrong coronary sinus, eccentric
origin from the correct sinus or a high aortic origin; (2) ostial
stenosis, i.e. an internal diameter that is smaller than the
artery 1-2 mm from the aorta (this includes a slit-like
ostium); (3) an abnomral angle of origin (Fig. 12.4), i.e. the
ang le between the aortic lumen and coronary ostial lumen
(normally 90 degrees) is between 30 and 45 degrees (proba­
bly abnormal) or less than 30 degrees (definitely abnormal);
or (4) an ostial 'flap' or 'ridge' - a type of ostial stenosis usu­
ally associated with an acute angle of aortic origi n or origin
from the wrong aortic sinus.
These various lesions may occur alone or in combina­
tion, as exemp lifi ed in a 12-year-old child who died sud­
denly and was shown to have several different anomalies
of the left coronary artely. These included a high aortic ori­
gin, an aortic intramural segment ('intussusception'), inter­
positio n between the pulmonary artery and ao rta, as well
as an intramural course. 58 Lesser degrees of complexity are
also descri bed. 55,57
An ectopic origin (anomalous origin from the wrong
aortic sinus) is the most common pattern, and origin of the
left coronary artery from the right coronary sinus is more

8r
LCA
RCA
(a) LAD
RCA
(bl LAD
Figure 12.5 Anomalous origin of the left (a) and right (b)
coronary artery from the wrong coronary sinus and their
relation ship to the aortic and pulmonary trunks.
common than origin of the right coronary artery from the
left coronary sin us. 59 The following, singly or in combina­
tion, may contribute to ischaemia of the myocardium: a
stenosed ostium or one with an acute angle of take-off; a
course interposed between the pulmonary and aortic trunk
(Fig. 12.5); and investment of the coronary artery in the
aortic adventitia as it departs at an acute angle. An origin
of one coronary artery from the other close to its com­
mencement is also reported. 56 Coronary arteries arising
from the wrong coronary sinus causes sudden death from
ventricular fibrillation owing to poor perfusion of the mus­
cle mass. Over two-thirds of patients do not have a history
of syncope or chest pain ; a history of palpitations may,
however, be apparent in old er subjects.
Because of the sma ll size of vessels and difficulty with
dissection in sma ll babies, the anomal ies are easily over­
looked, especially when the coronary ostia are normally
located. Infants do not usually show myocardial ischaemia
on microscopic examination. 56 By contrast, foci of myocar­
dial necrosis, fibrosis or contraction bands are often found in
the ventricular myocardium of older individuals. 3.5.41 .
Other anoma lies of the coronary arteries associated with
sudden death include aplasia/hypop lasia, aneurysm and fis­
tula. The significance of intramyocardial tunnelling (intra­
mural coronary artery) as a cause of sudden death is
controversial. 5o Some consider an intramyocardial mid­
segment of the left anterior descending artelY as a normal
variant,6l while others view as significant a deep and long
intramyocardial course.57 The latter authors recorded an
Cardiovascular causes of sudden death I 231
intramyocardial course from 10 to 25 mm and a depth from
4 to 8 mm in cases of sudden death. While these deaths were
mainly of older individuals, one was a 14-year-old boy.
Origin of the left coronary artery from the pulmonary
trunk (Bland-White-Garland syndrome) is a rare malforma­
tion. The relative low pressure of the pulmonary artery is
inadequate for perfusing the left ventricular myocardium and
most affected subjects die in infancy from cardiac failure. 29
Kawasaki's Disease
This is a systemic febrile vasculitis with multisystem
involvement, whose cause is unknown. Children under the
age of 5 years are predominantly affected. 3l Eighty per cent
of patients are diagnosed before the age of 5 years. 52 The
condition is reported world-wide and, in developed coun­
tries, males are affected more often than females. It is the
most common cause of acqu ired heart disease in children. 53
Kawasaki's disease and infantile polyarteritis nodosum are
now considered to be the same entity. An infective aetiol­
ogy is suggested by the observed clustering of cases.
The incidence in Blitain is estimated to be 3.4 per 100000
chi ld ren aged less than 5 years.63 The case fatality in Britai n
in 1990 was 3.7 per cent. Six deaths were reported in Britain
in 1990; in only one case was the diagnosis made in life.
Clinical features are the presence of fever for about 5 days,
lymphadenopathy, characteristic changes in the peripheries
(erythema of the palms and soles, desquamation of the fin­
gers and toes and pelipheral non-pitting oedema), bilateral
conjunctivitis, buccal inflammation (injected pharynx, dry
cracked lips, strawberry tongue, mouth ulcers) and a poly­
morphous exanthem. 64 Hydrops of the gall-bladder is
increasingly recognized. Three-quarters of affected patients
show a transient rise in the serum concentration of IgE. 62
Sudden clinical onset is usual. There are four major stages in
the development of the disease: 10
• an 'acute febri le' stage which lasts 1-11 days;
• a 'subacute' stage which lasts 11 to 21 days and is
associated with thrombocytosis;
• a 'convalescent' stage from 21 to 60 days; and
• a 'healed ' stage.
Vasculitis is often widespread but the coronary arteries
are always affected 52 (see Fig. ILl, p. 206). The heart is the
most severely affected organ, and coronary arteritis with
aneurysmaJ formation occurs in 20-30 per cent of cases. 53
Aneurysms more commonly affect the left coronary artery.
The prognosis for resolution is favourable unless the
aneurysms are 'giant', as these are more often associated
with myocardial infarction. 3! Recurrences may occur,65 and
coronary aneLllysms/dilata tion and giant aneurysms are
more common in this subset. Other organs often affected are
the kidneys, spleen, testis, pancreas, adrenal glands and
liver. There is correlation between the severity of the autopsy
findings and the duration of the illness. Most deaths occur in
232 I Sudden natural death in infants and children
the subacute and heali n g stages. Sudden unexpected death
may occur many years later, following recanalization of
thrombosed coronary arteries,lo and sudden death has been
reported in a fit athlete during exercise with evidence of pre­
ceding Kawasaki disease. JI
Overlapping features, including the clinical demonstra­
tion of coronary aneurysms, have been reported bet"feen
group A streptococcal infection and Kawasaki's disease. 66
Abnormalities of Cardiac Conduction
While myocardial disease, such as hypertrophic cardio­
myopathy, myocarditis, histiocytoid cardiomyopathy and
tumours, may cause sudden death consequent upon
involvement of the conduction pathway, abnormalities of
the cond uction tract may coexist with certain types of con­
genital cardiac malformations or the conduction tissue may
be damaged during their operative repair. 29 Fibromuscular
hyperplasia of the sinoatrial nodal an d atriovent ricul a r
nodal arte ri es may cause paediatric sudden death from car­
diac arrhythmia. 67
On the other hand, fatal arrhythmias may result from
intrinsic defects of the conduction tissue that are not obvi­
ous to the naked eye. 30 .6B For precise diagnostic categoriza­
tion, such cases require formal examination involving
extensive serial sectioning. In practice, this is time-consuming
and associated with such a low positive yield that it does not
gene rally form part of the routine examinatio n of the heart
in cases of sudden death. Conduction defects may thus be
easily overlooked and many such deaths are simply referred
to as showing no structural cardiac disease. 41
The long QT syndrome (LQTS) is significantly associated
with ventricular fibrillation and the risk of sudden death.
The condition can be inherited as an autosomal dominant
disorder (Romano-Ward syndrome),69 with ab norm ali ties
at six different loci identified.1° A recessively inherited
syndrome w ith deafness is also recognized .7o Vigorous
physical exertion often precedes vent ricula r fibrillation. In
these cases there is no discernible abno rmality on histolog­
ica l examination of the conduction pathway. IS Risk factors
for sudd en death in such indi vidua ls include a history of
syncope, congenital deafness and a fami ly history of sud­
den cardiac deathS
Preceding symptoms th at suggest a conduction defect
include syncope and dizziness. A preceding abnormal elec­
trocardiogram (ECG) recordin g or a family history of con­
duction defect is valuable in these cases, and likely to
ex plain an arrhythmic cardiac arrest. In their absence, the
significance of any histological cha nges of the conduction
tract is merely speCUlative. Opportunity for clinical diagno­
sis in these cases is precluded by the intervention of sudden
death of the patient.
Accessory pathways of th e conduction system may
cause re-excitation and re-entry lea ding to fatal arrhyth­
mias. The Wolff-Parkinson-White pre-excitation syndrome
may occur in isolation or be associated with other abnor­
malities, for example Epstein 's anomaly, rhabdomyoma of
the heart. These abnormal pathways produce early stimula­
tion of the ventricle and are associated with a variety of
electrocardiographic patterns. Many patients remain asymp­
tomatic however, but sudden death in this group is well
recognized.
Heart block may be congenital or result from acquired
causes. Marked degrees of atrioventricular (AV) block may
produce Stokes-Adams attacks and are rarely caused by
endodermal heterotopia of the AV node (mesothelioma of the
AV node). Some congenital abnormalities of the conduction
system (e.g. simple absence of the short segment of the right
bundle branch or the penetrating portion of the main bundle
of His) are stable and do not seem to progress. IS
Babies born to mothers with systemic lupus erythemato­
sus are at considerable risk of congenital AV block but the
prognosis is rem arkab ly good in such cases, and those
affected may not develop syncopal attacks until much later
in life. Death among infan ts with preceding congenital
heart block or those w h o die in association with maternal
connective tissue disease should be tested for anti-La (SS-B),
anti-Ro (SS-A) and even anti - UIRNP antibodies. 29
Heart block has been described w ith Kearns-Sayre
syndrome, Kartagener's syndrome and certain X-linked
myopathies. 10
Occlusive Disease of the Coronary Arteries
Infantile arterial calcification is a rare disease mainly affect­
ing infants,71 with reports of siblings affected. Low levels of
plasma cell membrane glycoprotein-l nucleoside triphos­
phate pyrophosphohydrolase have been demonstrated in an
affected infant. 72 The same group subsequently demon­
strated mutations in the EN??] gene 7J Alieries throughout
the body are involved, with the exception of the brain and
spinal cord. Of the 62 cases reviewed by Moran (1975),71 10
presented as sudden death, preceded by respiratory distress
in 5. The age of death ranged from 2 days to 28 months, with
85 per cent of affected infants dying within 6 months.
Clinical diagnosis is possible with rad iological study. The
ECG changes, when available, are mostly compatible with
myocardial ischaemia. Cardiac enlargement is common and
the myocardium is often infarcted. Intra-uterine ultrasono­
graphic diagnosis has been made in familial cases. 74
Microscopically, there are calc ific deposits with fragmen­
tation of the internal elastic membrane of arteries along wit h
variable degrees of fibrointimal proliferation and luminal
narrowing. 71 ·75 An inflammatolY reaction is typically absent.
An alieriopathy with features similar to those found in idio­
pathic infantile arterial calcification has been reported in
children with acquired immun e deficiency syndrome
(AlDS);76 one of the cases described by these authors showed
myocardial infarction associated with thrombus formation
in a coronary aneurysm.
 Cardiovascular causes of sudden death I 233

Table 12.1 Causes of coronary artery dysplasia in infants

and children

Fibrous muscular dysplasia

Idiopathic arterial calcification
Chronic arsenic poisoning
Congenital rubella
Tuberous sclerosis
Neurofibromatosis
Acquired immunodeficiency syndrome
Homocystinuria
Down's syndrome
Menkes'syndrome

Fibromuscular dysplasia is a segmental, non-atheros­

clerotic vascular disease of unknown aetiology mainly
causing renal vascul ar disease, but other arteries may be
involved. Young adults are usually affected and its occur­
rence is rare in infancy a nd childhood. Myocardial infarc­
tion due to fibromuscul ar dysplasia affecting the coronalY
artery has been reported as a cause of sudden death in
infants and children.77,78
Myocardial infarction with cardiomega ly has been
reported in children with chronic arsenic exposure from
drinking water. 79 This occurred as part of a ge neralized
arterial disease of irregu lar distribution and progressive
course, but apparently sparing the arteri es of the lung and
Figure 12.6 The cut surface of the heart sho ws multiple
brain. Microscopically, there was intimal fibroblastic thick­
tumours of the myocardium alternating with areas of congestio n
ening of the media of small-sized arteries. The endothe­
from an infa nt who died from cardiac rhabdomyomas.
lium, internal elastic lining, media and adve ntitia were
uninvolved.
Widespread vascular dysplasia in a child with tuberous ENDODERMAL HETEROTO PIA OF THE
sclerosis leading to variable luminal nan-ow ing of the coro­ ATRIOVENTRICULAR NOD E
nary, superior mesenteric, renal and common iliac arteries
Previously referred to as mesothelioma of the AV node, the
has been reported;80 these a uthors described a 9-month-old
lesions are frequently associated with heart block and sud­
infant who developed an aOliic aneurysm. Fatal haemor­
den death. Gross examination of the heart may show a small
rhage followed dehiscence of the aortic graft.
raised nodule immediately above the septal leaflet of the tri­
Occlusive artelial disease has also been reported in
cuspid valve, just antelior to the coronary sinus in the right
patents with homocysteinuri a,8 J and followin g congenital
atrium. Microscopically, the lesion comprises multiple cysts,
rubell a syndrome, S2 neurofibromatosis 1,83 Menkes' syn­
gland-like structures or nests of epithelioid cells within a
drome 84 and, rarely, Down's syndrome. 85
fibrous stroma. 86 The tumours may be missed if sections of
A list of conditions that may cause abnorm alities of the
the AV node are not exa mined in cases of sudden death.
corona ry arte ries is presented in Table 12.1.
These lesions have now been convincingly shown to be due
Corrado et al 57 referred to the sudden death of a 14­
to endodermal inclusions. Associated cardiovascular malfor­
year-old g irl whose coronary ostia were nan-owed by
mation may be present. 87 The entity is sometimes refen-ed to
Ta kayasu arteritis.
as the 'smallest tumour causing sudden death'.

Tumours of Cardiac Muscle RHAB DOMYOMA

-------
Primary cardiac tumours are uncommon but are associated
with potentially lethal conditions producing conductive
and haemodynamic abnormaliti es that often result in sud­
den death.
Rhabdomyoma is the most common tumour of the cardiac
muscle. In 30 per cent of cases, death occurs at birth or in the
neonatal period and one-third of cases are associated with
tuberous sclerosis.s7 The tumours may be isolated or multi­
ple (Fig. 12.6), but diffuse infiltrative patterns have been
234 I Sudden natural death in infants and children

Figure 12.8 The left ventricular myocardium from an 8-month­

old boy shows a large well-circumscribed tumour with a solid
trabeculated cut surface typical of cardiac fibroma.

Figure 12.7 The microscopy of cardiac rhabdomyoma shows

'spider' cel ls chara cte ri zed by ce lls with clear to vacuolated
cytoplasm wi th myofibrillar strands extendi ng between nuclear
and cytop lasmic membrane and a central or eccentric nucleus.

described. The cause of death is outflow tract obstmction or

cardiac arrhythmia. Microscopically, large cells with clear
cytoplasm and small nuclei (so-called 'sp ider cells') consti­
tute the microscopic picture (Fi g. 12.7). Most cases of sudden
death due to the tumour are sporadic. B6

FIBROMA
The seco nd most common cardiac tumo ur, fibroma of
the heal1, also ca uses death secondary to outflow tract
obstruction or arrhythmia. A well -circumscribed grey-white
mass betvveen 3 and 10 cm in diameter is_ the usual gross
appearance (Fig. 12.8), cysts may occur. The tumour is most
often located in the interventricular septum. Microscopy
shows fibroblast-like cells in a colla genous matrix, reminis­
cent of fibrom atoses of the soft tissues (Fig. 12.9).B6,87 The
tumour may occur as a component of Gorlin's syndrome.

MYXOMA
Myxomas are rare in infants and children. s7 Most occur as
sporadic lesions, arise withi n the left atrium and may reach
up to 10 cm in diameter. Su dden death is related to dis­
turbed cardi ac haemodynamics or systemic emb olization of
tumour fragments. Myxomas rarely involve the right side
of the heart, where they are capable of causing massive
pulmonary embo lism. Bs
- - "'. ."
Figure 12.9 Histology of the cardiac fibrom a shows fibrobl ast­
like cells in a coll ag enous matrix closely resemb li ng fibrom atosis.
OTHER CARDIAC TUMOURS
Cardiac haeman gioma is a rare cause of sudden death in
child re n. A previously asympto matic 13 -year-old girl who
was found unresponsive in bed was reported by Krous
et a\. B9 Autopsy showed haemangiomatous invo lvement of
 X-linked hypohidrotic (anhidrotic) ectodermal dysplasia I 235

the ventricular conduction system. Death was thought

most likely to have been arrhythmic in origin.
,
I , ,.
Swalwe11 90 repolted a benign teratoma of the interven­ ,
tricular septum causing sudden death in 2.5-year-old girl
. t
who had no preceding medical problems. .' •
Krous et al 51 analysed 68 cases of tumour-associated sud­ «. ,
den deaths in children from the literature. Of those involving \ .I ~ ~ G
\ t
the heart, there were 23 cases of oncocytic cardiomyopathy •

,,
(histiocytoid cardiomyopathy), 12 cases of cardiac fibroma, #

seven cases of rhabdomyoma, three cases of myxoma and \ ~

one case each of neuroma, teratoma and mesothelioma. In

•, I -/
~,
~ II
one case of oncocytic cardiomyopathy, one case of cardiac [
1
fibroma and the case of neuroma the lesion was considered
incidental and unrelated to the cause of death.
\ .
~

j
(
I '
•,
!
f J
\.
fI

­ .."
.1. I • t -

\' ' ,
.!.' .I I -~
Thromboembolism I .
I!

il
~
Pulmonary thromboembolism is an unusual cause of sudden
death in children. Predisposing factors include recent sur­
l ,
gery, congenital heart disease, indwelling venous catheters, , I
,.
J
£ ~ , t ..
sepsis, arteriovenous malformation and occult malignancy.91 .
.,
~

Of 17 500 autopsies reviewed by these authors, eight cases " .- ("' . • ~ ~

(0.05 per cent) showed pulmonary thromboembolism; their
ages ranged from 1 month to 13 years. The source of the
embolus may not always be demonstrated at autopsy.
Unexpected death from massive pulmonary embolism
was reported in three infants.92 The underlying conditions
were necrotizing enterocolitis , a ventriculoatrial shunt for
hydrocephalus and idiopathic arterial calcification. Birth
control medication may predispose to pulmonary embolism
in teenagers. 59 Sudden death from thromboembolism of the
left anterior descending coronary artery complicating acute
rheumatic heart disease in a 6-year-old girl and complete
atrioventricular canal defect in an 18-year-old girl with
Down's syndrome were reported by Stahl et al 4
Thrombosis and pulmonary embolism is a recognized
complication of homocysteinuria (cystathionine B synthase
deficiency).81 Any blood vessel can be affected, even the
intracranial dural sinuses. Microscopically, the arterial
lesion comprises marked fibrous intimal thickening, split­
ting and fraying of the muscle fibres , in the media with
increased interstitial collagen. The internal elastic lamina
may also be affected. Lipid deposition is not a feature.
Potential causes of sudden death in this condition include
pulmonary embolism, myocardial infarction and cere­
brovascular accident.
I I
\
'0
,
,
.'..· t ,
I
I
I " oJ
Figure 12 .10 Aortic wall with cystic medial necrosis. Death was
due to cardiac tamponade secondary to aortic rupture and an
undiagnosed aortic coarctation in a 14-year-old boy.
90 per cent of the mortality is related to cardiovascular com­
plications that include mitral valvular regurgitation or aortic
dilatation, dissection and nIptlire. Aortic dissection with
nIpture and haemorrhage into the pericardial cavity causing
cardiac tamponade is a common mode of death.3.4 1,93
An intimal tear is usually located in the ascending aorta
and histological examination of the aortic wall shows cys­
tic medial necrosis, characterized by fragmentation of the
elastomuscular media and accumulation of basophilic
material. Spontaneous dissection of the coronary a11ery
rarely causes sudden death, 57 but this is not confined to
cases of Marfan's syndrome. Cystic medial necrosis may
occur with aortic coarctation (Fig. 12.10) and in patients
with an isolated bicuspid a011ic valve in the absence of
Marfan's syndrome. 3 Patients with Ehler-Danlos type IV
syndrome are also at increased risk of vascular nIpture. 8

X-LINKED HYPOHIDROTIC (ANHIDROTIC)

Marfan's Syndrome ECTODERMAL DYSPLASIA

Patients with Marfan's syndrome display a characteristic
habitus: tall stature, arachnodactyly, bilateral ectopia lentis,
high arch palate, dolichostenomelia and striae distensae.
Linkage analysis has shown that the fibrillin gene on chro­
mosome 15 is associated with Marfan's syndrome. 8 Up to
The condition is characterized by an absence or diminution of
eccrine sweat glands, eczematoid rash, thin sparse hair,
oligodontia and peg-shaped teeth. The incidence is estimated
to be in 1 per 100000 live births.94 The condition carries a sub­
stantial mortality and morbidity, and death occurs in about
236 I Sudden natural death in infants and children
Figure 12.11 Mid-sagittal view of brain from a 12-month-old
girl with achondropl asi a. There is marked forebrain hydrocephalus
and necrosis of the upper cervical cord due to a small foramen
magnum.
30 per cent of cases. 95 There is an increased risk of chest infec­
tion and atopic disease. Because of the deficient number of
eccrine sweat glands, patients are unable to perspire and con­
sequently develop fatal heat intolerance 94 Mucus production
is also deficient in the respiratory and gastrointestinal tracts.
An inexplicable improvement in the patient's general condi­
tion occurs with increasing age.
Achondroplasia
Achondropl as ia is the most common type of osteochon­
drodysplasia; it is inherited as an autosomal dominant tra it.
Recent studies suggest that sleep apnoea is common among
affected individuals,96 involving both obstructive and cen­
tral mechanisms. Compressive myelopathy and dysfunction
at the cervicomedullary junction is caused by a small fora­
men magnum, and sudden infant and childhood deaths
have been reported. 97 Macroscopic or microsco pic necrosis
of the spinal tissue at the level of the foramen magnum may
be seen (Fi g. 12.11).
Tuberous Sclerosis Complex
Tuberous sclerosis is a dominantly inherited condition
characterized by genetic heterogeneity. The birth incidence
is about 1 in 6000. More than 70 per cent of cases will be
new mutations. Early skin manifestations are hypomelanic
macules and shagreen patch. Subungual fibromas and
facial angiofibromas occur later. Cardi ac rhabdomyomas,
giant-cell astrocytomas and renal disease are other patho­
logical lesions .93 The polycystic renal disease of tuberous
sclerosis usually affects infants and young children, while
the angiomyolipomas tend to occur at or after puberty. At
least 80 per cent of children presenting with cardiac rhab­
domyomas will have tuberous sclerosis. 93 The incidence of
tumours in patients with tuberous sclerosis approaches
50 per cent,99 but their size and number tend to regress with
age. The majority of individuals presenting in childhood
h ave epileptic seizures. 98
The preCise mechanism of death is not always clear. Sud­
den un expected death in infancy has been reported to occur
either from cardiac arrhythmia or blood flow obstruction
secondary to rhabdomyoma. Patients with tuberous sclerosis
may develop Wolff-Parkinson-White syndrome, which may
initiate fatal cardiac arrtbythmia and is probably related to
the presence of the cardiac tumour. 100
Rarely, a renal angiomyolipoma ruptures with life­
threatening retroperitoneal haemorrhage. 98 Aortic aneurysm
secondary to vascular dysplasia in a child w ith tuberous
sclerosis has been reported. so
INTRACRANIAL HAEMORRHAGE, NEOPLASMS
AND MALFORMATIOI\IS
Sudden unexpected death m ay result fro m intracranial
haemorrhage secondary to a generalized bleeding disor­
der lOI or the haemorrhage may be related to intracranial
pathology (Fig. 12. 12). Contrasting with subdural or
extradural haemorrhage, sudden unexpected death follow­
ing parenchymal haemorrhage is more likely to be associ­
ated with natural disease . Arteriovascular malformations
(Fig. 12.13) and aneurysms a re important sources of cata­
strophic bleeding. 2o ,JOI Intra-crani al a neurysms are found
with uncorrected aortic coarctation, familial multiorgan
cystic disease or autosomal dominant polycystic kidney
disease (ADPKD).
While fatal subarachnoid haemorrhage in ADPKD usually
presents in adulthood,I02 infants and children are rarely
affected. A family history is extremely helpful in making a
diagnosis. The cerebral anelllysm may be masked or destroyed
by the haemorrhage and, consequently, may be difficult to
find at post-mortem examination. Intracranial aneurysms in
childhood have been reported to occur as a complication of
renal hypertension . 103
The intracranial pathology mayo r may not be clinically
apparent prior to sudden death.l04 These authors reported
10 deaths from intracrania l haemorrhage, secondary to
tumours (four cases) , berry a neurysm (one case) and vascu­
lar malformations (five cases); the presence of four vascu­
lar malformations was assumed at autopsy as their positive
identification was difficult owing to the marked tissue
destruction. Such haemorrhages may be due to micro­
angiomas. lOl An additional cause of intracranial bleed in
older female patients is eclampsia.
Four cases of intracerebral haemorrhage reported by Liv­
ingston and Brown 101 compbcated a pre-existent coagulopa­
thy. In two cases this was thrombocytopenia (associated with
aplastic anaemia and acute lymphoblastic leukaemia) ; one
child had haemophilia and one case was thought to be a late
manifestation of haemorrhagic disease of the newborn.
Minor trauma was a factor in the case of the haemophilia.

Figure 12.12 Cut surface of the cerebellum shows haemorrhage
into a tumour which wa s subsequently demonstrated to be low­
grade astrocytoma. The patient was a 4-year-old girl who had
apparently been in good health and had died suddenly.
o adjacent to a tumour l04 obstructive hydrocephalus lO8. 109 and
Figure 12.13 Microscopic view of an arteriove nous malformation
which caused catastrophic intracerebral haemorrhage in a lO-year­
old girl. There is vascular prominence with irregular thickening of
some of the vessels. Elastic van Gei son stain (EVG).
The clinical course of idiopathic thrombocytopenic purpura
is rarely complicated by intracranial haemorrhage. The inci­
dence is estimated to be 1 in 1000 patients. lOS
Sudden death from undiagnosed primary intracranial
tumours per se rarely occurs. Abu al Ragheb et al (1986)106
found seven cases of sudden death secondary to intracranial
Intracranial haemorrhage, neoplasms and malformations I 2 37
J8,...
Figure 12.14 The tip of the intraventricular part of a
ventriculoperitoneal shunt that had become ensnared in choroid
plexus. causing obstruction to cerebrospina l fluid egress in a
young child with shunt- treated hydrocephalus.
neoplasm in 1055 autopsies. including one in a 6-year-olcl
and one in an 18-year-old. Preceding symptoms mayor may
not be present and, when present, relate mainly to increased
intracranial pressure. epilepsy, focal neurological deficit or
psychiauic manifestation. l07 Loss of consciousness may pre­
cede sudden death. Sudden mass effect, haemorrhage into or
epiIepsyilo are some of the mechanisms responsible for death.
An unusual case of spontaneous sub a rachnoid haemor­
rhage in a 5-month-old girl who died suddenly was
reported by Byard et al; III collagen analysis showed virtual
absence of type III collagen, diagnostic of type IV
Ehlers-Danlos syndrome.
The Dandy-Walker malformat ion is a well-recognized
but a rare cause of sudden death .112 Autopsy examination
may provide the first indication of the presence of the mal­
formation. The mechanism of death is unclear, but brain­
stem ischaemia from local pressure abnormalities in the
unshunted posterior fossa may be responsible.
Sudden death occasionally occurs in children with a
colloid cyst of the third ventricle. I08 Owing to its critical
location in the anterior part of the third ventricle, the
lesion may obstruct cerebrospinal fluid drainage and lead
to acute hydrocephalus. Histologically, these are unilocular
cysts conta ining eosinophilic debris and lined by
cuboidal/columnar epithelium, which may be ciliated.
Hydrocephalus may be congenital with and without
spina bifida. Acquired hydrocephalus may follow or
accompany intraventricular haemorrhage, meningitis or
neoplasms. Patients may present with respiratory arrest. 113
Focal and generalized convulsions are common with
shunted hydrocephalus and may result in status epilepti­
CUS.
114 Factors precipitating convulsions include infection
and shunt-related complications (Fig. 12.14).
238 I Sudden natural death in infants and children
Encephalitis and Leucodystrophies
Sudden deaths in infancy and childhood are rarely ascribed
to encephalitis. The clinical course of herpetic encephalitis in
chiJdren or young adults may be rampant, and in the neo­
natal period severe general ized disease caused by herpes
simplex virus infection can also be rapidly fatal. Cerebral
involvement as part of a severe disseminated infection in the
neonate may a lso occur with coxsackievirus infection, a nd
there is predilectio n for involvement of the brainstem and
spinal cord. liS Sudden death at home in two siblings associ­
ated with an encephalitic process has been reported by
Howat et al. I1 6 An aetiological agent was not discovered.
Shields et a l l1 7 reported sudden death in two female
patients secondary to the autosomal recessive form of
adrenoleucodystrophy and to Alexander's disease.
Adrenal Hypoplasia/Insufficiency
Adrenal hypoplasia in infancy may be secondary to brain
malformation or occur in isolation. liS Weakness, weight
loss, hyperpigmentation , hypotension, gastrointestinal
symptoms and vitiligo are the main symptoms of chronic
adrenocortical insufficiency. 119 By contrast, the symptoms
and signs of acute adrenal failure are non-specific and are
those found in any serious illness. Adrenocortical insuffi­
ciency presenting later in chi ldhood is more likely to be of
autoimmune origin (Addison 's disease) .11 9.120
Favara et al 121 reported severa l deaths in infants, with an
age range from a few hours to 24 months, with undiagnosed
adrenal hypoplasia without central nervous system malfor­
mation. There was a strong male predominance. Sudden
death was recorded in four of the infants. The morphology of
the adrenal glands was variable and all were normally
shaped but miniature in size. In a ll cases the combined adre­
nal weight was less than 2 g. One patient died during induc­
tio n of anaesthesia. Two infants were born small-for-dates.
Pregnancy-induced maternal hypertension was present
in just over 50 per cent of mothers whose infants had con­
genital adrenal hypoplasia. 122 The condition was unsus­
pected in life, and unexpected clinjca l deterioration
occurred in three cases.
An 'adrenal crisis' tends to occur in patients with previ­
ously diagnosed adrenocortical insufficiency and in whom,
during intercurrent infection, corticosteroid replacement
therapy is not increased. However, Molander lJ referred to a
sudden death during exertion from adrenal insufficiency in
a previously asymptomatic 19-year-old male. Fibrosis and
calcification of the adrenal glands were found at autopsy.
Al Sabri et al 120 reported the sudden death of a 12-year-old
gi rl with autoimmune Addison's disease. Biochemical
abnormalities included hyponatraemia, hyperkalaemia,
hypercalcaemia , haemoconcentration, hypoglycaemia and
metabolic acidosis. Autopsy revealed depletion and atro­
phy of the adrenal cortex.
Haemorrhagic Shock Encephalopathy
Syndrome
Haemorrhagic shock encepha lopathy syndrome is a rare and
devastating disorder, with a sudden-onset symptom com­
plex, usually in previously healthy infants and chil­
dren. 123 ,124 Symptoms include fever, shock, haemorrhage,
diarrhoea and encephalopathy (coma, seizures). Laboratory
investigation during hospitalization shows progressive dete­
rioration of renal function, falling haemoglobin and platelet
counts, evidence of disseminated intravascular coagulation,
hyper-natraemia, hypoglycaemia, metabolic acidosis, raised
serum transaminases and hyperammonaemia. 123 ,124
At autopsy the brain shows oedema, softening and infarc­
tion.1 24 Hepatic steatosis, hepatic necrosis and small intes­
t ina l vi llous blunting are other common findings, as are
bleeding from the nasotracheal tube and intravenous access
sites. Microbiological cultures and toxicological screening are
uniformly negative. The aetiology is unknown and the differ­
ential diagnosis includes heat stroke due to overwrapping,
toxic shock syndrome, septic shock, haemolytic uraemic syn­
drome, Reye's syndrome, metabolic disorders, viral haemor­
rhagic fevers and poisoning. 124
The age disttibution of the condition is between ) 7 days
and) 5 years. The modal age is 3 months, with more than 80
per cent of the patients presenting before 1 year of age. 12S
Boys are more commonly affected than girls. There may be a
family history of a neurologica l disorder or unexpected death
in infancy. The prodromal illness varies from a few hours to
several days. Of 33 patients with follow-up information, 22
died, and many of the survivors were neurologically dam­
aged.125 In my practice the incidence of this condition has
declined considerably over the past several years.
Reye's Syndrome
This is a rare and serious disorder chiefly affecting chi ldren,
characterized by a non-inflammatOlY encephalopathy with
hepatic dysfunction in which the diagnosis is made on com ­
bined clinical, biochemical and histological parameters. The
internationally accepted case definition is age under 16
years; unexplained non-inflammatory encephalopathy with
one or more of t he following: serum hepatic transaminases
raised more than three times the upper limit of normal,
plasma ammonia level raised to more than three times the
upper limit of normal and characteristic fatty infiltration of
the liver. 126 In Britain, the epidemiology ofReye's syndrome
differs from that in the USA in so far as the mean age is 14
months compared with) 1 years, and there is no clear asso­
ciation with influenza. 127
The onset of the illness is marked clinically by profuse
effortless vomiting, progressing to extreme lethargy and
coma. A coagu lopathy and hypog lycaemia are common.
Jaundice is rare. Poor outcome is associated with early
onset of seizures, profound hypo glycaemia and coma.
 Fatal anaphylaxis I 239

At autopsy the brain is oedematous and the liver

appears pale due to panlobular fa tty change . These light
microscopic findings, however, lack specificity and may
be seen in many other conditions. '28 Electron microsco py
is widely recommended to confirm the di agnos is, I29 and
swo llen and pleomorphic liver cell mitochondria are the
ultrastructural hallmarks. 126
The decline in the number of cases of classic Reye's syn ­
drome has been partially attributed to warnings against
aspirin exposure durin g viral prodromes in children under
12 years of age and the increased diagnostic awareness of
genetic metabolic diseases that mimic the syndrome. ' 27 ,130
Consequently, Reye's syndrome has become an entity
comprising a number of 'Reye-like' inherited metabolic
disorders,I3 1 includin g p-oxid ation defects, organic acid
disorders and urea cycle defec ts. '30 ,'J2 Classical Reye's syn­
drome has thus become a diagnos is of exclusion, l26
Medium-chain acyl CoA dehydrogenase deficiency is the
most common metabolic disorder masquerading as Reye's
syndrome. Previous unexplained sibling deaths or similar
illness, a previous history of ge netic metabolic disease or
unexplained illnesses (e.g. hypoglycaemia, fits) should
prompt appropriate investigation. Haemorrhagic shock and
encephalopathy syndrome and mitochondrial disease should
also be considered in the differential diagnosis,
Figure 12.15 Extensive haemorrhagic infarction of the small
bowel secondary to volvulus which caused the sudden collapse
GASTROINTESTINAL CAUSES and death of a 3-year-old girl.

Delayed presentation of congenital diaphragmatic hernia

has been reported as a cau se of sudden death in two male
infants and a 2-year-old girl. 133 The mechanisms of death
were mediastinal compression and cardiorespiratory col­
lapse. One of the infants died in his sleep and was thought
to have succumbed from sudden infant death syndrome.
The small size of the diaphragmatic defect in these cases
probably contributed to th e delayed onset of problems.
Small bowel incarceration and infarction as a result of
mesodi vertic ular band are reported to have caused sudden
dea th in i nfan ts aged 30 and 31 months. lJ4
We have previously encountered a 3-year-old girl
whose sudden death was caused by to rs ion of the intestin e
in the absence of malrotation (Fig, 12.15).
FATAL ANAPHYLAXIS
Acute ana phylaxis is an immediate type IgE antibody­
mediated hypersensitivity reaction. The clinical spectrum
includes hypo tensio n, bronchospasm, angio-oedema and
laryngop haryngeal oedem a. IJ5 ,l36 The full-blown clinica l
picture may be preceded in some patients by less fulminant
symptoms such as itching of the skin, flushin g, generalized
warm th or evidence of smooth muscl e contraction, l37
Fatal or near-fatal ana phylactic reactions occur in
children and ado lescents,l38 and even infants may be
affected. IJ5 The subjects are often highly atopic, with asthma ,
allergic rhinitis and atopic dermatitis. Apart from various
foods, recognized allergens include drugs, pollens and venom
from stinging insects. IJ7 " 39 The offending allergen can be
identified in most instances, but occasionally is not found.
Mast cell tryptase measurement has been recommended
as a useful test for anaphylaxis and has served as a diagnos­
tic marker at autopsy.lJ5 "'Vbilst tryptase remains stable in
samples up to 4 days at room temperature, freezing of sam­
ples at - 20°C is recommended if the assay cannot be carried
out immediately. J40 Elevated levels are not absolutely spe­
cific for anaphylaxis and may be found in post-mortem
serum in other conditions such as sudden infant death syn­
drome and trauma. 140 Total serum IgE and allergen-specific
IgE antib odies may also be usefully measured. l4l
The period between onset of ac ute anaphylaxis and
death varies considerably, but without medica l interven­
tion most patients expire within a few hours. 137
An allergic reaction may be conSiderably enhanced in
patients receiving p-blockers. l42 Patients with cow's milk
protein intolera nce who are challenged with milk following
a period of cow's milk protein avoidance are es pecially at
risk of developing acute anaphylaxis,
Autopsy findings in ac ute anaphylactic deaths are non­
specific, for example pulmonary congestion and oedema
with variable intra-alveolar haemorrhage, Changes may
240 I Sudden natural death in infants and children
be enhanced by efforts at resuscitation . 1J7 Other findings
include increased tracheobronchi al secretions, laryngeal
oedema and pulmonary emphysema. An autopsy diagnosis
of acute anaphylaxis cannot be made on morphological
grounds alone and appropriate clinical information is
esse ntial. IJ6 In cases of fatal reaction to insect stings, the
site of the sting may be found.
SICKLE CELL DISEASE
Sickle cell disease is more common in the black population
and the clinical course is generally more severe in homozy­
gous than in heterozygo us individuals. Most deaths occur
in patients with haemoglobin SS, but young subjects with
haemoglobin SC are also at risk.143 Polymerization of
deoxygenated sickle haemoglobin produces sickling of red
blood cells leading to veno-occlusive crises. Predisposing
factors include dehydratio n, fever, acidosis and hypox­
aem ia. Sudden death in infants and older children has been
repolted . 144 Generalized convulsions are also recognized.
Sudden death in infants with a sickle cell trait may be asso­
ciated with anaesthesia.
Infarction of the small bones of the hands and feet,
sickle cell dactylitis, aplastic and haemolytic clises and
splenic sequestration are recognized complications. Splenic
sequestratio n crisis, the most severe complication, causes
death from circulatory collapse due to marked pooling of
blood within the spleen. Th is is particularly common in
infants under 2 years of age. 144 Acute infection is a com­
mon precipitating factor. At autopsy the organs are pale
except for the spleen, which is enlarged and engorged.
Sickled cells in capillaries may be visualized at microscopy,
but sickling as such does not necessarily imply that it has
occurred ante-mortem. 144 Purulent meningitis, cerebral
infarction, cardiomegaly, splenic infarction and bone mar­
row hyperplasia are other auto psy findings.
In the USA the peak incidence of death was found to be
between one and three years and the major precipitating
event was infection. 145 The causative agent most commonly
isolated was Streptococcus pneumoniae. Pneumococcal septi­
caemia in childhood is observed less frequently because of
the widespread use of immunization and antibiotics. 143 The
major cause of death in the second decade is a cerebrovascu­
lar accident. Other causes of death include acute chest syn­
drome secondary to pUlmonary vasc ular occlusive disease,J46
a condition that tends to affect older subjects and may lead to
chronic lung disease and pulmonary hypertension.
HAEMORRHAGE AS A CAUSE OF
SUDDEN DEATH
Haemorrhage produces death through mass effect and dis­
tOltion of vital structures (intracra niai), asphyxia (lung) or
exsanguination (gastrointestinal or intraperitoneal).
Sources of intracranial ha emorrhage have been dis­
cussed earlier. One study l5 found 13 out of 169 sudden
deaths to be primarily due to haemorrhage; these were
mainly intra cra nial haemorrhage and usually secondary to
an arteriovenous malformation. One death followed rup­
ture of a tubal pregnancy.
The aetiology of pulmonary haemorrhage has been out­
lined by Cutz l47 and, broadly, may be separated into pri­
mary idiopathic pulmonary haemosiderosis and second ary
pulmonary ha emorrhage. The latter includes immunolog i­
cal causes and vascular, infectious and bleeding disorders.
Idiopathic pulmon ary haemosiderosis has been recognized
as a cause of sudden death.lo
Gastrointestinal sources of haemorrhage include oesophageal
varices, peptic ulceration and vascular malformations.
RESPIRATORY CAUSES OF SUDDEN DEATH
Abnormalities of the Trachea
Tracheomalacia may be congenital or acquired. The con­
genital type is exceedingly rare,148 and cases have been
reported in association with Larsen's syndrome, pulmonary
vascu lar sling, bronchopulmonary dysplasia and tracheo­
oesophageal atresia. Tracheal collapse with life-threatening
airflow obstruction is a recognized complication following
repair of tracheo-oesophageal fistula. 149
Acute Epiglottitis
Acute ep iglottitis, an acute life-threatenin g condition, is
one of the most serious manifestations of infection with
Haemophilus injluenzae type b. The clinical diagnostic cri­
teria are a red swollen epiglottis (Fig. 12.16), inspiratory
stridor or difficulties with swallowing and pyrexia. 150 Chil­
dren are most commonly affected, but the case fatality is
lo w in all age groupS.1 50, 151 Trollfors et ali SO reported six
childhood deaths in their series of 485 children. Four died
at home or were dead on arrival in hospital. The other two
arrived in hospital deeply comatosed, With current immu­
ni zation schedules, the incidence of acute epiglottis in the
Western world may be expected to decrease sharp ly.
Retropharyngeal Abscess
Retrophary ngeal abscess is most commonly reported in
children less than J years of age. Clinical features include
fever, neck swelling, stridor and pharyngeal swelling. In
one series about half of the patients were less t han 12
months old, one-third were less than J months of age and
three patients presented in the neonatal period. Preceding
upper respiratory infection was present in 45 per cent of
 Respiratory causes of sudden death I 241

Figure 12.16 Marked reddening and swelling of the epiglottis

Figure 12.17 A large amount of acute inflammatory exudate
in a child whose death was due to acute epiglottitis.
and necrotic slough are seen in the lumen of the trachea from a
child who died from bacterial tracheitis.
patients. 152 Clinical diagnosis can be difficult in infants
and young children as the onset of infection may be insid­
cases. Parainfluenza virus (I, 2 and J) is the most common
ious with few signs and symptoms. 153
preceding viral pathogen. 155
The condition causes acute upper airways obstruction .
Clinical presentation is characterized by fever, barking
Stridor is unlikely in children over J years of age. Two
. 152 Inone 0 f cough, hoarsness, shortness of breath, cyanosis, stridor and
deaths were reported among t he J 1 patients.
respiratory distress. 156The epiglottis is noted to be normal but
these (a J-month-old baby) the diagnosis was made at
the trachea contains copious purulent secretions (Fig. 12.17).
autopsy. Another mechanism of death in patients with
Complications included pneumonia, septic shock, adult-type
retropharyngeal abscess is rupture of the abscess. leading . . 156
respiratory distress syndrome and tOXIC shock syndrome.
to pneumonia and haemorrhage.
Rarely a pseudomembrane extends to involve the whole tra­
A useful diagnostic method in life and at autopsy is a ,
cheobronchial tree, and the oesophagus an d stomac I1. 157
lateral radiograph of the neck to demonstrate the increased . k . 158
. II y at ns
Children with Down's syndrome are especla
depth of the retropharyngeal space. Other important radio­
In a review of 110 patients,155 the mean age was 54
logical features are a visible fluid level in the abscess cav­
months, the majority being less than J years of age. Males
ity or gas in soft tissues.
were more often affected than females . More than 80 per
The bacteriology of retropharyngeal abscess has been
cent of patients required endotracheal intubation and res­
weLl documented and includes mixed aerobic and anaero­
piratory support. Cardiopulmonary arrest occurred in IJ
bic infections;153.154 others l52 have reported a mixture of
pat ients, four of whom died.
Gram-negative bacilli and anaerob es, although Staphylo­
Corynebacterium diphtheriae rarely causes tracheitis in
coccus aureus was the most common single organism iso­
the absence of marked supraglottic involvement.
lated. Three children had pure isolates of Klebsiella.

Bacterial Tracheitis Acute Bronchiolitis

Bacterial infection of the trachea is usually superimposed Bronchiolitis is an infection of the lower respiratory tract
on preceding viral infection. Staphylococcus aureus and caused by one of a number of different viruses. It is prima­
Haemophilus influenzae are responsible for the majority of rilya clinical diagnosis 159 and the clinical picture is typified
242 I Sudden natural death in infants and children
Figure 12.18 Microscopic view showing intraluminal inflammatory
exudate and peribronchiolar inflammation in an infant who died from
acute bronchiolitis due to respiratory syncytial virus infection.
by acute wheezing following an upper respiratory illness.
Hyperinflation is the most frequently observed radiological
abnormality.
Respiratory syncytial virus (RSV) is by far the most com­
monly isolated pathogen. Less common agents include
parainfluenza viruses types I and 3, adenovirus, rhinovirus,
and Mycoplasma pneumoniae. RSV is a major cause of the
disease in infants, whereas the other agents tend to affect
older children. '59
The mortality from RSV bronchiolitis in infants who are
otherwise healthy is less than one per cent. 160 Some infants
are at greater risk of severe or fatal RSV infection, and these
include premature infants l61 and those with congenital heart
disease, 162 bronchopulmonary dysplasia, 160.1 63 immune defi­
ciencyl 64 and pulmonary hypertension. 160 Life-threatening
complications of RSV bronchiolitis include supraventricular
tachycardia and pneumothorax. 159
RSV replicates in epithelial cells. 159 Epithelial necrosis
with cilial destruction is the earliest microscopic change
and, subsequently, an inflammatory infiltrate composed of
lymphocytes, plasma cells and macrophages invades the
peribronchial spaces (Fig. 12.18). Airway swelling, slough­
ing of necrotic debris, loss of cilia and increased mucus pro­
duction predispose to luminal obstruction. Bronchioles with
a diameter ranging from 300/.Lm to 7 5 ~.m are affected, 160
and become progressively narrowed and obstructed by sub­
mucosal oedema and mucus plugs. Air trapping and lobular
collapse ensue. 160
The presence of RSV can be determined by viral culture
or commercially available methods such as enzyme-linked
immunosorbent assay (ELISA), immunofluorescent anti­
body and enzyme immunoassay tests. The immunofluores­
cent technique is more sensitive and specific than viral
cuI ture. 159
Pulmonary Veno-Occlusive Disease
This is a rare and usually fatal condition. 165 The aetiology
is unknown. The clinical picture is inconsistent, but
patients usually present with progressive or exertional dys­
pneoa. Sudden infant death has been reported. 166 Histolog­
ically, the pulmonary veins show total or partial occlusion
by intimal fibrosis with or without recanalization. Fresh
thrombi may be superimposed. The veins are not uniformly
involved and some may be normal. 165 Secondary changes
may be observed in pulmonary arteries.
Pulmonary Arterial Hypertension
Severe primary or secondary pUlmonary vascular obstruc­
tive disease is an important cause of sudden death. I ,59
Eisenmenger's pathophysiology is secondary to a large
underlying shunt, usually a ventricular septal defect,
reversed patent ductus, atrioventricular canal or secundum­
type atrial septal defect. The age at death spans the first
decade of life, but sudden death resulting from Eisen­
menger's syndrome tends to occur in older children and
adolescents. A large number of primary pulmonary dis­
orders may also lead to pulmonary arterial hypertension .29
EPILEPSY AND SUDDEN DEATH
A shorter lifespan among epileptics compared with the gen­
eral population is well recognized. 167,1 68 Status epilepticus,
which is a life-threatening condition, is reported to occur in
some 16-24 per cent of children with epilepsy. 169 The risk of
death in status epilepticus is directly related to the speed
with which the seizure is controlled. 17o Sudden unexpected
deaths in epilepsy, in the absence of continuous fitting, take
longer to recognize. 170 Such deaths are referred to as sudden
unexpected death due to epilepsy (SUDEP) and may be
related to the underlying disorder or to the effect of earlier
seizures.
The majority of these deaths are unwitnessed, 171 and victims
are often found in bed. 20,1 67,170 Most deaths are seizure related
and no gross or microscopic pathology at autopsy are appar­
ent, pulmonary oedema and organ congestion usually being
the only findings. The underlying pathophysiology of these

deaths is uncertain, but proposed mechanisms involve auto­
nomically related cardiac arrthymias l67 and ictal apnoea .J72
Poor seizure control and poor compliance with
an tiepileptic therapy do not ap pear to be significant factors
in SUDEP. 173 A toxico logy screen should fonn pa rt of the
au topsy proto col and the heart should be examined in detai l.
A serum dmg level lower tha n the therapeutic ran ge do es
not necessari ly imply fa ilure of compliance, and sh ould not
be in terpre ted by the pathol ogist as a factor in the cause of
death. 174 Moreover, the relevance of post-mortem blood lev­
els of anticonvulsants many hours after death is unclear, as
so me degrade more rapidly than others.1 6S The possib ility of
adverse cardiovascular effects of carbamazipine in SUDEP in
children cannot be ignored. 175
When evidence su ggestive of a seizu re is found, for
example a bitten tongue, limb injuries or an empty bladder,
this sh ould be record ed. However, should this evidence be
absent, a seizure canno t be excluded. It is recommended
that when children are fou n d wi th thei r head immersed in
water, de ath should be recorded as ep ilepsy- rel ated death,
rather than SUDEP.
Card iac arrhythmias rarely give ris e to 'epilepsy',1 76. 177
and death in such cases may be erron eously attributed . A
prolonged QT interval may be responsible and au topsy
examination is unlikely to r eveal any abnorm ali ty.
DEATHS FROM ACUTE ASTHMA
Asthma is a common disease and, des pite advances in tbe
treatment, there is still a hig b ra te of mortality and morbid­
ity amon g children. Deaths from as th ma are usually as a
result of prolonged attacks tha t fail to respond to conve n­
tional measures, status asthmaticus. Sudden unexp ected
death is a lso recognized but occurs less freq uen tly.17B
Zach an d Karner l79 reported sudd en deaths in two gi rls,
one aged 14 years and one aged 9 years, receiving anti­
asthmatic medication. In both cases the gi rls ' condition
was regarded as moderately severe and well stab ilized.
These deaths might be explained by inappropriate percep­
tion of the severity of the attack by the p atie nt.
Of 30 childhood asthma-related deaths reviewed by Car­
swell,l so the mean age at death was 8.5 years; 19 died at
home or in transit to hospital and 11 died in hospita l. One
child died within one hour of the sta rt of the attack; three
died less than 12 hours after the start and seven died after
more than 12 hours. Deaths occurred mainly at night or in
th e evening. Only 12 patients in this stud y we re known to
be receiving prophylaxis.
Unexpected death in asthma patients is especia ll y pro­
nounced in adolescent an d preadol escent years, and may
be related to treatment non-use/abuse, which is ren owned
to occur at this age. lSI The increased frequ en cy of deaths
from asthma at night or in the early morning has been
lin ked to diurnal va ria tion in airflow limitation. ls2 Risk
factors for sudden death include the occurrence of previous
Diabetes mellitus I 243
life- threatenin g asthmatic episodes, hospital admissions
for as thm a in the year preceding death, poor access to
medical ca re, inadequate medica l ma nagement a nd psy­
chological and psychosocial problems. IS} Previous stero id
treatment, by suppressing the ad ren a l axis, m ay lead to
adren al insufficiency and a pred isposition to unexp ected
de ath.
Recognized mechanisms of death are severe asphyxia
du e to airflow restriction IS4 or cardiac arrhythmia from
myocardial irri ta bility secondalY to excessive use of ,B2
an tag onists. Hypokalaemia, which may be partly expl ained
by ~ 2 antagonists, can cause cardia c a rrhythmia or gene r­
ali zed muscle weakness, co ntributing to sudden dea th. 135
Life-threa tenin g eve nts are includ ed among the adverse
effects of aminophylline. IS)
The hi gh mortali ty from asthma in t he 1960s was attrib­
uted to the excessive use of press urized beta ago nist
aerosols, the so-called 'aeroso l hy pothesis'. ls6 Whilst the
debate concerning the role of th ese aerosols con tinues, it is
difficult to discount fa ilure to promptly and adequatel y
treat the asthmatic attack in su ch cases. It has been ques­
ti oned whether lun g fun ction in such cases is ever normal
shortly befor e the fatal attack JB2
Apart from abrupt airway narrowing and medication­
related deaths, an imp ort ant factor may be u nsuspected
pulmonary pathology. lSI Among 13 unexpected asthma
deaths in subjects aged between 9 and 19 years, Kravis and
Kolski lSI id entified a cause in five cases: pneumothoraces,
histiocytic bronchopneumonia, iso lati on of Kl ebsiella
pneumoniae, eosinophilic pn eumonia and foc al bronchio­
litis obliterans.
Macroscopic fi n din gs in acute deaths from asthma are
bulky, hyperinflated lungs, surgical emph ysema and, rarely,
pneumothorax. Microsco py shows intraluminal bronchial
and brochiol a r mucinous plu gs, epithelial basement mem­
brane th ickening, peribronchial smooth muscle hypertro­
phy, mucus gland hyperpl as ia a nd submucosal cellular
infiltration, often with large numbers of eosinophils. 17s
Asthmatics are also at increased risk of developing severe
or fa tal anap hyl ax is,14 1.142 and this possibility sh ould be
considered in the event of sudden death in a patient who
suffers from asthma. Thi s is especially relevan t as acu te ana­
phyla xis may present cl inically as respiratory distress. Mea­
surement of the semm tryptase level may assist in making
the distinction.
DIABETES MELLITUS
Despi te modern treatment, in sulin-depe ndent diabetes mel­
litus (IDDM) in childhood still carries Significant morta l­
ity.IB7·IBB Most deaths are attributable to metabolic
complications, and those related to diabetic ke toa cidosis
a re most common, foll owed by hypog lycaemia. IB9 Sudden
unexpected death is reported in cases of di abeti c keto ac i­
dosis,190 and the a dol escent age group is mostly affec ted .
244 I Sudden natural death in infants and children
Cerebral oedema is an important comp li cation in such cases
and carries a poor prognosis. 191 After death , the blood glu ­
cose is raised and the urine contains ketones, but ketones
may be absent in the aketotic form of diabetic coma.
Post-mortem determination of blood glucose is unreli­
able. 190 The presence of glucose in the urine may also be
misleading as glucose may have been administered intra­
venously. The site of blood sampling at autopsy has an
important influence on the blood glucose level. 192 Thus,
samples taken from the right side of the heart often show
sp uriously high gl ucose levels as a result of glycogenolysis ,
espec ially if death is accompanied by an 'alarm' reaction or
cardiac massage is performed . Sampling from the periph­
eral vein, by contrast, often shows a glucose level that is
spuriously low due to continuing glycolysis after death.
Glucose in the vitreous can be more usefully meas­
ured, 190.192 but the level decreases rapidly after death. So
lon g as the sample has been taken peri mortem into a fluo­
ride tube, this w ill reflect blood glucose level s at that time
and a normal glucose level should exclude hypoglycaemia
providing, of course, that the patient has not received glu­
cose infusion prior to admission. 193
High levels of vitreous or peripheral venous glucose
probably do indicate uncontrolled diabetes, especially if
concentrations of ketone bodies in the blood or urine are
also raised. Measurement of glycated haemoglobin has
been suggested as a more reliable indicator. 194 Absence of
insulin in ante-mortem blood carries more weight than in
a post-mortem sample; by contrast, raised insulin values
in post-mortem blood are useful in excluding diabetic
ketoac idosis. 190
Histology of the pancreas may provide useful informa­
tion providing the post-mortem interval is short and
a utolysis is minimal. 195 Changes include acinar atrophy,
reduction in the number and size of Langerhans islets,
insulitis, distortion of islet architecture and histochemical
demonstration of a reduced number of insulin-producing B
cells. 195 Glycogen may be increased in the renal tubular
epithelium (Armani-Epstein nephropathy), but this is not a
constant feature. Othe r histo log ical findings are microvesic­
ular fatty change of the liver and vacuolization of hepatic
nuclei; these are, however, non-specific and may be found
in other conditions.
The 'dead in bed syndrome' constitutes a puzzling group
among patients with IDDM. The subjects are without clini­
cal evidence of late complications. They are usually on
insulin and there is often a history of one or more noctur­
nal hypoglycaemic attacks in the previo us 6 months. 196
They are generally observed to be in good health on the
preceding day and are found undisturbed in bed the fol­
lowing morning. 196 ,197 Children and young adults account
for some six per cent of all such deaths under the age of 40
years. The most likely cause is thou ght to be hypogly­
caemia with associated events such as ca rdi ac a rrhyth­
mia 198 or respiratory depression. 196 Hypoglycaemia is
difficult to confirm and, as the post-mortem often does not
reveal a cause, the diagnosis is usually inferred from t he
clinical history.
GENETIC IVIETABOLIC DISORDERS
Factors that should prompt consideration of genetiC meta­
bolic disease in cases of sudden death are parental consan ­
guinity, maternal HELLP (haemolysis, elevated liver
enzymes, low platelets) syndrome, previo us Reye-like ill­
ness or infant deaths in the family (from known genetic
metabolic disease or unex plained), a previous acute life­
threatening even t (near-miss cot death), fasting or recur­
rent hypo g lycae mia, neonatal hypotonia, dysmorphism,
en la rgement of the live r and/or spleen. A pale (fatty) liver
or hepatic fibrosis, cardiomegaly and severe brain oedema
are important markers of metabolic disease at autopsy. The
possible metabolic causes of sudden death in the paediatric
group are wide-ranging. 199
Mitochondrial Abnormalities
These comprise defects of fatty acid oxidation (FAO) and
oxidative phosphorylation (OXPHOS). A Reye-like illness or
sudden death is exceptionally reported in association with
defects of OXPHOS. In defects of OXPHOS abnormalities are
likely to involve a single organ or several systems simulta­
neously. A diagrammatic scheme of intramitrochondrial
~ -oxidation of fatty acids an d its interrelationship with the
respiratory chain complex is shown in Fig. 12.19. 200
FAm ACID OXIDATI ON DEFECTS
Enzymes that have been associated with sudden death are
shown in Table 12.2. All of these inborn errors are inherited
as an autosomal recessive trait. Very long-chain acyl coen­
zyme A dehydroge nase (VLCAD) deficiency usually presents
in early infancy with hypoglycaemia, recurrent vomiting,
liver dysfunction, cardiomegaly and a tendency to cardiac
arrest.200.201 Presentation as sudden infant death is well rec­
ognized.202.203 Compared with the other defects, VLCAD is
more likely to be associated with cardiomegaly, although
sudden death and severe illness in the neonatal period have
been reported. 204 Only a few patients with short-chain acyl
CoA (SCAD) deficiency are recorded, 202 and its association
with sudden death is unclear. 203
Primary carnitine defici ency and carnitine palmitoyl­
transferase type II (CPT II) deficiency often present with
cardiomyopathy, whereas carnitine palmitoyltransferase
type I (CPT I) deficiency usually presents with a Reye-like
illness, but occasionally cardiac arrhythmia occurs in the
neonatal period. 205 Fat oxidation is impaired as the trans­
fer of long-chain fatty acids across the mitochondrial
membrane is dependent on carnitine transporter and carni­
tine palmitoyltransferase. 202
 Genetic metabolic disorders I 245

Long-chain fatty acid

1
acyl-CoA
Ligase

Carnitine palmitavl
transferases I and /I

1 Carnitine aeVI-earnitine translaease

SCAD, MCAD
vLCAD

Enavl-CoA
hvdratase
-I
:lJ
3-hydroxyacyl-CoA : "
C
Z
NAD+ ~ g
3-Hvdroxvaevl-CoA o
z
dehvdrogenose ,:t>
NADH+H + 1 m
Z
N
-<
3-oxoacyl-CoA ':m
s

Thiolase

acyl(n_2j-CoA

Figure 12.19 Diagram to show the fatty acid 0-oxidation pathway and the link with the respiratory chain complex. (Modified from ref 200.)

Table 12.2 [3-0xidatian of fatty acid defects that have been lead to depleted glycogen reserves, initiating ketone body
associated with sudden death production and muscle fatty acid oxidation.
In the UK, medium-chain acyl CoA dehydrogenase
Very long-chain acyl CoA dehydrogenase deficiency (VLCAD)
(MCAD) deficiency is the most commonly encountered
Long-chain 3-hydroxyacyl CoA dehydrogenase deficiency (LCHAD)
metabolic disorder leading to sudden death in infants and
Mitochondrial trifunctional protein deficiency (TFPD)
children. 206 It is clinically the least severe of the i3-oxidation
Medium chain acyl CoA dehydrogenase deficiency (MCAD)
defects. 20I,202 According to the review by Touma and
Carnitine palmitoyltransferase type 1 deficiency (CPTl D)
Charpentier 207 the average age at presentation is 13 .5
Carnitine palmitoyltransferase type 2 deficiency (CPT2D)
months and the mean age of death is 18.5 months. Patients
Plasma membrane carnitine transporter deficiency (PMCTO)
often present clinically with hypoglycaemia, acute
Multiple acyl CoA dehydrogenase deficiency (MADD)
encephalopathy (mimicking Reye's syndrome), sudden death
Carnitine acylcarnitine translocase deficiency (CACTO)
and an acute life-threatening event. 201.202,208,209 One-quarter
of patients die with a Reye-like illness or experience sudden
infant death. A previous sibling death is often recorded.
A preceding viral prodrome is common.
Long-chain 3-hydroxyacyl CoA dehydrogenase (LCHAD) The gene for MCAD has been identified and sequenced;209
deficiency is associated with a high mortality in infancy.200 up to 85 per cent of subjects with MCAD deficiency are
Cardiomegaly may be severe and hypo ketotic hypogly­ homozygous for a single gene mutation of an A-to-G
caemia may be a presenting feature. When the fetus is nucleotide at position 985. A recent prospective surveillance of
affected by LCHAD, the later stages of pregnancy may be the prevalence of the condition in the UK210 found 1.3 cases per
complicated by the HELLP syndrome in the mother. 10 000 births. The precise mechanism of death in MCAD defi­
Disorders of fatty oxidation tend to be clinically silent ciency is unclear, but hypoglycaernia, cardiac decompensation
in the absence of significant physiological stress. 20 ) Meta­ from fatty infiltration of the myocardium and arrhythmogerric
bolic events such as prolonged fasting and febrile illness blood levels of acylcamitines have been implicated.
246 I Sudden natural death in infants and children
L-..I
1cm
Figure 12.20 Liver from an infant whose sudden death was due
to medium chain acyl-CoA dehydrogenase deficiency, showing
marked pallor due to fatty in filtration.
At autopsy, pronounced fatty infiltration of the liver
(Fig. 12.20), kidney and muscle is usually found and should
prompt appropriate investigation. Occasionally, individuals
dying with MCAD deficiency do not exhibit significant
fatty infiltration of the liver,21I ·212 and its absence should
therefore not preclude investigation along these lines. A
post-mortem diagnosis can be made by demonstrating:
• a characteristic profile of medium-chain
dicarboxylicaciduria and hexanoylglycine on urinary
organic acid analysis;
• a raised plasma cis4 decenoate;
• octanoylcarnitine by tandem mass spectrometry
(MS/MS) on dried blood spots (or using the original
Guthrie screening card); and
• ~-oxidation of fatty acids in intact cultured fibroblasts.
Urine, blood, vitreous humor, tissues and a skin biopsy
are important for establishing a post-mortem diagnosis 20J
(Table 12.3). The minimal requirement of urine is 0.1 mL and,
even if the bladder seems empty, swabbing the bladder
mucosa with a cotton ball may provide a sufficient volume
for analysis. Urine (or bladder swab) samples should be
stored at -20°e. Organic acid analysis of vitreous humour
may be useful in the absence of urine. 193 Liver, skeletal and
cardiac muscle should be collected and stored without fixa­
tive at - BO°C prior to analysis. Blood should be collected
into an anticoagulant and centrifuged; blood cells and
plasma should be stored separately at - 20°e.
With the advent of gas chromatography-mass spectrome­
try (GC-MS) technology, bile, blood and cerebrospinal fluid
can be spotted onto a Guthrie card at post-mortem examina­
tion for acylcamitine analysis. 206 Vitreous humour has also
been used for the diagnosis ofMCAD deficiency. A small skin
biopsy sample taken under sterile conditions, as soon as pos­
sible after death, should be placed in tissue culture medium
containing one per cent dimethyl sulfoxide and frozen
at -70°C, or sent directly to the culture laboratory if metabolic
Table 12.3 Useful tissue samples and their storage
recommended for autopsy investigation ofpossible fatty acid
i3-oxidation defects
Urine (- 20°C)
Liver (- BO°C)
Skeletal muscle (- BO°C)
Cardiac muscle (-BO°C)
Blood into anticoagulant (centrifuged)
Blood cells - plasma (- 20°C)
Small skin sample - tissue culture + 1% dimethyl sulphoxide
(Store at -70% or send to culture laboratory straight away if
suspicion is high)
Whole blood, bile and/or cerebrospinal fluid spotted onto a
Guthrie card
disease is strongly suspected. The predominant MCAD defi­
ciency mutation can be detected by DNA analysis of the stored
dried blood spot sample taken for mass neonatal screening of
phenylketonuria,2°9 but DNA may also be obtained from
frozen tissue samples and cultured fibroblasts. 193
DISORDERS OF OXIDATIVE PHOSPHORYLATION
Compared with ~ -oxidation fatty acid defects, deaths due to
defects of OXPHOS are more likely to have demonstrable
pathology at autopsy (e.g. cardiomyopathy, liver fibrOSis,
cinhosis). These disorders are wide-ranging in their clinical
manifestation and patients only rarely present as sudden
death cases. Isolated myopathy, or a multisystem disorder
that includes encephalopathy, liver disease and cardio­
myopathy are features of the disorder. 2IJ The cardiomyopa­
thy is usually symmetrical (concentric), without outflow tract
obstruction. Death before the age of 1 year from heart failure
is the usual outcome.2l3 Histology and electron microscopy of
the myocardium reveals swollen myocytes and an increase in
the number of mitochondria. The mitochondria may also be
morphologically abnormal. The most common defects,
according to the review by Guenthard et al,213 were deficien­
cies of complexes I and lV, in isolation or combination.
An Il-month-old previously healthy boy reported by
Smeitink et al 214 died following a short Reye-like illness. A
defect in the coenzyme Q region of the respiratory chain
was found. His 6-month-old female sibling had previously
died following a similar illness.
An acute apparent life-threatening event (ALTE) was
reported as the first sign of respiratory chain complex defi­
ciency in a 4-month-old girl whose parents were consan­
guineous. 215 She was found to have markedly reduced
complex I activity.
By the time of the post-mortem examination, it is usu­
ally too late to measure the function of the respiratory
chain as enzyme activity deteriorates rapidly.19J However,
 Other bacterial infections I 24 7

Table 12.4 Clinical and biochemical indicatians far autapsy

investigation of oxidative phosphorylation disorders

Recurrent apnoea
Reye-like syndrome
Aminoaciduria
Metabolic acidosis
Hypoglycaemia
Hyperaminonaemia
Dica rboxylicacid uria
A metabolic screen with negative resu lts

adenosine triphosphate (ATP) production can be evaluated

in cultured fibroblasts and complexes II, III and IV can be
measured. It is important to note that secondalY morpho­
logical abnormalities may be found in the mitochondria of
patients with MCAD deficiency and plasma membrane car­
nitine transporter deficiency.43,2J6 Clinical and biochemical
features that should prompt consideration of OXPHOS dis­
orders as a cause of death are listed in Table 12.4.

OTHER BACTERIAL INFECTIONS

Meningococcal infection often has a fulminant onset with a Figure 12.21 A widespread maculopapular rash typical of
high mortality variously attributed to adrenal haemorrhage, meningococcaemia.
endotoxic shock or a generalized Schwalizman reaction. The
disease has two clinical forms: meningitis and meningococ­
caemia. 217 The main clinical manifestations are fever, rash
(Fig. 12.21), vomiting, lethargy, meningeal irritation and cir­
culatory collapse. The rash may be maculopapular, purpuric
or mixed. 2 18 There was a sustained increase in the incidence
of meningococcal disease in the 1990s.219
An increased incidence of the condition is observed in
late winter and spring. A wide age range is affected, with a
peak incidence in the second year. Infection in the first 2
years of life and meningococcaemia with or without menin­
gitis, compared with meningitis alone, are associated with
increased mortality. Other poor prognostic indicators are the
onset of petechiae within 12 hours prior to admission, shock,
normal or low white cell count, and a normal or low erythro­
cyte sedimentation rate.217 Autopsy often demonstrates
bilateral adrenal haemorrhage (Fig. J 2.22). This is, however,
not an invariable or specific finding, as adrenal haemor­
rhage also may be seen in haemorrhagic shock encephalopa­
thy syndrome (see above) and other forms of septicaemia.
Laboratory findings include isolation of Neisseria meningi­
tidis in the blood or spinal fluid, or the presence of Gram-neg­ Figure 12.22 Bilateral adrenal haemorrhage secondary to
ative diplococci in the spinal fluid or a petechial scraping in a men ingococcaemia.
patient with a typical clinical course. Unsuccessful attempts at
isolating the microorganism from the blood and spinal fluid ,
clinically or at autopsy, are often due to preceding antibiotic Sudden unexpected deaths in infants and young children
treatment. Antigen screening ofbJood, cerebrospinal fluid and have been ascribed to overwhelming infection as a result
urine are other diagnostic measures that may be adopted, as is of group A 0-haemolytic streptococcus, or Streptococcus
PCR amplification from blood or cerebrospinal fluid .2J9 pneu111oniae, with the organism being isolated from several
248 I Sudden natural death in infants and children
sites at autopsy.1J,220 However, such reports generally lack
microscopic evidence of sepsis or associated features and a
causal association is unclear.
Sudden death in infancy due to Haemopi1ilus injluenzae
septicaemia has been reported in infants with asplenia,
splenic hypoplasia and polysplenia, occurring in isolation
or associated with complex heart disease [with or without
accompanying visceral abnormalities).221 At autopsy, adre­
nal haemorrhage was a common finding in these cases.
Colonization or infection with Staphylococcus aureus
and group A streptococcus is implicated in the toxic shock
syndrome. 222 - 224 The condition appears to be toxin medi­
ated; the exact pathophysiology is unknown and the clini­
cal course may be rapidly fatal. Identifying the site of
infection may be difficult. Both adults and children may be
affected. The streptococcal toxic shock syndrome is charac­
terized by the sudden onset of fever, rash, vomiting and
diarrhoea, hypotension, conjunctival injection and straw­
berry tongue, followed by skin desquamation.
Necrotizing fasciitis is a rare, but often fatal, soft-tissue
infection with rapidly spreading inflammation and necrosis
of the muscle fascia and subcutaneous fat, in some instances
involving the epidermis. The condition is life-threatening
and carries a high mortality rate. Most cases are a result of
polymicrobial or group A streptococcal disease. 225 ,226 Adults
and children are affected. Pyrexia, leucocytosis, cellulitis
and oedema are the main clinical manifestations. The
affected skin becomes progressively warm, erythematous,
swollen and painful.
Invasive group A streptococcal infection [GAS), with or
without 227 necrotizing fasciitis, is the most serious compli­
cation of varicella infection in children. 22B Progression of
GAS may be very rapid, with death at home or shortly after
arrival in hospitalYs
Some patients with necrotizing fasciitis share many of
the features of streptococcal toxic shock syndrome,226 and
the onset of streptococcal toxic shock syndrome may be
preceded by necrotizing fasciitis.224 Neonatal omphalitis
can progress to necrotizing fasciitis, followed by rapid dete­
rioration and death.229 Mortality is increased by delayed
treatment and any underlying debilitating disease.226
DEATHS RELATED TO OBSTETRIC EVENTS
AND PREMATURE BIRTH
Babies born pre-term are at risk of sudden unexpected
deaths at home. The autopsy findings include no anatomi­
cal changes, coincidental disease and residual pathology
from prematurity-related complications.
Prolonged tracheal intubation and positive-pressure ven­
tilation in pre-term babies are associated with a range oftra­
cheal abnormalities including granulations, pseudopolyps,
subglottic stenosis, bronchomalacia and tracheomala­
cia. 230 - 232 Tracheobronchomalacia may cause life-threaten­
ing episodes from acute airways collapse in infants with
bronchopulmonary dysplasia.233 The pathophysiology of the
condition is not well understood but is thought to be
acquired rather than congenital. 231 Significant factors in its
production are low gestational age at birth and a high mean
airways pressure during ventilation in the first week of life.
In our own experience of 40 prematurity-related unex­
pected post-neonatal deaths, two infants were found to
have died from conditions unrelated to the prematurity
[RSV-related bronchiolitis and viral myocarditis). Twelve
infants displayed abnormalities directly or indirectly as a
result of their prematurity; these were post-hypoxic
ischaemic encephalopathy, resolving brochopulmonary
dysplasia, tracheal stenosis and stricture of the small
intestine, alone or in combination. The cerebral lesions
included periventricular gliosis, mineralization and
haemosiderin deposits, and periventricular cysts. Infants
with more extreme degrees of prematurity tended to show
more severe pathology, but even in these instances the role
of the pathology in the cause of death was often unclear.
Pathology was not found in the remaining 26 infant
deaths.
A limited respiratory reserve in infants with resolving
bronchopulmonary dysplasia has been proposed as a
mechanism of death in such cases, especially when there
was superimposed viral or bacterial infection, producing
sudden decompensation of cardiac and/or respiratory func­
tion. 23o Injury to the pulmonary vascular bed, a feature of
bronchopulmonary dysplasia, is commonly associated with
pulmonary arterial hypertension and right ventricular
hypertrophy.29 From a clinical perspective, however, the
picture is quite different, with Tammela and Koivist0 2J4
reporting an absence of fatality in a cohort of 86 low­
birth-weight babies, 23 with bronchopulmonary dysplasia,
after discharge from hospital and followed over a period of
12 months, indicating a low mortality among infants with
resolving bronchopulmonary dysplasia.
MISCELLANEOUS CAUSES OF SUDDEN
NATURAL DEATH
Sudden death has been reported in cases of congenital
myotonic dystrophy,235 although the mechanism was
unclear. Cardiac dyslythmia is also known to cause sudden
death in patients with Emery-Dreifuss muscular dystro­
phy.2J6 An occult adrenal phaeochromocytoma as a cause
of sudden death in a young boy following a kick to the
anterior abdomen has been ascribed to a sudden surge of a
large amount of noradrenaline into the circulation, produc­
ing violent myocardial contractility.2J7 The sudden death of
an infant who subsequently demonstrated widespread his­
tological features indistinguishable from Kikuchi-Fujimoto
disease has been reported. 238
Encephalopathy in association with acute pancreatitis is
welJ recognized in adults and has been rarely reported in chil­
dren.239 The condition is thought to be a result of damage to

Figure 12.23 The myocardium shows frequent contraction band
necrosis. Sudden death followed a kick to the anterior abdomen in
a boy with a right adrenal pha eochromocytoma .
brain tissue caused by the circulatin g pancreatic enzymes.
The initial clinical picture may be confused with Reye's syn­
drome.
Premature atherosclerotic coronary disease in subj ects
aged between 14 and 19 years is rarely described as t he
cause of sudden death .41 Precocious coronary atheroscle­
rotic disease in young patients is alm ost always secondary
to type II hypercholesterolaemia, system ic lupus elyth e­
matosus or juvenil e- onset diabetes. 59
Cerebral disease has long been known to cause electro­
cardiographic changes and has been linked to overactivity
of the sympathetic limb of the autonomic nervo us system.
Autopsied cases h ave shown hi stologica l changes of the
myocardium ranging from normal muscle to severe
necrotic lesions with mononuclear cell 'infiltration;26 con­
traction band necrosis of t he myoca rdium is a characteris­
tic finding (Fig. J2.23) . Similar les ions of the myocardium
are reported to be ca used by stress, catechol a mine infusion
and reperfusion inj u ry.
Opening of the calcium channel with infl ux of calcium into
the cell and efflu x of potassium causes interaction between
the actin and myosin filaments, leading to cell death from
hypercontraction, is the likely mechanism. A similar mecha­
nism may provide an explanation for sudden death occurring
with epilepsy, asthma and 'stress' fro m other causes. 26
Sudden death associated with 'inte rm ed iate ' pathology I 249
SUDDEN UNEXPLAINED DEATH IN
OLDER CHILDREN
Many of the reported series of sudden death include a propor­
tion of older children in whom the cause is unexplained.
Thus, 11 of 169 (6.5 per cent) sudden deaths in children and
ado lescents aged 2- 20 yea rs reported by Keeling and
Knowles l5 were unexplained. However, most of the cases had
not been investigated toxicologically or microbiologically,
and sampling for microscopy was limited. The study reported
by Molander,1 3 from southem Sweden, included four unex­
plained deaths in 389 sudden natural deaths in subjects
between the ages of 1 and 20 years. Similarly, sudden death
in 29 of 207 subj ects aged 1-21 yea rs was reported as being
of undetermined cause. 20 The subjects were mainly between 1
a nd 4 years of age, but were not discussed in any detail.
' Bangungut' is a con dition that affects yo ung men from
t he Philippines, J apan, Vietnam and Thailand. The victims
a re excl usively male and usually die mysteriously at night.
The mechanis m of death appears to be ventric ular fibrill a­
tion witho ut underly ing card iovasc ul ar disease. Thiamine
deficiency h as been suggested as a contributing factor. 59
SUDDEN NATURAL DEATH IN THE EARLY
NEONATAL PERIOD
Hypoxic ischaemic encephalopathy, bacterial infection,
undiagnosed cardiac malformation and genetic metabolic
disease are important causes of sudden death at this age.
Several of the enzyme deficiencies that interfere with ~ -oxi­
dation of fat may also be responsible for sudden death.240
Disseminated herpes simplex infection has been reported 241
(see Fig. 11.7, p. 209). Maternal evidence of infection may
be absent and the disease in the baby may not be clinically
manifest.
Mercuri et al 242 reported several in fants born with no r­
mal Apgar scores who subsequently developed seizures
after a peliod ranging from 10 hours to 4 days. In all cases,
ischaemicjhaemorrhagic les ions, sustained either pe rin a­
tally or post seizure, were demonstrated on bra in im aging.
One of these infants died at 3 days of age.
SUDDEN DEATH ASSOCIATED WITH
'INTERMEDIATE' PATHOLOGY
Norman et al 243 have already drawn attention to findings
in paediatric a utops ies w hose significance is questionable.
Examples are s ma ll ca rdi ac fibrom as, abnorma l cerebra l
gyral pattern and atrial septa l defects. The issue has already
been raised in relation to premature babies who die unex­
pectedly at home a nd whose brains reveal changes of a
previous hypoxicjischaemic event. Mi ld degrees of hydro­
cephalus, neuronal heterotopias or a ventricular septal
defect are other entities that may be difficult to eva luate in
cases in sudden in fa nt deaths.
250 I Sudden natural death in infants and chi ldren
Figure 12.24 Microscopy of a microglial nodule in the
brainstem of an infant with cytomegaloviral parotitis w ho died
suddenly at home.
Cytomegalovirus-related parotitis is found in a small
number of infants who die suddenly. These cases may show a
few microglial nodules in their brainstem (Fig. 12.24),244.245
but it is unclear whether they have role in the cause of death.
REFERENCES
Lambert EC, Menon VA, Wagner HR, Vlad P. Sudden
unexpec ted death from cardiovascular disease in children.
A cooperative international study. Am J Cardiol 1974;
34:89 - 96.
2 Abu -Harb M, Hey E, Wren C. Death in infancy from
unrecognised congenital heart disease. Arch Dis Child 1994;
71:3-7.
3 Basso C, Frescura C, Corrado D et al. Congenital heart
disease and sudden death in the young. Hum Pathol 1995;
26(10):1065-72.
4 Stahl J, Santos LD, Byard RW. Coronary artery
thromboembolism and unexpected death in childhood and
ado lescence. J Forensic Sci 1995: 40:599-601.
Klitzner TS. Sudden cardiac death in children. Circulation
1990; 82:629-32.
6 Steinberger J, Lucas RV, Edwards JE, Titus JL. Causes of
sudden unexpected cardiac death in the first two deca des of
li fe. Am J Cardiol 1996; 77:992-5.
7 Glew RH, Varghese PJ, Krovetz LJ et al. Sudden death in
congenital aortic stenos is. Am Heart J 1969; 78:615-25.
8 McManus BM, Babul S. Pathology of sudden death in th e
young. Can J CaTdioil996; 12:1150-6.
9 Byard RW, Bourne AJ, Adams PS. Subal1erial ventricular
septal defect in an infant with sudden unexpected death:
Cause or coincidence? Am J Cardiovasc Pathol 1990; 3:333 -6.
10 Byard RW. Vascular conditions. In Byard W, Cohle SD (ed s)
Sudden Death ill Injanty, Childhood and Adolescence.
Cambridge: Cambridge University Press, 1994, pp. 175- 261.
II Bird LM, Billman GF, Lacro RV et al. Sudden death in
Will iams syndrome: report of ten cases. J Pediatr 1996;
129:926-3 1.
12 DeSa DJ. Isolated myocarditis in the first year. Arch Dis
Child 1985; 60:484-5.
13 Molander N. Sudden natural death in later childhood and
adolescence. Arch Dis Child 1982; 57:572-6.
14 Siboni A, Simonsen J. Sudden unexpected natural death in
young persons. Forensic Sci lilt 1986; 31 :159-66.
15 Keeling JW, Knowles SAS. Sudden death in childhood and
adolescence. J Patliol 1989; 159 :221 -4.
16 Davies MJ. The cardiomyopathies: a review of terminology,
pathology and pathogenesis. Histopathology 1984; 8:363-93.
17 Noren GR, Staley NA, Bandt CH, Kaplan EL. Occurrence of
myocard itis in sudden death in children. J Forellsic Sci
1977; 22: 188-96.
18 Davies MJ, Popple A. Sudden unexpected cardiac death: a
practical approach to the forensic problem. Histopathology
1979; 3:255-77.
19 Lim C-H, Toh CCS, Chia B-L. Low L-P et al. Stokes-Adams
attacks due to acute nonspecific myocarditis. Am Heart J
1975; 90:172-8.
20 Neuspiel DR, Kuller LH. Sudden and unexpected natural
death in childhood and adolescence. J Am !VIed Assoc 1985;
254: 1321-5.
21 Woodruff JF. Vira l myocarditis. Am J Patho1 1980;
101 :427-42.
22 Leslie K, Blay R, HaiscJl C et al. Clinical and experimental
aspects of vira l myocarditis. Ciin Microbiol ReI! L989;
2:191-203 .
23 Hilton DA, Variend S, Pringle JH. Demonstration of
coxsackie virus RNA in fonnalin-fixed tissue sections from
childhood myocarditis cases by ill sirt! hybridization and the
polymerase chain reaction. J Pathol 1993; 170:45-51.
24 Lo zinski GM, Davis GG, Krous HF et al. Adenovirus
myocarditis : retrospective diagnosis by gene amplification
from formalin-fixed, paraffin-embedded tissues. Hum Pathol
1994; 25:831-4.
25 Shimizu C, Rambaud C, Cheron G et al. Molecu lar
indentification of viruses in sudden infant death associated
wi th myocarditis and pericarditis. Pediatr lilject Dis J 1995;
14:584-8 .
26 Samuels MA . Neurogenic heart disease: A unifying
hypothesis. Am J Cardiol 1987; 60:15-1 9.
27 Frustaci A, Loperfido F, Gentiloni N et al. Catecholamine­
indu ced cardiomyopathy in multiple endocrine neoplasia: A
histolo gic, ultrastructural and biochemical study. Chest 1991;
99 :382-5.
28 Olsen EGJ. Myocarditis. In Olsen EGJ (ed.) The Pathology
oj the Heal1. London: The MacMillan Press Ltd, J980;
178-99 .
29 Patterson K, Donnelly WH, Dehner LP et al. The
cardiovascular system. In Stocker JT, Dehner LP (eds)
Pediatric Pathology Vol. I. Philadelphia: Lippincott
Company, 1992 ; 575 -651.
30 Haupt HM, Hutchins GM. The heart and cardiac conduction
system in polymyositis-dermatomyositis: A clinicopathologic
study of 16 autopsied patients. Am J Cardiol 1982;
50 :998-1006.
31 Bissenden JG, Hall S. Kawasaki syndrome: lessons for
Britain. Br Med J 1990; 300: 1025-6.

32 Davies MJ, McKenna WJ. Dil a ted card iomyopathy: an
introduct io n to pathology and pathogenes is. Br Ht'Grt J
1994; 72(Supp l.):24.
33 Kelly DP, Stra uss AW. Inherited cardiomyopathies. N Eng l J
Med 1994; 330:913-19.
34 Lawson MR. Southwestern Interna l Medicine Conference:
hypertrophic cardiomyopathy: current views on etiology,
pathophys iology and managem ent. Am J Med Sci 1987 ;
294:191 - 210.
35 Burch M. Hypenrophic ca rdiomyopathy. Arch Dis Child
1994; 71:488-9.
36 McKe nn a WJ, S tewa n JT, Nihoyannopoulos P et al.
Hypertrophic cardiomyopathy without hypertrophy : two
families w ith myocardial disarray in the abse nce of
increased myocardial mass. Br Hea rt J 1990; 63 :287 -90.
37 Gow RM. Sudden cardiac death in the young. Can J Cardiol
1996; 12: 1157-60.
38 Burch M, Mann JM, Sh a rl and M et a l. Myocardial disarray in
Noonan sy ndrome. Br Heart J 1992; 68:586- 8.
39 Ha lliday HL. Hypenrophic ca rdiomyopathy in infa nts of
poo rl y-co ntroll ed diabetic mothe rs. Arch Dis Child 198 1;
56 :258-63.
40 Burch M, Runcima n M. Dilated ca rd io myopathy. Arch Dis
Child 1996; 74:479-81.
41 Maron BJ, Shirani J, PoJiac LC et a l. Sudden death in young
competitive athletes: clini cal, demog raphic, and path o logical
profiles . J Am !VIed Assoc 1996; 276: 199-204.
42 Huber S. Viral myocarditis - a tale of two diseases. Lab
hllJesr 1992; 66:1- 3.
43 Be nnett MJ, Hale DE, Pollitt RJ et al. Endocardial
fibroelastosis a nd prima lY ca rnitine deficiency due to a
defect in the plasma membran e ca rnitine tran sponer. Ciil1
Cardiol 1996; 19: 243 - 6.
44 Virman i R, Robin owitz M, Clark MA, McAllister HA. Sudden
death and partial absen ce of the right ventricular
myocardium. Arch Pathol Lab Med 1982; 106:163-7.
45 Thi ene G, Nava A, Corrado D e t a!. Right ventricul ar
ca rdiomyopathy and sudd en deat h in young peopl e. N Engl
J Med 1988 ; 318:129-33.
46 McKe nna WJ, Thi ene G, Nava A et al. Diagnos is of
a rrh y thmoge nic ri g ht ventric ul a r dysplasia/ca rdi o myopathy.
Br Heart J 1994; 7 1:2 15-1 8.
47 Fonta ine G. Arrhythmoge ni c rig ht ven tricula r dyspla sia. Curr
Opin Cardiol 1995; 10: 16-20.
48 A hmad F, Li D, Karibe A et al. Localisation of a gene
responsible for a rrh ythmoge ni c right ventricul ar dyspla sia to
chromosome 3p23. Circu lation 1998; 98 :2791 - 5.
49 Gerlis LM, Schmid t-O tt SC, Ho SY, Anderson RH. Dysp lastic
cond itions of the right ventricu lar myocardium: Uhl's
anoma ly versus arrhythmoge nic right vent ricu lar dysplasia .
Br Heart J 1993; 69: 142 - 50.
50 Tansey OK, A ly Z, Sheppa rd MN. Fat in the ri g ht ventricle of
the no rmal heart. Histopath ology 200 5 ; 46 :98- 104.
51 Krous HF, Ch adw ick AE, Isaacs H. Tum ors associated w ith
sudden infan t and childhood death. Pedia tr Dev Pathal
2005; 8:20-5.
52 Shehata BM, Patterson K, Thomas JE et al. Histiocytoid
cardiomyopathy: Three new cases and a review of the
literature. Pediatr Dev Parhol 1998; 1:56-69.
53 Suarez V, Fuggle WJ. Cameron AH et al. Foamy myocardia l
transformation of infa ncy: a n inherited disease. J Ciill Pathol
1987; 40:3 29 - 34.
54 Koponen MA, Siegel RJ. Hi st iocytoid ca rdi omyopathy a nd
sudden death. Hum Pathol 1996; 27:4 20-3 .
55 Ness MJ. McManu s BM. Anomalous right coronary artery
origin in otherw ise unexplained infant death. A rch Pat/101
Lab Mcd 1988; 112:626-9.
Referen ces I 251
56 Lipsett J. Byard RW, Carpe n ter BF et at. Anomalous co ro nary
aneries arising from the ao rta assoc iated with sudden death
in infancy and ea rly childh ood. Arch Pa thol Lab Med 1991 ;
115 :770- 3.
57 Corrado D. Thiene G. Cocco P, Frescura C. Non­
atherosclerotic coronary artery disease and sudd en death in
the young.
Br Heart J 199 2 ; 68 :60 1-7.
58 Garfia A, Rodrigu ez M, Chava rria H. Garrid o M. Sudden
ca rdiac death during exe rcise due to an isolated multiple
anomaly of the left co ronary a rtelY in a 12-year- o ld g irl:
Clinicopatholog ic fin dings. J Fo rens ic Sci 1997 ; 42:330-4.
59 ScuJly RE, Mark EJ . McNeely W F. McNeely BU. Case reco rds
of th e Massachusetts Genera l Hospital . Case 22 -1 989. N Engl
J lVled 1989; 320: 1475 - 83 .
60 Thiene G, Pennelli N, Rossi L. Cardiac conduction system
abnormaliti es as a possible cause of sudd en de a th in yo ung
athletes. Hum Pathol 1983; 14:704-9.
61 Angelini P. Normal and anomalous coronalY arte ries :
detinitions and classification . A 111 Heart J 1989; 117 :41 8- 34.
62 Price J . Kawasaki sy ndro me. Br Med J 1984 ; 288:262 - 3.
63 Curti s N. Kawasaki disease. Br Med J 1997 ; 315:322-3 .
64 Ti za rd EJ, Suzuki A. Levin M. Dill on MJ. Clinical as pects of
100 patients wit h Kawasa ki disease . Arch Dis Child 1991 ;
66:18 5- 8.
65 Nakamura Y. Yanagawa H, Ojima T et al. Cardiac sequelae of
Kawa saki disease among recurrent cases . Arch Dis Child
1998; 78:1 63 -5.
66 Hoare S. Abinun M. Cant AJ. Ov erl ap between Kawasaki
disease and Group A streptococcal infection (Letter). Pediatr
/l ljixt Dis J 1997; 16:633 -4.
67 Bya rd RW. Cardiac co nditions. In Byard RW (ed.) Sudden
Death in Infancy Childh ood and Adolescellce. 2nd edn.
Cambrid ge: Ca mbrid ge University Press. 2004 , pp. 202 - 48.
68 Bharati S. Bauernfiend R, Scheinm an M et al. Congenita l
abnormalities of the conducti on system in two patients with
tachyarrhythmias. Circulation 1979; 59:593-606.
69 Splawski I, S he n J, Timothy KW et al. Spectrum of mutations
in long Q-T syndrome genes KVLQTI. HERG. SCN5A.
KCNE I and KNCE2. Circula tiol1 2000 ; 102: 1178-85.
70 Towbin JA. Molecular basis of sudden ca rdiac death. Pediatr
Ciin N Am 2004; 51: 1229 - 55.
71 Moran JJ. Idiopathic arterial calcificatio n of infancy: a
clin icopatho logic study. Pa t/wi Ann 1975; 10:393-417.
72 Rutsch F. Vainganka r S. Johns on K et al. PC-I nucleos ide
triphosphate pyrophosphohydrola se defiCiency in idiopathi c
infanti le arte rial calcification. Am J Parha l 2001;
158 :543-54.
73 Rutsch F. Ruf N. Vaingankar S et al. Mutations in ENPP 1 a re
associated with 'idiopathi c' infantile arterial calcification.
Nat Geller 2003; 34:379-81.
74 Juul S. Ledbetter D. W ight T N, Wood rum D. New in sights
into idiopathic infantile arteria l ca lci nos is. Three patient
repons. Am J Dis Child 1990; 144 :229-33 .
75 Byard RW. ldiopathic arterial calcification and unexpected
infant death . Pediatr Pathol Lab Med 1996; 16:985-94.
76 Joshi VV. Pawel B. Connor E et al. Arteriopathy in children
with acquired immune de fi ciency syndrome. Pediatr Pathol
1987; 7:261-75.
77 Siegel RJ. Dunton SF. Systemic occlusive arteriopathy with
su dden death in a 10-year- old boy. Hum Parhol 1991;
22:197- 200.
78 Im a mura M. Yokoya ma S. Kikuchi K. Coronary
fibromuscular dys plasia presen tin g as sudden infant deat h.
Arch Pathol Lab Med 1997; 121 : 159-6 1.
79 Rosenberg HG. Systemic arterial disease and chronic
arse ni cism in infants. Arch Pathol 1974; 97:360- 5.
252 I Sudden natural death in infants and children

80 Rol fes DB, Towbin R, Bove KE. Vascular dysplasia in a chil d 104 Byard RW, Bo urn e AJ, Hanieh A. Sudden and unexpected
with tube rou s sclerosis. Pediatr Pathol 1985 ; 3 :359-73. death due to hemorrhage from occult central nervo us system
81 Mudd SH, Levy HL, Skovby F. Disord ers of transsulfuration. lesions. Pedia tr Neurosurg 1991 -92; 17:88-94.
In Sc river CR, Beaudet Al, Sly WS, Va lle D (ed s) The 105 Lill ey man JS. Intrac rania l haem orrhage in idiopathic
Meta bolic and Molecular Bases of Il1h erited Disease. New thrombocy topeni c purpura. Arch Dis Child 1994; 71 :251-3.
Yo rk: McGraw-Hili Inc. , 199 5; 1279- 327. 106 Abu AI Ragheb SY, Koussous KJ, Arm SS. Intracra nial
82 Fo rtu in NJ, Morrow AG, Roberts We. Late vascular
neop lasms associa ted with sudd en death : a rep ort of seve n
manifesta tions of the Ru be lla synd rom e. Am J Med 1971;
cases a nd a review of the literature. Med Sci Law 1986 ;
51: 134-40. 26:270-2.
83 Hamilton SJ, Allard MF, Friedm a n JM. Cardiac Findin gs in 107 DiMa io SM, DiMaio VJM, Kirkp atrick J B. Sudden ,
an indiv idual with neurofibromatos is I and sudden dea th . un expec ted death s due to primary intracranial neopl as ms.
Am J Med Genet 2001; 110:95-9. Am J Foren sic Med Pathol 1980 ; 1:29-45.
84 Uno H, Arya S, Laxova R, Gilb ert EF. Menkes' syndro me
108 Bya rd RW, Moo re L. Sudd en and unexpec ted death in
with vascula r and ad ren ergic nerve abn ormalities . Arch
childhood du e to a co lloid cyst of the third ven tricle.
Path ol Lab Med 1983; 107 :286- 9.
J Forens ic Sci 1993; 38 :210 -13.
85 Fleishe r GR , Buck BE, Co rnfeld D. Prim ary intimal 109 Buzzi S, Verdura C, Arlati S, Colecchia M. Sudden deat h in a
fibroplasia in a child with Down's syndrome. Am J Dis Child child due to ra re endo crani al neo fo rm ation . Med Sci Law
1978 ; 132 :700-3 . 1998; 38: 176-8.
86 Cina SJ, Smialek, JE, Burk e AP et al. Primary ca rdiac tumors 110 Leestma JE, Walcz ak T, Hughes JR et al. A prospective study
causing sudden death: a review of the literature. Am J on sudden une xpected death in epilepsy. Ann Neural 1989;
Forensic Med Pa thol 1996 ; 17:271-81. 26:195-203.
87 Siega l GP. Pri mary tumors of mu scl e. In Stocker JT, Askin
II I Byard RW, Keeley FW, Smith CR. Ty pe IV Ehlers-Danl os
FB (eds) Pathology of Solid Tum on il1 Ch ildren. London:
sy nd rome presenting as sudd en infant death. Am J Clin
Chapman and Hall Medical, 1998, pp. 161-81.
Patl101 1990 ; 98:579-82.
88 Parker KM , Embry .IH. Sudden dea th due to tri cuspid va lve 11 2 Elterman RD, Bodensteiner JB, Barnard JJ. Sudden
myx om a with massive pulmonary emb olism in a 15-m onth unexp ec ted death in patien ts with Dandy-Walker
old male. J Forensic Sci 199 7; 42 :524-6. malformation . J Child Neurol 199 5; 10:382-4.
89 Krous HF, Chapman AJ, Alts hul er G. Cardiac hema ngioma: a 113 Vaishnav A, Ma cKinn on AE. Prog ressive hydrocephalus in
rare (or poss ible) cause of su dden death in child ren. J teena ge spina bi fid a patients. Z Kindercilinl rg 1986;
Foren sic Sci 1978 : 23:375-8. 41(Suppl. 1):36-7.
90 Swalwell CI. Be nign intra cardiac teratoma - a case of
114 Stellm a n GR, Banniste r CM, Hillier V. The incidence of
sudd en death. A rch PatilOl Lab Med 1993; 117 :739-42.
seizure disorder in children with acq uired a nd congenital
91 Byard RW, Cutz E. Sudden and un expec ted dea th in in fa ncy hydrocephalus. Z Kinderchirurg 1986 ; 41( Suppl. 1):38-9.
and ch ildhood due to pulm ona lY thro mb oemb olism. Arch 115 Friede RL. Meningoencephalitic pro cesses in th e perinata l
Pathol Lab Med 1990 ; 114:14 2-4. period. Deue!opmental Neuropathology. New York: Sprin ge r­
92 Champ C, Byard RW. Pulmonary thromboembolism and
Verlag, 1975 ; 178-87.
unexpected dea th in infancy. J Pediatr Child Health 199 4;
116 Howat AJ, Smith CML, Variend S. Sudden infant death a nd
30:550-1.
encephalitis in two sib lings. Dev Med Child Neu ra l 1987;
93 Pa tton DJ, Ga lliani CA, Johnson WH, Hed lund GL. Sudd en
29:539-40.
death in Marfan syndrome. A m Joumal of Roentgenology
117 Shield s LBE, Handy TC, Parker JC, Burns e. Postmoltem
1995; 165:1 60.
diagnosis of leukodystrophi es. J Forensic Sci 1998 ;
94 Masse J-F, Perusse R. Ectoderm al dysplasia. Arch Dis Child 4 3:1068 -71.
1994; 71: 1- 2. Jl8 Kereny i N. Conge nital ad renal hyp opla sia. A rch Pathol 196 1;
95 Clarke A, Phillips DIM , Brown R, Harpe r PS. Clini cal aspects 71 :336- 3.
of X-linked hypohidrotic ectodermal dyspl as ia. Arch Dis 119 Irvine WJ, Toft AD. Diagnosing adrenocortical insufficiency.
Child 198 7: 62:989-96. Practitioner 1977; 218 :539-45.
96 Waters KA, Evere tt F, Sillen ce D et al. Breathing
120 AI Sabri ANI, Sm ith N, Busuttil A. Su dden death due to a uto­
abno rm a lities in sleep in achondroplasia. Arch Dis Child
immun e Addiso n's disease in a 12-yea r-old girl. Int J Legal
1993; 69:191-6.
Med 199 7; 110 :278-8 0.
97 Bla nd JD, Emery JL. Unexp ected death of ch ildren with
121 Fava ra BE, Franc iosi RA, Mil es V. Idiopathic adrena l
achondroplasia a fter th e perin a tal period . Dev Med Child
hypop lasia in children. Am J ((in Pa tl10 I 1972 ; 57 :287-96.
Neurol 1982: 24:489-92.
122 Brown W, Singer DB. Pregnan cy-indu ced hyp ertension and
98 Webb DW, Osbo rne JP. Tuberous scl erosis. A rch Dis Child congenital adrenal hypoplasia. Obstet Gynaecol 1988;
199 5; 72:471-4. 72: 190-4.
99 Nir A, Taj ik J , Freeman WK et a l. Tuberous scl erosis and 123 Bacon CJ , Hall SM. Haemorrh agic s hock encephal opa thy
cardia c rh abdomyosarcoma. Am J ·Cardiol 1995; 76:4 19 - 21. syndrome in the British Isles. A rch Dis Child 199 2;
100 O'CalJaghan FJK, Clarke AC, Joffe H et al. Tuberous scle ros is 67:985-9 3.
complex a nd Wo lff-ParKinson - Wh ite sy ndrome. Arch Dis 124 Li ttle D, Wilkin s B. Hemo rrhagic shock and encephalopa thy
Child 1998 ; 78: 159-62. syndrom e : an unusual cause of sudd en death in children.
101 Livingsto n JH , Brown JK. Intrac erebra l haemorrhage after Am J Forensic Med Pathol 1997; 18 :79-8 3.
the neon ata l period. Arch Dis Child 1986; 61 :538-44. 125 PHLS Communicab le Disease Surveill ance Centre. Joint
102 Gieteling EW, Rinkel GJE. Ch aracteristics of intracranial British Paediatric Association a nd Communicable Disease
a neu rysms and subara ch noid haemorrhage in patients with Surveillan ce Scheme for haemorrhagic sho ck encephalopathy
polycysti c kidn ey disease. J Neural 2003; 250:418- 23. syndrome surveill a nce report for 198 2-4. Br Med J 1985 ;
103 Lilova MI , Petkov DL. Intracranial aneurysms in a child with 290: 1578-9.
autosom a l. recessive poly cysti c kidney disea se. Pediatr 12 6 Glasgow JFT, Mo ore R. Reye's sy ndro me 30 years on. Br
Nephra l 2001 ; 16: 1030- 2. Med J 1993; 307 :9 50-1.

127 Anon. Rey e's syndrome and aspirin: epidemiolog ical
associations and inborn erro rs of metabol ism. Lancet 1987;
ii:429-31.
128 Bonnel HJ, Beckwith JB. Fatty liver in sudde n childhood
death. Am J Dis Child 1986; 140:30-3.
129 Forsyth BW, Shapiro ED, Horwitz Rl et al. Misdiagnosis of
Reye's-like illn ess (Letter). Am J Dis Child 1991 ; 145:964-6.
130 Green A, Hall SM. Investigation of metabolic disorders
resemblin g Reye's syndrom e. Arch Dis Child 1992; 67: 13 13-17.
131 Hardie RM, Newton LH, Bruce JC et al.. The chan ging clinical
pattern of Reye's sy ndrome 1982- 1990. Arch Dis Child
1996; 74:400-5.
132 Rowe PC, Valle D, Brusilow Sw. Inborn errors of metabolism
in children referred with Reye's synd rome . JAm Assoc 1988;
260:3167-70.
133 Byard RW, Bohn DJ, Wilson G et al. Unsuspected
diaphragmatic hernia: a potential cause of sudden and
unexpected death in in fancy and early childhood. J Pediatr
Surg 1990; 25:1166-8.
134 Pfalzgraf RR, Zumwalt RE, Kenny MR. Mesodivelticular
band and sudden death in children. Arch Puthol La.b Med
1988; 112:182-4.
135 Pumphrey RSH, Stanw orth SJ. The clinical spectrum of
anaphylaxis in north-west England. c/in Exp Allergy 1996;
26: 1364-70.
136 Pumphrey RS, Robert IS. PostmOltem findings after fatal
anaphyl actic reactions. J c/in Pa thol 2000; 53 :273-6.
137 Delage C, lrey NS. Anaphylactic deaths: a clinicopathologic
study of 43 cases. J Forensic Sci 1972; 17:5 25 - 39.
138 Sampson HA, Mendelson MD, Rosen JP. Fatal and near-fatal
anaphyl actic reaction s to food in childre n and adolescents.
N Engl J Med 1992 ; 327:3 80-4.
139 Riches KJ, Gillis D, James RA. An autoJlsy approach to bee
sting-related death s. Pathology 2002; 34:257-62.
140 Byard RW. Miscellaneous conditions. In Byard RW (ed.) Sudden
Death ill Infancy, Childhood and Adolescence, 2nd ed n.
Camb ridge: Cambridge University Press, 2004, pp. 469-88.
141 Assem ESK, Gelder CM, Spiro SG et al. Anaphylaxis induced
by peanuts. Br i\lIed J 1990; 300: 1377 -8.
142 Patel L, Radivan FS, David T1. Management of anaphylactic
reactions to food. Arch Dis Child 1994; 71:370-5.
143 Powars D, Chan LS, Schroeder WA. The variable expression
of s ickle cell di sease is genetica lly determined. Semil1
Hematol1990; 27:360-76.
144 Jenkins ME, Scott RB, Bairs RL Stud ies in sickJe cell
anaemia. J Pedia tr 1960; 56:30-9.
145 Leikin SL, Gallagher D, Kinney TR et al. Mortality in children
and adolescen ts with sickJe cell disease. Pediatrics 1989;
84:500-8.
146 Clini copathologic Conference. Sudden death in a young
woman with sickle cell anaemia. Am J Med 1992;
92:556-60.
147 Cutz E. Idiopathic pu lmonary hemosiderosis and related
disord ers in infan cy and childhood. In Rosenb erg HS,
Bernstein J (eds) Respiratory I1l1d Alimentary Tract Diseases.
Basel: Boehm-Hutter Press, 1987, pp. 47-81.
148 Stocker JT. The respiratory tracr. In Stocker JT, Dehner LP
(eds) Pediatric Pathology Vol. 1. Philadelphia: Lippincott
Company, 1992; 505-73.
149 Schwartz MZ, Fill.er RM. Tracheal compression as a cause of
apnea following repair of tracheo-oesophageal fistula:
treatment by aortopexy. J Pediatr Surg 1980; 15:842-8.
150 Trollfors B, Nylen 0 , Strangert K. Acute epiglottitis in child ren
and adults in Sweden 1981-3. Arch Dis Child 1990; 65:491 -4.
151 Claesson B, Troll fors B, Ekstrom-Jodal B et al. Incidence and
prognosis of ac ute ep iglottitis in children in a Swed ish
region. Pediatr Infect Dis 1984; 3:534-7.
References I 253
152 Coulthard M, Isaacs D. Retroph ary ngeal abscess. Arch Dis
Child 1991; 66: 1227-30.
153 Barrat GE, Koopmann CF Jr, Coulthard Sw. Retropharyngeal
abscess - a ten-year experience. Laryngoscope 1984;
94:455-63.
154 Brook J. Microbiology of retropharyngeal abscesses in
chi ldren. Am J Dis Child 1987 ; 141:202-4.
155 Donne lly BW, McMillan JA, Wei ner LB. Bacterial tracheitis:
Report of eight new cases and review. Rev Infect Dis 1990;
12:729-35.
156 Britto J, Habibi P, Walters Set al. Systemic complications
associated with bacterial tracheitis. Arch Dis Child 1996;
74:249-50.
157 McKenzie M, Norman MG, Anderson m, Thiessen PN. Upper
respiratory tract infection in a 3 year old girl. J Ped iatr
1984; 105:129-33.
158 Cant AJ, Gibson PJ, West RJ. Bacterial tracheitis in Down's
synd rome. Arch Dis Child 1987; 62:962- 3.
159 Pan itch HB, Callahan CW, Schidlow DV. Bro nchiolitis in
children. Clin Chest Med 1993 14:715-31.
160 Price JE Acute and long-term effects of vi ra l bronchiolitis in
infancy; Lung 1990; 168(Suppl.):414-21.
161 Meert K, Heidemann S, Abella B, Sarn aik A. Does
prematurity alter the course of respiratory syncytial virus
infection? Crit Care Med 1990; 18 : 1357- 9.
162 MacDonald NE, Breese Hall C, Suffin SC et al. Respiratory
syncytial viral infection in infants with congenital heart
disease. N Eng! J Med 1982; 307:307-400.
163 Groothuis JR, Gutierrez KM, Lauer BA. Respiratory syncyt ial
virus infection in children wit h bronchopulmonary
dysplasia. Pediatrics 1988; 82:199 - 203.
164 Hall CB, Powell KR, Ma cDonald NE et al. Respiratory
syncytial viral infection in chi ldren with compromised
im mun e function. N EI1g1 J Med 1986; 315:77-81.
165 Hastleton PA, Ironside JW, Whittaker JS et al. Pulmonary
veno-occlusive disease. A report of four cases.
Histopathology 1986; 10:933-44.
166 Cagle P, Langston C. Pulmonary veno -occlus ive disease as a
cause of sudden infa nt death. Arch Pat/wi Lab Med 1984;
108:338-40.
167 Jay GW, Leestma JE. Sudden death and epilepsy. Acta
Neurologica Scandinavica 1981; 63(Suppl. 82): 11 -16.
168 McGugan EA. Sudden unexpected deaths in epileptics­
a literature review. Scot iVIed J 1999; 44:137-9.
169 Shields WD. Status epilepticus. Pediatr Clin North Am 1989;
36:383 - 93.
170 Leestma JE. Forensic aspects of complex neural
dysfunctions. In Leestma JE (ed.) Forensic Neuropatho logy.
New York: Raven Press, 1988, pp. 396-428.
171 Coyle HP, Baker-Brian N, Brown Sw. Coroners' autopsy
reporting of sudden unexplained death in epil epsy (SUDEP)
in the UK. Seizure 1994; 3:247-54.
172 Nashef L, Walker F, Allen P et al. Apnoea and bradycardia
during epilept ic seizures: relation to sudd en death
in epil epsy. J Neurol Ne'urosuTg Psychiatry 1996;
60:297-300.
173 Lip GYH, Brodie JVIJ. Sudden death in epilepsy: an avoidab le
outcome? J Royal Soc Med 1992; 85:609-11.
174 Brown SW, Mawer GE, Lawler W et al. Sudden death and
epilepsy (Letter). Lancer 1990; 335:606-7.
175 Timmings PI.. Sudden unexpected death in epi lepsy: is
carbamazipine implicated? Seizure 1998 ; 7:289-91.
176 Gars on A Jr. Medicolegal probl ems in the management
of cardiac arrhythmias in children. Pediatrics J 987 ;
79(1):84-8.
177 Brown DC, Godman MJ. Life-threatening 'epi lepsy'. Arch Dis
Child 1991; 66:986-7.
254 I Sudden natural death in infants and children
178 Ch amp C, By ard RW. Sudden death in asthma in chi ldhood .
Forens ic Sci Int 1994; 66:117-27.
179 Zach MS , Karner U. Su dden death in asthma. Arch Dis Child
1989; 64: 1446-51.
180 Carswell F. Thirty deat hs from asthma. Arch Dis Ch ild 1985;
60: 25 -8.
181 Kravis LP, Kolski GB. Unexpected death in childhood asthma.
Am J Dis Child 1985; 139:558-63.
182 Benatar SR. Fatal asthma. N Engl J IVIed 1986; 314:423-9.
183 Larsen GL. Ast hma in children. N Engl J IVIed 1992;
326:1540- 5.
184 Molfino NA, Nannini LJ. Martelli AN, Sl utsky AS.
Respiratory a rrest in near-fatal asthma. N Engl J Med 1991 ;
324:285- 8.
185 Haalboom JRE, Deenstra M. Stmyvenberg A. Hypokalaemia
induced by inhalation of fenot erol. Lallcet 18 May 1985 ;
2 :1125-7.
J 86 Pearce N. Crane J. Burgess C et al. Beta ago nists a nd asthma
mOl1a lity: deja vu. Clin Erp Allergy 1991 ; 21 :401 - 10.
187 Warner DP, McKinn ey PA. La w GR, Bod ansky HJ. MOl1ality
and diabetes from a population-based register in Yorkshire
1978-93. Arch Dis Child 1998; 78:435-8.
188 Dahlquist G, Kallen B. Mortality in childhood-onset type 1
diabetes: a population-based study. Diabetes Care 2005;
28:2381-7.
189 Edge JA, Ford-Ad ams ME, Dunger DB. Causes of death in
children with insulin- depend ent diab etes 1990-96. Arch Dis
Child 1999; 81:318-23.
190 DiMaio VJM. Sturn er WQ, Coe JI. Sudden and unexpected
deaths after the acute onset o f diabetes mellitus. J Foren sic
Sci 1977 ; 22: 147-5 1.
191 Rosenbloom Ai. Kohrman A. Sperling M. Classifica tion a nd
diagnosis of diab etes mellitus in children and adolescents.
Joul11al of Pediatrics 1981; 99:320-3.
192 Coe JI. Post-mortem bioch em istry of bl ood and vitreous
humour in paediatri c practi ce. In Mas on JK (ed.) Paediatric
Forensic Medicine and Pa thology. London: Cambridge
Universi ty Press, 1989, pp. 191-203.
193 Olpin SE. The metabo lic investigation of sudd en infant
death. A/1I1als of Clinical Biochemistry 2004; 41:
282 -93.
194 John WG . Scott KWM, Hawcroft DM. Glycated haemoglobin
an d glycated protein and glucose concentrations in necropsy
blood samples. J Clin Pa thol 1988 ; 41:415-18.
195 Rozin L, Perper JA. Jaffe R, Drash A. Sudden unexp ected
death in childhood due to unsuspected diab etes mellitus.
Am J Forensic Med Pa thol 1994; 15:25 1-6.
196 Tattersall RB. Gill Gv. Unexplained deaths of type 1 diabetic
patients. Diabetic Medicin e 1991; 8:49-58.
197 Thordarso n H, Sovik O. Dea d in bed syndrome in young
diabetic patients in Norway. Diabetic Medicine 1995;
12:782-7.
198 Matyka K, Ford-Adams M. Dunge r DB. Hy poglycaemi a and
co unterregulation during childh ood. Horm Res 2002;
57(Suppl. 1):85-90.
199 Emery JL. Howat AJ, Variend S. Vawter GF. Investigat ion of
inborn errors of metabolism in un expec ted infant deaths.
Lancet 1988; ii: 29 -31.
200 Pollitt RJ. Disord ers of mitochondrial long -cha in fatty acid
oxidation. J Inh erited Metob Dis 1995; 18:473-90.
201 Pollitt RJ. Inherited disorders of straight chain fatty ac id
oxidation. Arch Dis Child 1987; 62:6-7.
202 Anon. Sudden infant death and inherited disorders of fat
oxidation. Lan cet 8 Nov ember 1986; ii: 1073- 5.
203 Benn ett MJ, Hale DE, Coates PM , Sta nley CA. Postmortem
recognition of fatty acid oxidation disord ers. Pediatr Pathol
1991; 11:365-70.
204 Largilliere C, Vianey-Saban C, Fontaine M et al.
Mitochondrial very long chain acyl-CoA dehydrogenase
defiCiency - a new disorder of fatty acid oxidation. Arch Dis
Child 1995; 73:FI03-S.
205 Olpin SE. Allen J. Bonham JR et al. Features of Carnitine
Palmitoyltransferase Type I Disease. J Inherit lvIetab Dis
2001; 24:35-42.
206 Chace DH. DiPerna JC, Mitch ell BL et al. Electrospray tandem
mass spectrom etry for analysis of acylcarnitines in dried
postmortem blood specimens collected at autopsy from infa nts
with unexp lained cause of death. Clin Chem 2001 ; 47:JJ66-82.
207 Touma EH. Charpentier C. Med ium chain acyl-C oA
dehydrogenase deficiency. Arch Dis Ch ild 1992 ; 67:142-5.
208 Roe CR. Millington OS. Maltby DA. Kinnebrew P. Recognition
of medium-chain acyl-CoA dehydrogenase deficiency in
asymptomatic siblings of children dy ing of sudden infant
death or Reye-like syndromes. J Pediatr 1986; 100:13-18.
209 Anon. Med ium chain acyl CoA dehydrogenase deficiency.
Lancet J991; 338:544-5.
210 Shortland G. Besley G, Bonham J et al. Newborn screening for
medium chain acyl CoA dehydrogen ase deficiency (MCADD):
Find ings from a multicentre prospective UK collaborative
study. J Inherit Metab Dis 2006; 29(Suppl. 1): 19.
2JJ Losty He. Lee P, AI Faham M et al. Fatty infiltration in the
liver in medium chain acyl CoA dehyd rogenase deficiency.
Arch Dis Child 199 1; 66:727-8.
212 Bove KE. Letter to the Editor. Pediatr Patlwl 199 2; 12:621-5.
213 Guenthard J. Wyl er F. Fowler B, Baumgartner R.
Cardiomyopathy in respiratory chain disorders. Arch Dis
Child 1995; 72: 223 -6.
214 Smeitink JAM. Fischer JC. Ruitenbeek W et al. Sudden infant
death associated with defect ive oxidative ph osp horylation.
Lancet 199 3; 341: 1601.
215 Konstantopouloll V, Sperl W. Wohl gena nnt J et al. Nea r
missed sudden unexpected infant dea th (SUID) as the firs t
sign of a respiratory chain co mplex I deficiency. J Inh eri t
lvfetab Dis 2001; 24(SuppL 1):78.
216 Tripp ME. Katcher ML. Peters HA et al. Systemic carnitine
deficiency presenting as familial endocardial fibroelastosis.
N El1gl J Med 1981; 305:385-90.
217 Steihm ER. Damrosch DS. Factors in the prog nosis of
meningococcal infection. J Pedialr 1966; 68:457-67.
218 Marzou k 0, Thoms on APJ, Sills JA et al. Features and
outcome in meningococca l disease presenting with
maculopapul ar rash. Arch Dis Child 1991; 66:485-7.
219 Morl ey SL, Lev in M. Bacterial meningitis. Prescr J 1998;
38:129-4l.
220 Sh arief N, Kh an K, Conlan P. Overwhelming sepsis
presenting as sudden unexpected death. Arch Dis Child
199 3; 69:381-3.
22 1 Dyke MP, Martin RP, Berry PJ. Septicaemia and adrenal
ha emo rrhage in congenital aspl enia. Arch Dis Child 1991;
66: 636 -7.
222 Larkin SM, Williams ON, Osterholm MT et al. Tox ic shock
syndrome: clinical. laboratory, an d pathologic findings in
nine fatal cases. Annals of internrillvfedicine 1982 ;
96:858-64.
223 Con e LA. Woodard DR. Schlievert PM. Tomory GS. Clinical
and bacteriologic observations of a toxi c shock-like
syndrome due to Streptococcus pyogenes. N Engl J Mer!
1987; 317:146-9.
224 Steve ns DL. Tanner MH . Winship J et al. Severe g roup A
streptococcal infection s associated with a toxic shock-like
syndrome and scarlet fever toxin A. N Eng l J Med 1989;
321 :1-7.
225 Stamenkovic I. Lew PD. Earl y recognition of potentially fatal
necrotizing fasciitis. N Engl J lvled 1984 ; 310:1689-93.

226 Chelsom J, Ha lstensen A. Haga T, H0iby EA. Necro tisi ng
fasciitis due to g roup A streptococci in western Norway:
incidence and cI inical features. Lancet 1994: 344: II J 1-15.
227 Tyrrell GJ, Lovgren M, Kress B, Grimsrud K. Varicella­
associated invasive group A streptococcal disease in Alberta,
Canada - 2000-2002. Ciin Infect Dis 2005; 40: 105 5-7.
228 Vugia DJ, Peterson CL, Meyers HB et aJ. Invasive group
A streptococcal infections in children with varicella
in Southern California. Pediatr Infect Dis J 1995;
15:146-50.
229 Sawin RS , Schaller RT, Tapper D et al. Early recognition of
neonatal abdomina l wall necrotizing fasciitis. Am J Surg
1994; 167:481 -4.
230 Stocker JT. Pathologic features of long-standing
'healed' bron chopulmonaty dysp lasia. Hum Pathol 1986 ;
17:943-61.
231 Miller RW, Woo P, Kellman RK. Tracheobronchial
abnormalities in infants with bronchopulmonary dysplasia.
J Paediatr 1987; 111 :779-82.
232 Greenholz SK, Hall RJ, Lilly JR, Shikes RH. Surgical
implications of bronchopulmonary dysplasia. J Pedia tr Surg
1987; 22:1132-6.
233 Doull lJ, Mok Q, Tasker RC. Tracheobronchomalacia in pre­
term infants with ch ronic lung disease. Arch Dis Child 1997;
76:F2OJ- 5.
234 Tammela OKT, Koivisto ME. A I-year foHow-up of lo w birth
weight infants with and without bronchopulmonalY
dysp lasia: hea lth , growth, clinical lung disease,
cardiovascular and neurological sequelae. Early Hum Dev
1992 ; 30: 109-20.
References I 255
235 Reardon W, Newcombe R, Fenton I et aJ. The natural history
of congenital myotoniC dystrophy: mortality and long term
clinical aspects. Arch Dis Child 1993; 68:177-81.
236 Thomas PK, Caine DB, Elliott CF. X-linked scapuloperoneal
syndrome. J Neurol Neuroslirg Psychiatry 1972; 35:208-1 5.
237 Primhak RA, Spicer RD, Variend S. Sudden death after minor
abdominal trauma: an unusual presentation of
phaeochromocytoma. Br Med J 1986; 292:95-6.
238 O'Neill D, O'Grady J, Variend S. Child fatality associated
with pathological features of histiocytic necrotizing
lymphadenitis (Kikuchi-Fujimoto Disease). Pediatr Pafhol
Lab Med 1998; 18 :79-88.
239 Morens OM , Hammar SL, Heicher DA. Idiopathic acute
pancreatitis in Children. Am J Dis Child 1974; 128:401-4.
240 Wilcken B, Hammond J, Silink M. Morbidity and mOltality
in medium chain acyl coenzyme A dehydrogenase
deficiency. Arch Dis Child 1994; 70:410-12.
241 Traisman ES, Young S, Lifschultz BD et al. Sudden death in a
neonatal as a result of Herpes simplex infection. J Forensic
Sci 1988; 33:267-71.
242 Mercuri E, Cowan F, Rutherford M et aJ. Ischaemic and
haemorrhagic brain lesions in newborns with seizures and
normal apgar scores. Arch Dis Child 1995; 73 :F67-74.
243 Norman MG, Taylor GP, Clarke LA. Sudden, unexpected,
natural death in childh ood. Pediat,. Pafho11990; 10:769-84.
244 Variend S. Infant mortality, microglial nodules and parotid
CMY-type inclusions. Ear/y Hum Del) 1990; 21:31-40.
245 Variend S, O'Neill D, Arnold P. The possible significance of
cytomegaloviral parotiti s in infant and early childhood
deaths. Arch Pat/wi Lab Med 1997; 121 :1272-6.
I CHAPTER 13 I
RECENT ADVANCES IN PAEDIATRIC
TOXICOLOGY
Patrice Mangin and Christian Giraud
Scope of the problem 256
Specificity of paediatric toxicology 258
Techniques used in drug testing 259
Special techniques for analysis of volatile substances 262
Alternative specimens for drug testing 262
Pitfalls and limitations of drug screens 267
During the past decade paediatric toxicology has substan­
tially advanced in line with all other branches of medical
toxicology; such advances should be considered in the
context of general forensic toxicology, with particular
attention being directed at the various aspects of the sub­
ject that a re so mewhat unique to the paediatric age group.
In this respect, this chapter focuses on several important
areas of forensic toxicology that have undergone signifi­
cant changes in the past decades, but which also are of par­
ticular interest and specific importance to paediatrici ans
and pathologists involved in a medicolegal or forensic
practice.
After a brief insight into the major issues raised during
the last few years in this specialty, this chapter will outline
what the paediatrician and/or the forensic pathologist can
expect to learn from toxicological investigations in the
light of the recent advances in analytical toxicology with
reference to drug testing, and the circumstances in which it
may be app ropriate to initiate laboratory tests looking for
drug use.
SCOPE OF THE PROBLEM
The age distribution of children involved in poisonings is
bimodaL I Ingestion of, or exposure to, toxic substances by
children before the age of 5 is almost invariably accidental,
with the peak of instances occurring in the toddler years.2
Specific applications 271
The importance of paediatric toxicology in
specific cases 274
Conclusions and future considerations in forensic
paediatric toxicology 274
References 275
Poisoning agents are usually those found in the home and
may include household chemicals, such as cleansers, hydro­
carbon fuels, paints and thinners, plants and fungi, and
medications that belong to other household members. 3 Poi­
soning appears to be frequent in this age group: according
to the Toxic Exposure Surveillance System compiled by the
American Association of Poison Control Centers, 50 per
cent of poisoning cases involved children under 6 years of
age. In contrast, fatalities in this age group are uncommon,
accounting for only about 4 per cent of all poisoning fatal­
ities. 4 As a result, in any given year, the probability of a
forensic pathologist or toxicologist encountering paediatric
patients in whom drugs or poisons can be detected is quite
lowS Even in these cases, the exact role played by the
drug(s) or substance(s) in the death of the child is not
always clear; indeed , the presence of the drug or poison
may be categorized as either causing or contributing to
death by direct toxic effect, or by idiosyncratic or hyper­
sensitivity reaction, or through drug-induced disease.
The second peak of incidence for poisoning occurs in the
adolescent years, when drug ingestion is more common.
Teenagers are more likely to ingest dangerous substances
intentionally and, in this respect, they generally choose
pharmaceutica l agents. 6 - 10 Within the teenage age group,
girls are more likely than boys to ingest drugs. Agents
involved in adolescent poisonings are often those found
in the family medicine cabinet, which include aspIrIn,
paracetamol (acetaminophen), iron supplements (ferrous

Table 13.1 Substances that are lethal in young children, even in small doses, as reported by Criddle lO
Benzocaine Camphor
Beta-blockers Clonidine
Calcium channel blockers Cocaine
Table 13.2 Substances most frequently involved in fatal
exposures in adolescents (7 3- 79 years old) as reported by Criddle lO
for 2004
1. Analgesics
2. Sedatives, hypnotics, antipsychotics
3. Antidepressants
4. Stimulants and other street drugs
5. Cardiovascular drugs
6. Alcohol and volatile compounds
7. Gases and fumes
8. Anticonvulsants
9. Chemicals
10. Muscle relaxants
sulphate), antidepressant medications, non-steroidal anti­
inflammatory drugs (NSAlDs), antipsychotic drugs, drugs
of abuse and recreational drugs. Inhalants, generally
volatile hydrocarbons and carbon monoxide, must also be
considered as well as caustic agents. In this age group,
fatalities remain rare, with poisoning accounting for less
than 10 per cent of the total fatalities. 4 The 2006 annual
report of the Swiss Toxicological Information Centre
(www.toxi.ch) indicates that the highest number of caJis
received for human poisoning involved children under
5 years of age (41.4 per cent). However, the proportion of
cases with potential health risk was somewhat higher in
adults (35.7 per cent) than in children (21.2 per cent). Boys
were more frequently represented among the children, and
women were more frequently represented among the ado­
lescents and adults. Acute accidental intoxications repre­
sented the largest group, with children ingesting easily
accessible household products, pharmaceuticals or plant
parts. Acute intentional poisoning was mostly due to
attempted suicide and less frequently to drug abuse or
criminal behaviour.
Small amounts of some drugs can be very toxic to
young children; one pill or one mouthful can require
aggressive treatment. Table 13.1 lists 12 toxic compounds
or classes of substances which are known as 'one pill can
kill' substances. 10, II Table 13.2 lists the 10 leading sub­
stances responsible for teen fatalities.
There is a growing concern in the medical community
over the maternal use of illicit drugs during pregnancy,
since intra-uterine exposure to drugs of abuse and nicotine
Scope of the problem I 257
Oi phenoxylate/atropine Quinidine
Lindane Su Ifonylu reas
Methyl salicylate Tricyclic antidepressants
is associated with a high incidence of perinatal complica­
tions.12,13 Approximately 30 per cent of the substance­
addicted population in the USA is female, and most are
women of child-bearing age. Among pregnant women
aged 15-44 years, 4.0 per cent had used illicit drugs and
1L8 per cent had used alcohol within 1 month of the 2006
National Survey on Drug Use and Health. The complica­
tions in pregnant women include an increased incidence of
still birth related to abruptio placentae and placenta prae­
via; a reduction in birth weight, birth length and Apgar
scores; intra-uterine growth retardation; and pre-term
birth. 14 - 16 After bilih, withdrawal symptoms including
hyperactivity, coarse tremor, poor feeding and poor weight
gain are present in almost all infants of heroin-, cocaine­
or amphetamine-addicted mothers. 17 Sudden infant death
syndrome (SIDS) is more common among infants of smok­
ing mothers. IS, 19 In addition, children born to smoking
mothers develop more slowly, both physically and men­
tally, through their teen years.20
Breast-feeding mothers constitute another group in
whom there may be a risk of infantile poisoning. Because
many drugs, especially lipophilic substances, are excreted
into milk, infants are exposed to adverse effects of these
drugs if breast fed. During lactation, multiple situations can
arise that require maternal pharmacological treatment.
Generally, paracetamol (acetaminophen), low-dose aspirin
(acetylsalicylic acid) (up to 100 mg/day) and short-term
treatment with other NSAlDs, codeine, morphine and
propoxyphene are considered compatible with breast­
feeding and lacking significant danger. 21 However, recently,
codeine intake in a breast-feeding mother resulted in a
baby's death. Genotype analysis for cytochrome P450 2D6
(CYP2D6) indicated that the mother was classified as a
codeine ultrarapid metabolizer. The clinical consequence
was a relatively high morphine blood concentration in the
baby resulting from metabolism of morphine, leading to
central nervous system depressant effect and neonatal
death. 22 Diazepam and its metabolite N-desmethyldiazepam
enter breast milk. The accumulation of diazepam in breast­
fed babies whose mothers are taking the drug may result in
infantile lethargy and weight 10ss.23 Drugs and substances
that require a careful assessment of risk before prescription
to breast-feeding women have been published by Ito in
2000. The same author listed the main drugs that are pre­
ferred for women who are breast-feeding. 24
Attention deficit hyperactivity disorder (ADHD) is the
most common neurobehavioural disorder of childhood that
258 I Recent advances in paediatric toxicology
affects school-aged children, wi t h a prevalence ranging
between 4 per cent and 12 per cent of the general popula­
tion. 25 The traditional therapeutic approach involves the
use of psychostimulants such as methylphenidate. As ther­
apeutic use of methylphenidate has increased, so the risk of
accidental intake by to ddlers and of abuse, misuse or use in
suicide attempts by pre-teenagers and adolescents has
become greater. Many cases of poisoning or abuse have
been reported in the scientific 1iterature. 26 ,27
The use of anabolic steroids among teenagers and
young adults is becoming increasingly popul ar, particu­
larly among those involved in sports requiring great
strength, such as weight-lifting and body-building. 28,29 All
anabolic steroids are derivatives of the male hormone
testosterone and all are intended for 'tissue building' and to
increase masculini zing effects, Reported psychological
side-effects associated with their use include aggression,
psychosis and changes in libido, The real incidence of ana­
bolic steroid consumption is hard to evaluate, but the
res ults of the National Household Survey on Drug Abuse
indicate that more than 1 million Americans are current or
former users.30 In Germany, the estimated number of juve­
nile users is about 100000.3J In the USA, 2 per cent of
college-aged men have been estimated to use anabolic
androgenic steroids. In anonymous US high school sur­
veys, as many as 5-6 per cent of adolescent male athletes
and 1-2 per cent of adolescent female athletes have
reported experimenting with anabolic steroids in conjunc­
tion with weight training to improve sports perform­
ance. 32 ,33 In this context, forensic medicine is involved in
several different areas. The forensic pathologist examines
doping-associated deaths. The forensic psychiatrist and
forensic physician deal with affective and mood disorders
including the results of aggressiveness and Criminality, and
problems of dependence on anabolic steroids.
It is perfectly legitimate to raise concerns about the use­
fulness of toxicological investigations if poisoning is sus­
pected in a paediatric case. Indeed, recent literature has
confirmed the conclusion that, in most cases, neither com­
prehensive toxicology nor drugs-of-abuse screening of
poisoned patients significantly influences their clinical
management. 34 However, these analyses are useful for
diagnostic confirmation and are a prerequisite in certain
forensic investigations, The utility of focused quantitative
serum assays to determine serum levels of particular
poisons is more helpful to paediatric clinicians. 35 In a retro­
spective study of all comprehensive emergency department
toxicology screenings performed in paediatric patients, 463
cases were reviewed by Belson et al. 36 In this study, more
than 550 toxins were screened, Of 234 positive screening
tests, only seven were positive without a documented sus­
picion of an exposure. In these cases, detection of these
specific drugs did not result in a change in medical man­
agement a nd their presence did not affect the patient's
clinical outcome. However, toxicological analyses are jus­
tified, particularly if the suspected pOison (e.g. carbon
monoxide, CO) can endanger the life of other people.
Infants and children are at greater risk of CO poisoning
than adults. Fetuses are particula rly vulnerable, because
maternal CO crosses the placenta. 37 The rapid identification
of the toxic compound may also help in selecting the
appropriate antidote, especially if it is potentially toxic. For
example, the standard antidote for acetaminophen (para­
cetamol) toxicity is N-acetylcysteine (NAC) . Intravenous
administration of NAC may induce anaphylactoid reac­
tions. J8 In a few rare cases, administration of the false anti­
dote may therefore increase the symptoms of toxicity.
SPECIFICITY OF PAEDIATRIC TOXICOLOGY
Absorption of a drug from the gastrointestinal tract is
slower in infants than in older children an d adults.J9 The
skin of infants is more permeable to drug diffusion than in
adults, and the blood-brain barrier of infants is also more
readily penetrated. 4o Enzyme systems involved in drug
metabolism are usually functionally immature at birth,
leading to prolonged elimination half-times for drugs. The
oxidation P450 and conjugation glucuronidation systems
require several weeks after birth to reach maturity. In
contrast, the sulphation pathway is more mature and may
partially compensate for the deficiencies of the other meta ­
bolic pathways41 For multiple enzyme families, a'develop­
mental switch' is observed that characterizes the transition
between predominant fetal enzyme form to the predomi­
nant adult enzyme form. For example, CYF3A7 expression
dominates in the fetus whereas hepatic CYF3A4 expression
dominates in the adult. 42 Differences in pharmacokinetics
between newborn children and adults are presented by
Soldin and Steele,43 Soldin and Soldin 44 and Alcorn and
McNamara. 45 There are also important pharmacokinetic
differences between children and adults that influence drug
toxicity: (1) the fat compartment of children is generally
smaller, providing less storage for lipid-soluble drugs; (2)
the propoliion of unbound drugs in the blood is higher in
children, indicating that more drugs are available to pro­
duce a pharmacological effect; and (3) the ratio of liver
weight to total body weight in children is 50 per cent
higher at 2 years old and 30 per cent higher at 6 years old
than for the adult. Drug biotransformation rates are therefore
enhanced in children and reduced in neonates compared
with adults. In addition to pharmacokinetics differences,
pharmacodynamics responses also differ. 46
A final point concerns the marketing of therapeutic sub­
stances. Only a small minority of approved drugs have
received adequate paediatric study in appropriate clinical
trialsY Furthermore, a licensed formulation for the admin­
istration of certain pharmaceutical products to children is
often also lacking.48 The situation could be even worse
with off-label/unlicensed drugs. Indeed, many cases of
adverse events involving off-label/unlicensed prescriptions
have been reported. 49 To fill these gaps, new regulations

and patent protections have been issued for new drugs by
the US Food and Drug Administration, requiring specific
paediatric studies.
Another point that deserves consid eration rel ates to
the clinical research involving children and therapeutic
drugs. Several studies of drugs that bring only very minor
therapeutic benefits to neonates and children are ethically
questionable. 50 ,51
TECHNIQUES USED 11\1 DRUG TESTING
Analysis of biological specimens is the necessary step that
determines whether or not an individual has been intoxi­
cated with illegal drugs, with excessive amounts of prescrip­
tion drugs or with any other toxic substances. It is important
to remember that toxicologists, especially in paediatric
rather than adult cases, often do not know what to look for
and therefore have to take into account a vast number of
toxicologically relevant substances, which makes the task
difficult but very challenging. 52.10
Drug Screening
The technology of drug screening has greatly improved
throughout the last two deca des. Idea lly, a screening test
for a drug should be rapid (a short turnaround time). highly
specific (without false-positive results), highly sensitive
(allowing the detection of a very low concentration of the
drug in the specimen), reliable (same result from one lab o­
ratory to another for a given specimen), easily performed
technically and inexpensive. Since none of th e available
analytical technology meets all of these requirements per­
fectly, drug screening is usually achieved by means of the
combination of two different techniques; this is done in
order either to extend the ambit of the screening test to a
larger group of substances not detectable by the other
method or to confirm the result from another ana lysis. In
practice, screening tests are design ed for max imum sensi­
tivity at the expense of selectivity. Confirmatory analyses
are often just as sensitive (if not more so) but additionally
provide the requisit e specificity. It is important, however, to
point out that screening procedures must not give false­
negative results since this inva lidates the testing process,
False-positives at this stage are acceptable because all pre­
sumptive samples wiU subsequently be reanalysed by a
confirmatory technique. 52
The methodologies most frequently employed for drug
screening are immuno assays and chromatography, and
sometimes capillary electrophoresis. Serum (or blood).
urin e and oral fluid (OF) (sa liva) can be analysed by these
techniques. The pharmacokinetics of distribution and eJ im­
inati on of dru gs are such that most are detectable in urine
for a much longer time tha n in blood or OF so that urine is
usually the specimen of choice for screening.
Techniques used in drug testing I 259
IMMUNOASSAY TECHNIQUES
Immunoassays are very useful to filter out the negative
samples from the positives, a nd thereby reduce the amount
of further analytical work required for toxicological inves­
tigation. 53 Immunoassay techniques are now the most
widely used screening pro ce dures in drug testing. Advan­
tages include high sensitivity, no extraction steps and
applicability to a large number of samples. Popular ver­
sions of these tests are represented by radioimmunoassay
(RIA). enzyme-multiplied immunoassay technique (EMIT),
enzyme-linked immunosorbent assay (ELISA) and fluores­
cent polarization immunoassay (FPIA),s4 These techniques
use antibodies specific to the drug being assayed and a
labelled form of the same drug, The label itself may be a
radio active isotope (RIA). an active enzyme (EMIT and
ELISA) or a fluorescent label (FPIA), which is incorporated
synthetically. A fixed quantity of antibody and labelled
drug are added to the test sample. The binding sites on the
antibody attract both the labelled drug type and the unla­
belled drug type in the sample. The amount of labelled drug
bound is inversely proportional to the number of unla­
belled drug molecules present. 53,55
In RIA, a rad ioactively labelled drug competes fo r the
same antibody binding site as the unlabelled drug. The ana­
lytical measurement of radioactivity remaining in the solu­
tion determines the amount of unlabelled drug in the
sample. RIA kits were ava ilable for several classes of drugs
of abuse but now have been virtually abandoned, being
replaced by non-radioactive immunoass ays. For instance,
some larger companies stopped their marketing of most RIA
kits during the period 2004-2006.
In EMIT, ELISA and FPIA testing, the analytical measure­
ment is based on an optically detected change, such as ultra­
violet (UV) absorption, fluorescence or luminescence. These
systems avoid the use of radioactive isotopes but also have
reduced sensitivity compared with RlAs because the optical
signal is measured in the presence of the original biological
fluid. An EMIT kit is available for opiates, barbiturates,
cocaine, amphetamines, benzodiazepines, methaqualone,
methadone, phencyclidine, cannabinoids and lysergic acid
diethylamide (LSD) . The advantages of EMIT as a screening
technique include its rap idity, its semiquantitation and its
ability to detect a number of drugs. Disadvantages include
the high cost of the reagents and its susceptibility to both
false-positives and false-negatives so that confirmation by
an alternative methodology is recommended for positive and
even for negative results. 56 Microplate ELISA tests are also
competitive immunoassays. They are speCifically optimized
for use with a number of different biological matrices. Kits
exist for about 20 pharmaceuticals and 'drugs of abuse'.57 An
FPIA is both sensitive and specific; however, the equipment
is expensive and susceptible to sample alteration, affecting
its accuracy. The application of the FPIA technique to foren­
sic blood work and to alternative specimens such as hair or
bile has been described .58 Additional antibody kits are also
260 I Recent advances in paediatric toxicology
available, expanding the uses of these systems. 55 New
immunoassays are available to test the OF from drivers at
the roadside and workers at the workplace. For instance, the
Cozaft® drug detection system (DDS) test is used to detect
impaired drivers suspected of driving under the influence of
psychoactive substances. DrugWipe® 5+, developed by
Securetec, is another test evaluated for the detection of psy­
choactive drugs in the oral fluid of suspected impaired driv­
ers. Improvements are still needed for the similar detection
of cannabinoids and benzodiazepines.
SAMPLE PREPARATION
Sample preparation is an important step, which still
remains the most tedious and time-consuming part of an
analytical method. Optimum sample preparation leads to
enhanced sensitivity and selectivity while reducing the
amounts of interfering matrix components and increasing
analyte concentration. Sample preparation represents a
major part of analysis that is capable of taking up to 80 per
cent of the total time of a complete separation-based ana­
lytical process, which typically encompasses five steps:
sampling, sample preparation, separation, detection and
data analysis. Major techniques rely on liquid-liquid,
solid-phase and headspace extraction. Recent advances
have been made in miniaturization, high throughput, selec­
tivity, integration and hyphenation and automation of
sample preparation.
Liquid-liquid extraction (LLE) transfers the target ana­
Iytes from a liquid matrix into another immiscible liquid
phase according to solubility difference. Solid-phase
extraction (SPE) is used to extract and concentrate analytes
from a liquid matrix by partitioning the compounds
between a solid and a liquid phase. An SPE procedure con­
sists of four consecutive steps: column conditioning, sam­
ple loading, column washing, and elution of selected
analytes. The headspace technique is dedicated to the
extraction of volatile compounds. The analyte is parti­
tioned between the sample and the gas phase in a closed
system. Then, the vial is pressurized and the headspace is
sampled and injected into the gas chromatograph. In recent
years, selective extraction methods with molecularly
imprinted polymers (MIPs) and affinity columns with
bound antibodies have also been considered. Template
molecules are used to produce MIPs by creating cavities in
a polymer that will recognize the target molecule. 59 The
MIPs somehow mimic natural antibodies. Miniaturiza­
tion of the SPE methods resulted in the development of
solid-phase microextraction (SPME). Other materials cur­
rently under investigation are restricted-access materials
(RAMs).5o The RAMs are used to exclude large molecules,
such as proteins, and to extract low-molecular-mass ana­
Iytes by use of hydrophobic, ionic or affinity interactions,
typically from complex blood matrix. They allow the direct
injection of plasma or serum samples on to a high­
performance liquid chromatography (HPLC) column without
rapid clogging of the column, facilitat-ing sample prepara­
tion and analysis coupling and automation. 51 .52
HIGH-PERFORMANCE LI QUID CHROMATO GRAPHY
The separation of non-volatile substances from each other
or from other components of an extraction residue can be
obtained using HPLC. When a mixture of substances is
injected onto the column, each component is partitioned
between the stationary phase (column) and the liquid
(mobile) phase. Molecules with greater affinity for the col­
umn spend more time in that phase and, therefore, take
longer to reach the detector. The time taken from injection
to the peak maximum is known as the retention time. The
detector responds in direct proportion to the concentration
of material passing through it, hence peak heights and area
shown on the chromatogram are directly related to the
concentration of each analyte.
Good sensitivity and high specificity, depending upon
the detection system used, are shown by HPLC. The most
common detectors utilize absorption of UV light by the
drug. A diode array detector (DAD) will allow a full UV
spectrum of the analyte to be obtained. This can be com­
pared to a standard spectrum of the drug, and identifica­
tion is then based on both retention time and UV
spectrum. The identification of a very broad number of
unknown substances in blood specimens can be obtained
using HPLC-DAD in combination with a suitable spec­
trum. 53 .54 High selectivity and sensitivity for molecules
with characteristic excitation and emission fluorescence
spectra is shown by HPLC coupled with fluorimetric detec­
tion. In this way, LSD and its metabolite nor-LSD can be
quantified in blood by HPLC fluorescence detection down
to 20 pg/ml. 55
Pre- and post-column reactions for increased sensitivity
have also been described. Owing to its qualities, HPLC can
be used to screen a sample for many drugs of abuse. The
drawbacks of HPLC include its expense, the high degree of
expertise necessary for operation of the equipment, the
need for sample extraction and the need for preparation of
separate specific columns. 55 New filling column materials
have been marketed over the last few years. For example,
polar metabolites, such as ecgonine or hydroxyl metabo­
lites of cocaine, can be analysed without prior derivatiza­
tion through hydrophilic interaction chromatography
(HILlC).57 Similarly, the polar metabolites of morphine
(morphine 3- and 6-glucuronide) can be readily analysed
on reversed-phase columns, which are compatible with
almost 100 per cent water eluent. 58 More recent develop­
ments in HPLC concern the use of sub-2 p.m particles and
mobile phases delivered at high flow rates. The recent com­
mercialization of porous hybrid organic-inorganic silicon­
based paliicles with a narrow size distribution in the range
of 1.7 pm has enabled a new level of performance, but only
through the use of newly developed pumps that permit
pumping and injection of liquids at pressures in excess of

10 000 pSi. Full implementation of such technology, termed
ul tra performance liquid chromatography (UPLC), is further
challenged by the requirement for new detectors fast
enough to record a minimum number of data-points in a
very narrow chromatographic peak (width of a few seconds
only).69.70 The use of UPLC columns speeds up analyti cal
runs to times of as low as a few minutes for complex mix­
t ures. Another way to reduce the overall analysis time is to
use monolithic supports, which co nsist of a continuous
porous silica rod. Separations with monolithic columns can
be performed at high flow-rat es, reducing the analysis time
to less than 10 min.7!
GAS CHROMATOGRAPHY
Gas chromatography (GC) is another method of separating
substances of analytical interest. 66 It is one of the most
efficient techniques available for separating drugs in body
fluids. The separation is carried out on an analytical co l­
umn containing a station ary phase (liquid or solid, depend­
ing on temperature), which is maintained at a given
temperature inside an oven. The whole GC system com ­
prises six components: gas supply and flow controllers,
injector, oven, column, detector and recording device. In
drug testing, GC capillary (rather than packed) co lumns are
co mmonly used . A com pound is identified by matching its
retention time with that of a drug standard under the same
cond itions. The use of more sens itive and selective detec­
tors than flame ionizat ion detectors, such as electron cap ­
ture and nitrogen-phosphorus detectors, has greatly
improved the reliabili ty of Gc. The method of choice for
volatile compounds a nalysis is GC, which can be easily
coupled with mass spectrometlY detectors. Unfortunately,
many drugs and poisons conta in polar functional groups
and require chemica l derivatization to improve thermal
stability, volatility and detection. 72 Other dra wbacks include
the time necessary to prepare the specimen (extraction,
purification and derivatization), the expense of the equip­
ment and the requisite expertise of GC technicians. There­
fore, GC is usu ally reserved for use as a screening and
confirmatory technique in the expert laboratory setting.
Fast GC is a new development in GC technology. Fast GC
allows rapid analyses and reduction of retention time (up to
X 10) while main taini ng acceptable analyte reso lution. This
is made possible by reduction of column bore size, down to
100 !Lm, increasing the oven temperature ramp up to
120· C/m in, and by using high-pressure ca rrier gas control.
Because peaks a re very narrow, fast detectors and data sam­
pling are also required. Applications include analysis of
drugs of abuse in urine, 30 different drugs in oral fluid,73
benzodiazepines in blood 74 and cannabinoids in blood. 75
CAPILLARY ELECTROPHORESIS
During the past decade, capillary electrophoresis (CE)
emerged as a promising, effective and economic approach
Techniqu es used in drug testing I 261
for sepa ration of a large variety of substances, including
those encountered in forensi c toxicology.76 Capillary elec­
trophoresis comprises a family of electrokinetic separation
tech niques that separate co mpounds based upon differ­
ences in electrophoretic mobili ty, phase partitioning, iso­
electric point, molecular size or a combination of one or
severa l of these propeliies. Electroki netic capillary methods
employed for forensi c toxicology include capillary zone
electrophoresis (CZE) and an electrokinetic capillary chro ­
matography technique called micellar electrokinetic capil­
lary chromatography (MECC). In CZE, a few nanolitres of a
sample are applied to the beginning of a fused-silica capil­
lary filled with buffer. On app lication of a high-voltage
direct current field, charged solutes begin to separate and
are swept through the capi llary by the combined action of
electrophoresis and electro-os motic bulk flow and are
on-column detected towards the capillary end . In MECC,
the buffer contains charged micelles (dodecy l sulphate
micelles), and uncharged and charged solutes separate on
the basis of differential partitioning between the micelles
and the surrounding buffer and, if charged, also by differ­
ential charge effects, including electrophoresis. In both
techniques, the most commo n detection principles app lied are
on-column absorbance and fluorescence. Using on-column
multiwave len gth detection, this technology is well suited
for toxicological drug scree ning. Compared with HPLC and
GC, CE has distinct advantages, including automation ,
small sample size, minimal sample preparation , use of very
small amounts of organic solvents and low cost of capillary
columns. 77 ,78 Recent developments concern the co upling of
CE with tandem mass spectrometry and the analysis of
biofluids focused on less analysed matri ces, such as amni­
otic fluid, saliva, cerebrospinal fluid, sweat or airway sur­
face fluid and sputum. 79
Drug Confirmation
DlUg confirmatio n is based upon the use of two different
techniques, each confirming the other's results. In this
resp ect, the combination of two of the previously men­
tioned techniques adds confidence in the accuracy of the
resul t. 55,80
On the other hand, the introduction of the mass spec­
trometry detection method has revolutionized the analyti­
ca l tox ico logy in providing the capac ity of a formal
identification of unknown compounds. Gas chromatography­
mass spectrometry (GC-MS), combining GC with mass
spectrometry, has become the so - ca ll ed 'gold standard ' of
forensic drug testing and doping ana lysis. 8o - 82 In the elec­
tron impact ionization mode, the sepa rated compounds are
bombarded with high-energy electrons, causing them to
break apart. The fragments produced are separated on the
basis of their mass-charge ratio. Under the same condi­
tions, a molecule will fragment in exactly the same way
evelY time, producing the sa me spectrum of fragm ents.
262 I Recent advances in paediatric toxico logy
This spectrum, along with the chromatographic retention
times, constitutes a virtual 'fingerprint' that is very specific
for the particular chemical compound of interest. Detection
sensitivity may be as low as to measure the volume of the
drug in nanograms per millilitre. The definitive method for
drug identification is considered to be GC-MS.
Chemical ionization (CI), is a softer ionization method
that utilizes a charged reagent gas. Under these conditions
of ionization, fewer fragments are formed, increa sing the
chance of detecting the pseudomolecular ion. High-tlu'oughput
procedures mean that thousands of relevant toxic sub­
stances can be screened in a single procedure, i.e. the so­
called systematic toxicological analysis (STA).
Urine still remains the standard specimen for compre­
hensive screening, especially when a general screen for
unknown substa nces is required. To detect the largest num­
ber of xenobiotics , conjugated metabolites are first hydrol­
ysed and then derivatized prior to GC-MS analysis.
Other confirmatory techniques include HPLC coupled
with mass spectrometry (LC-MS); this constitutes a straight­
forward method for the analyses of polar or thermolabile
compounds without a derivatization step as required in
GC-MS. Gas chromatography or HPLC coupled with tan­
dem mass spectrometers (GC-MS-MS and LC-MS-MS)
enhance the selectivity and the sensitivity by a specific sec­
ondary fragmentation of selected ions provided by the first
fragmentation. All these sophisticated techniques, includ­
ing GC-MS, can offer valuable information for the identi­
fication of unknown compounds. Using these hyphenated
chromatography and mass spectrometry methods also allows
acute or chronic poisonin g to be definitively excluded, which
might be sometimes as important as the detection of a toxic
substance.
Time-of-flight mass spectrometers (qTOF-MS) accu­
rately determine the mass-charge ratios of ions by measur­
ing flight time after acceleration in a vacuum tube by a
high voltage. qTOF-MS highly improves mass resolution
and accuracy allowing exact mass determination. Accurate
mass measurement can be used for qualitative identifica­
tion of unknown molecules. Pavlic et a18 ) used ES1-qTOF­
MS with mass spectral libraries to identify over 300
different drugs. New and very expensive MS tools which
are emerging in toxicology laboratories are Fourier trans­
form ion cyclotron resonance mass spectrometers (FTMS)
and OrbitrapT". These are now considered as the ultimate
instruments for high mass resolution a nd accuracy84
Method Validation
Correct interpretation of toxicological findings is only pos­
sible provided reliable analytical data a re available. There­
fore, new analytical methods to be used in clinical and
forensic toxicology require careful method development
and thorough validation. For method validation, a whole
bunch of analytical parameters must be considered and
determined, namely the calibration model, the accuracy
and precision of the method, the low er limit of quantifica­
tion and the limit of detection. 8s .B6 Participation in external
quality control programmes allows laboratOlY performance
to be evaluatedY
SPECIAL TECHNIQUES FOR ANALYSIS OF
VOLATILE SUBSTANCES
Epidemiological studies indicate that a substa ntial number
of children and teenagers worldwide experiment with or
abuse volatile substances with the intention of experienc­
ing a euphoric state of consciousness. A large range of
vo latile compounds are toxic, especially for young people.
They can be classified in three broad classes: the hyd rocar­
bons (e.g. butane), oxygenated compounds (e.g. butanol,
amylnitrite) and halogenated compounds (e.g. halothane,
trichloroethylene). In case of suspicion of poisoning with
volatile substances, proper sample collection, storage and
handling are critical points to guarantee accurate toxico­
logical results. Samples should be taken rapidly and stored
in gas-tight, welI~sea led containers with minimal head­
space. Storage, transport and handling of t he sample
should always occur at low temperature (- 5°C up to 4°C).
The headspace technique is the most appropriate and
popular method for the extraction of volatile substances in
body fluids and tissues. Solid-phase microextraction, cryo­
genic oven trapping, cryogenic focusing, and purge-and­
trap extractio n techniques are typical headspace methods.
Usually, electron capture detection (ECD), flame ioniza­
tion detection (FlO) and mass spectrometry (MS) are used
for detection and quantification purposes, while MS and
Fourier transform infrared (FTIR) spectroscopy are used for
identification. 88
ALTERNATIVE SPECIMENS FOR DRUG TESTING
The rapid growth and development of drug testing technol­
ogy has created a number of testing methodologies . that
can assess a broad range of biological specimens, including
besides urine and blood, hair, oral fluid, sweat, meconium
and amniotic fluid. 89
Urine
For many reasons, urine is the most common matrix
analysed for drug testing. The specimen is easy to collect
and is not considered as an invasive sample to request.
Large volumes can often be collected , allowing extensive
screening and storage for any further analyses by other
laboratories if additional expertise is needed. The matrix is
one of the simplest that can be used, and it is the easiest to
analyse compared with other fluids or tissues. Drugs and

metabolites are usually stabl e in frozen urine, allowing
long-term storage of positive samples.
Drugs are usually metabolized by the liver to form polar
metabolites that are conjugated before their elimination
from the body via the urine. The metabolism of parent
drugs (which are chemically non-polar) is relatively rapid
compared with their urinary excretion. Conversely, the uri­
nary excretion of most metabolites (which usua lly are
chemically polar) is more rapid and extensive tha n with the
parent drugs. Drug metabolites are therefore frequently
found in urine in concentrations mu ch greater than the
parent drugs. Therefore, many immuno assay urine screen­
ing assays for drugs of abuse utilize antibodies to drug
metabolites.
Urine, however, can be easily adulterated since observed
specim en collection is not a common practice. 90 ,91 Addi­
tionally, most drugs are present in urine for only a few
days after consumption so that only recent drug use will be
detected.
The determination of drug-to-creatinine ratios takes
account of the potential dilution of urine. Dilution can result
from deliberate adulteration or through fluctuations in fluid
intake and elimination. Ulinary excretion profiles are nor­
malized and smoothed when reported to the creatinine con ­
centration. New intake of drugs can be better predicted by
comparing creatine-normalized drug concentrations meas­
ured in two successive Ulinary voids 92
Blood
Blood is co nsidered to be an invasi ve sample. It represents
a complex matrix particularly if haemolysed. The detec tion
window of the majority of drugs and metabolites to be
tested is limited from a few minutes to several hours at
most, and much less frequently days. However, the toxic
effects or the pharmacological activities of drugs are gen­
erally correlated with their blood concentrations, which
allows the effects of these levels on the living person to be
predicted, a nd thus the extent of involvement of the drug
in the death to be understood. 9J To assess the significance
of drug levels measured in blood, the concentrations are
compared with ranges of therapeutic, toxic and fatal levels
found in large databases. 94 ,95 These databases compile the
results of toxicological investigations of published case
reports. They are mainly based on tox icology data obtained
from adult patients or fatalities. It is assumed that these
compilations can be used to interpret drug levels in blood
specimens taken from young adults and teenagers,
However, extrapolation to the fetus, neonates and infants
is probably more hazardous and prone to error. In the
developing chi ld there are rapid changes occurring in terms
of organ maturation, changes in body composition and the
ontogeny of drug elimination,96
Genotype and the temporal acquisition of drug bio­
transformation are criti cal determinants of a drug response
Alternative specimens for drug testing I 263
in infants. For instance, if the O-demethylation of dextro­
methorphan (reflecting CYF2D6 activity) appears to be well
developed by 2 weeks post-natal age, the N-demethylation
(reflecting CYFJA4 and possibly also CYP2B6 activities)
increases more slowly over the first year of life. 97
All of these changes have a profound effect on the phar­
macokinetics of drugs across the paedi atric age range.
Because of the non-linear nature of these changes, simple
allometric scaling methods based on body weight or body
surface area often fail in the prediction of drug dosage,
especially in neonates and infants.98 Several sca lin g mod­
els in predicting maintenance doses for children from those
used in adults have been suggested. 99
Altered drug metabolism in children may result in adverse
drug reactions because of inappropriate high-dose adminis­
tration. Idiosyncratic adverse drug reactions may also occur.
For instance, hepatotoxicity after valproate administration is
much more frequent in children. The metabolism of va lproi c
acid in children is increased, resulting in the formation of a
higher amo unt of a toxic metabolite. In some cases, children
are more resistant to drug toxicity than adults. Compared
with adults, children appear to be more resistant to the hepa­
totoxic effects of paracetamol (acetaminophen) overdose. Sul­
phation, whic h is more active in children, may diminish the
formation of toxic metabolites by P450 enzymes. 2J Hence,
giving a dose extrapolated from adult dosage may result in
unexpected blood levels in paediabic patients. Therapeutic
ranges for so me drugs are quite different in adults and chil­
dren, e.g. therapeutic theophylline concentrations are lower
for neonates (5-15 mg/L) than for adults (10- 20 mg/L). 100
Protein binding can be quite different in neonates, producing
vely different therapeutic ranges for highly bound drugs such
as phenytoin.
The blood (pl asma) drug metabolite concentration (or the
ratio of parent drug and metabolite) can also be used to dif­
ferentiate between acute and chronic drug administration.
In the case of acute drug administration, the blood (plasma)
metabolite concentration in relation to parent drug concen­
tration can be helpful in determining the approximate
length of time between drug administration and death, or
specimen collection at autopsy or in a living person.
Neonates may also show different ratios of the pa rent drug
and metabolite, the hydroxyl metabolite of phenobarbitone
and its glucuronide conj ugate being present in much higher
concentrations in neonates than in older subjects. 100
Hair
Detection of dru gs and their metabolites in hair has gained
much attention over the past decade. Drugs incorporated in
hair remain in the keratin matrix for a long time, thus open­
ing a much wider window of detection than there is for
drugs in urine. 101-104 Reported drug recovery from the hair
of ancient Peruvian mummies suggests that hair analysis
may provide inform ation about drug use or exposure
264 I Recent advances in paediatric toxicology
almost indefinitely under the correct sample storage
conditions. l05 However, it has also been reported that a
small amount of the entrapped substances may slowly be
hydrolysed spontaneously. lOG
In addition to the promising diagnostic power of hair
analysis, hair sample collection is easy to perform without
the embarrassment commonly associated with urine collec­
tion. Hair samples can be stored or transported without
refrigeration, pH control or the preserving agents that are
normally needed for other biological samples, such as
blood and urine.
Hair may also serve as a 'diary' of exposure, producing
a drug history. 107 Scalp hair, in the average person, gener­
ally grows approximately 1 cm per month ; hair that is 3 cm
from the scalp would have been formed by the hair-forming
cells in the hair follicle approximately 3 months earlier.
Therefore, it is theoretically probable that a drug incorpo­
rated into the hair via the hair-forming cells in the hair
follicle would appear 3 cm from the scalp 3 months after
consumption. Unfortunately, the interpretation of drug pres­
ence in hair is not that straightforward. Not all drugs found
in the hair may enter through the hair follicle. Drugs may
also enter the hair shaft via blood, sweat or sebum, or by
diffusion from the skin surrounding the hair. 108
Hair pigmentation 109 and gender differences no also appear
to influence the concentration of drugs measured in hair.
Several investigations suggest that basic drugs bind to
melanin, which explains the higher concentrations gener­
ally found in dark hair. On the other hand, levels of acidic
drugs and metabolites, such as ll-nor-9-carboxy-delta
(9)-tetrahydrocannabinol, are unaffected by hair cOIOUr. III ,112
Our understanding about hair as a distribution site for
drugs is still limited, and much research is necessary before
drug concentrations in hair can be interpreted with accu­
racy.IJJ Although there is an increasing body of literature
being developed about the detection of drugs of abuse in
hair, there remain substantial and significant questions
about the interpretation of the analytical data.
Hair is somewhat unique among tissues in that
substances that are incorporated into hair may not have
subsequent access to the blood, and are thus retained long
after the parent drugs and metabolites have been elimi­
nated from the body. Although hair (particularly hair bulb)
does contain cytochrome P450 enzymes and has the ability
to metabolize drugs, it is considerably less able to do so
than most other organs. 107
Substances may become permanently bound in hair in
the form in which they are distributed into the hair follicle.
Cocaine appears to be an exception in that its hydrolysis to
benzoylecgonine is thought to occur in the hair shaft. 106
The blood supply to the hair follicle is extremely rich, and
the hair-forming cells of the hair follicle are among the most
rapidly dividing cells in the body. Drugs in the blood circula­
tion will be delivered rapidly to the hair follicle and, if they
are capable of crossing the cell membrane, will enter the hair­
forming cells. Parent drug concentrations in hair generally
exceed the concentrations of metabolites even when both
the parent drugs and their metabolites are no longer
detectable in plasma, blood or urine. 114 ,115 However, it is also
possible to detect the metabolites of drugs of abuse in hair,
which may provide proof of drug ingestion. II G
The finding of parent drugs in the hair may prove to be
an advantage in the use of hair testing by forensic toxicol­
ogists. For example, hair may become the matrix of choice
when differentiation between heroin use and prescription
use of morphine or codeine is required. IOI , IOJ
The finding of greater concentrations of parent drugs
than metabolites in hair may result from several mecha­
nisms. 1l 7 Parent drugs are generally less chemically polar,
more lipophilic and better able to cross cell membranes
than their metabolites. If substances enter the hair through
the hair-forming cells in the hair follicle, then the process
would favour the incorporation of parent drugs. Increased
lipophilicity may also favour parent drug secretion into
sweat and sebum, where it could come into contact with
the hair shaft distal to the hair-forming cells. In this
respect, Henderson et al 118 administered deuterium-labelled
cocaine to known cocaine users and found that in some
subjects the drug moved in a distinct band while in others
the labelled cocaine appeared first at a site more distal to
the scalp. These data suggest that in some individuals
cocaine may be secreted rapidly into the sweat or sebum,
and then deposited on the hair. It is also possible that the
components to which drugs bind in hair favour the binding
of parent drugs over metabolites , as has been shown for
codeine, which binds to hair in vitro with a greater affinity
than morphine.
Besides the polarity of drugs, it has been suggested that
other chemical characteristics should play an important
role for the incorporation of drugs into hair. In this respect,
Nakahara et al 11 9 and Nakahara and Kikura 120 have shown
that structural changes that increase the lipophilicity of a
compound increase its incorporation into hair (for exam­
ple, an increase in the length of the carbon chain on the
nitrogen of amphetamine), whereas structural changes that
decrease a drug's basicity will reduce its incorporation. On
the other hand, subtle structural changes that influence a
drug's lipophilicity and ionization to a cation or anion at
physiological pH greatly affect the incorporation of the
parent drug and its metabolite into hair. At a given plasma
concentration, drugs that form cations at physiological pH
appear to be incorporated into hair at greater concentra­
tions than drugs that form anions. The reason for this
observation is not known, but evidence suggests that it
may be a result of the binding with melanin (see below).
In addition to the impact of changes in the chemical
structure of drugs on their incorporation into hair, there is
also the possibility that the structure of the hair (fine ver­
sus coarse) or chemical treatments that affect hair structure
are important for the incorporation of drugs. Blank and
Kidwell 12 1 demonstrated that the absorption of cocaine into
cut hair specimens was greater for thick hair than for fine

hair. The hair specimens used in these experim ents were
black (thick hair) and brown (fine hair), and it is possible
that the difference in pigmentation also influenced the
absorption of cocaine. Cirimele et aj122 measured drug con­
centrations in the hair of a female drug addict who had
brown hair with strands of hair bleached with hydrogen
peroxide. They found approximately threefold greater con­
centrations of cocaine and codeine in the brown hair than
in the bleached hair from the same person.
The role of pigmentation in influencing hair concentra­
t ions has been clearly demonstrated for several drugs such as
cocaine, codeine, methadon e or nicotine. For example,
Joseph et al 123 showed that the specific binding of cocaine is
157 times greater in female African black hair than in female
Caucasian blonde hair. Furthermore, these authors showed
that bleaching black hair decreased the specific binding of
cocaine J3-fold. Nakahara et al l1 9 fo und a good correlation
between the in vitro melanin affinity and the incorporation
ratio for many drugs of abuse. Hold et al 124 showed that
stanazolol, an anabolic steroid, is incorporated to a greater
extent in pigmented than in non-pigmented hair.
If these data clearly show that pigmentation is an
important factor in the incorporation of drugs into hair, it
must be pointed out that drugs are also incorporated into
the hair of albino animals who have non-pigmented hair
owing to the absence of ty ros inase. Therefore, one cannot
exclude the existence of a racial bias that could contribute
to the difference in incorporatin g drugs into hair other
than by the degree of pigmen tation of hair. 1l3
Proced ures used in hair testing include a preliminary
phase fo r specimen collection and storage, a sample prepa­
ration phase for decontamination and drug isolation from
hair structu re, and a drug identification and quantification
phase by instrumental analysis. lOG ,125
Nails can be also used as an alternative to hair analysis
for long -term drug detection . It has been established that
drugs are incorporated into growing nails at levels similar
to those 0 f hair. 126
SPECIMEN COLLECTION
Hair samp les are generally collected by cutting, as near as
possible to the scalp, in the posterior vertex zone. In some
cases, different types of body hair, such as pubic, axillary
and beard hai r, can be used. 127 - 129 The amount of hair
needed for toxicological analysis is about 30-50 mg. If a
segmental hair analysis is required to evaluate drug use
history, th e hair sample is generally cut into pieces of
}-1.5 cm in length, which represent about 1 month's
growth, Hair samples are to be stored at room t emperature
in paper, plastic envelopes or in plastic or glass tubes.
SAMPLE PREPARATION
Hair specimens are normally cut or pulverized to produce
small fragments yielding more surface contact with the
Alternative specimens for drug testing I 265
extraction medium. However, prior to the extraction step,
removing external contamination is mand atory in order to
avoid misinterpretin g positive results. The most frequent
contaminating factors include passive exposure to dust,
aerosol, smoke (when other people smoke drugs), drug
powders (by handlin g drug stuff or any contaminated
material), and sweat and sebum, which carry drugs actively
consumed by the subject. Many decontamination proce­
dures have been proposed, but all involve washing steps
using solvents, detergents and aqueous media, such as
buffer solutions, diluted acid solutions or distilled water
alone. 125
The other fundam ental preparation phase is the isolation
of the drugs from the keratin matri x. In contrast to body
fluids, in which drugs and metabolites are dissolved and can
be directly extracted and analysed, xenob iotics in hair are,
in fact, entrapped into th e solid keratin, Therefore, keratin
matrix must be digested, extracted, hydrolysed or dissolved
before drug separation and identification, Factors affecting
the choice of the most suitable procedure include the chem­
ical structure of the drug, the system in use for particular
dnlg detection, analytical recovery and time of analysis.
DRUG IDENTIFICATI ON
Methodologies applied to drug detection and quantifica tio n
are simi lar to those usually carried out in laboratories of ana­
lytical toxicology.130 lmmunochemical and chromatographic
methods coupled with mass spectrometry or tandem mass
spectrometry to enhance sensitivity and specificity are com­
monly adopted. 126 ,131 To date, more than 60 pharmaceuticals
or drugs of abuse have been reported to be detectable in hair,
including, besides the classical drugs of abuse (opiates,
cocaine, cannabis, amphetamines, methamphetamine, 3,4­
methylenedioxymethamphetamine [MDMA] 3,4-methylene­
dioxyamphetamine [MDA] and other designer drugs), opioids
(i.e. semisynt hetic or synthetic morphine derivatives), hallu­
cinogens, psychostimulants (including nicotine), barbiturates,
benzodiazepines, other sedatives (hypnotics, an tidepressants,
neurol eptics), cardiovascular drugs, anti-infectious drugs and
other miscellaneo us compounds.67. 131,132
CUT- OFF VALUES
Two main threshold values should be mentioned. The first
one concerns the analytical thresho ld, which relies on the
limits of detection and quantification of the chromato­
graphic method used for drug analysis. A second cut-off
value is used to decide whether an analytical result can be
interpret ed with enough reliability with regards to the cir­
cumstances of use or exposure. This second cut-off value is
always higher than the first one. A low cut-off value will
be selected to demonstrate a single dru g expos ure while
high cut-off values will be chosen in case of repetitive '
abuse to exclude external contamination, Several scientific
266 I Recent advances in paediatric toxicology
societies have put forward recommendations for cut-off
values used for interpretation of hair results {e.g. the SoHT
or the Society of Hair Testing) . lOB
Oral Fluid
In recent years, OF has attracted much attention as a pos­
sible alternative to urine for drug testing and, under certain
conditions, to plasma in the area of clinical investigation,
therapeutic drug monitoring and the assessment of recent
drug use. IJJ - 136 Oral fluid is a complex biological matrix
consisting of the secretory products of salivary glands
(saliva) mixed with other fluids, substances and cellular
debris that are present in the oral cavity. In fact, OF has
specific advantages over both urine and blood in being
readily accessible for sampling. Furthermore , it constitutes
a non-invasive sample to collect. Drug concentration-time
profiles measured in OF are generally repolied to resemble
those determ ined in whole blood, although saliva-plasma
ratios may vary considerably depending on the drug prop­
erties. As such, a positive OF drug test may indicate a
recent use, probable intoxication and, in some cases, psy­
chobehavioural impairment, whereas urine testing cannot.
For these reasons, OF drug testing for forensic purposes has
been shown to be reliable compared with other means of
drug testing. 137 In this respect, it is noticeable that the lit­
erature on OF testing is expanding at a substantial rate,
which is indicative of the increasing interest in this unique
biological fluid. 7J ,'38
Testing OF for drugs can provide both qualitative and
quan titat ive information about the drug status of an indi­
vidual. I39 Generally, drug concentrations in saliva are
lower than those found ill urine or blood, and the major
compound detected is the parent drug, not the metabolites.
Initially, oral, intranasal or smoking routes may produce
high concentrations of drugs in OF for several hours owing
to local contamination of the oral cavity. Thereafter, the
concentration of drug in OF is thought to reflect the free
fraction of drug in blood. Thus, the m ajor disadvantage of
saliva is that, after the contamination phase that follows
local administration, many drugs are retained for a shorter
period of time than they are in urine. Also, many drugs are
weak bases and saliva concentrations may be highly
dependent on pH conditions, which itself is dependent on
salivary flow. These factors lead to highly variable saliva­
plasma ratios for many drugs.
Although highly sensitive methods of detection a re
required, most drugs can be detected in salivary secretions.
Therefore, saliva testing is expected to offer many potent ial
applications in the general areas of drug screening and
forensic investigations. Saliva drug tests can reveal the
presence of a pharmacologically active drug in an individ­
ual at the time of testing. Significant correlations have been
found between saliva concentrations of drugs, behavioural
and physiological effects. It can be anticipated that saliva
testing for drugs will develop further in the near future,
with new domains of application being found as soon as
the mecbanisms by which drugs enter the saliva have been
c1 ari fied more speci fically. 140
The marketing of new immuno assays (ElAs) specifica lly
designed for the detection of drugs in OF has enabled the fast
screening and selection of presumably positive samples.
Some of these immunoassay kits are commercially available
with specific collection devices, e.g. Orasure® microplate,
Intercept kits and Cozaft® microplate ElA. Confirmation of
presumably positive results is generally performed with
LC-MS methods. Atmospheric pressure chemical ionizat ion
(APCl) method is the preferred ionization method since it
considerably attenuates ion suppression effects due to OF
matrix components. 141
Sweat
Since 1911, it has been shown that drugs are excreted by
the body in sweat, but no one has developed a practical
solution to the problem of capturing sweat before testing
until recently.142 Occlusive bandages consisting of one to
three layers of filter paper or pieces of cotton, gauze or
towel were proposed to collect sweat. By using these home­
made collectors, it was nevertheless possible to identify
various drugs including quinine, salicylic acid, antipyrine,
ethanol, methadone, phenobarbitone, morphine, cocaine,
cannabinoids, methamphetamine and phencyclidine. 142 ,143
More recently, clothes have been shown to retain opiates
excreted in sweat, demonstrating that it is possible to
obtain evidence about the drug use status of the owners of
pieces of c10tbing (e.g. underwear),144
In practice, systematic collection of sweat specimens is
difficult bec ause of unequal distribution of sweat glands on
different pans of the body.'43 Approximately 50 per cent of
the total volume of sweat is produced by the trunk, 25 per
cent by the legs and 25 per cent by the head and upper
extremities. Sweat is approximately 99 per cent water, tbe
most concent rated solute being sodium chloride. Sweat
production is irregular and is highly dependent upon an
individual's physical activity and emotional state and the
ambient temperature and humidity. Furthermore, sweat
collection methods used in most studies actually obtain a
mixture of sweat and sebum. Drug transport in se bum has
not been examined thoroughly.145
Significant advances have been made in recent years to
develop a sweat patch technology for the routine collection
of sweat samples over an extended period of time. '46 The
sweat patch collection device PharmChek T.", marketed by
Sudormed, Inc., consists of an adhesive layer on a thin
transparent film of surgical dressing to which a rectangular
absorbent pad is attached. The sweat patch acts as a speci­
men container for non-volatile and liquid components of

sweat, including drugs of abuse. Non-volatile substances
from the environm ent cannot penetrate the transp are nt
film, which is a semip ermeable membrane over the pad that
allows oxygen, wa ter a nd carbon dio xi de to pass through
the patch, leaving the skin underneath healthy. Over a
period of several days, sweat saturates the pad and drugs
present in sweat are retained. The patch is generally worn
over a period of 1-14 days, and drugs and metaboli tes
accumul a te over this time period.
Patch testing includ es an extraction step followed by a n
identification step us ing either immuno assay or GC-MS
techniques. 146 The predo minant species fou nd in sweat are
parent compounds, not drug metaboli tes. Analyses of
duplicate patches in controlled studies have revealed tha t
intra-subject variability is low but inter-subject variability
is high. 143
Advantages of this type of testing includ e non-invasive
sampl e collection (the patch being worn on the back, biceps
or chest), ability to carry out normal activities including
swimming and showering without removing the patch, and
the ability to tell if a patch has been removed and reap­
plied . 146,147 Thus, w hen applied to drug testing, the method
increases the window of drug detection to several days and
up to several weeks. Furthermore, the test appears to be
very sensitive since the administration of low doses of
drugs such as cocaine or opiates produces detectable
amounts in sweat after a few days. 147 By offering a cumu­
lative estimate of drug exposure over a period of several
days, this technology is particularly sui table in t he trea t­
ment and monitoring of substance abusers,148 when it is
importan t for medical personnel to obtain information on
the behav iour and activities of these patients. Patches can
be worn continuously and constitute a record of drug
intake during that period .
Meconium and Amniotic Fluid
The determination of drugs of abuse in meconium for the
purpose of determining maternal drug use during preg­
nancy is becoming increasing ly popular owing to its ability
to provide a larger historical record of drug exposure.149-154
Meconium is a dark green mass of water, cells, mucus, sterol
precursors a nd bile pigments which is formed in the fetal
gut from swallowed amniotic fluid and sloughed gastroin­
testinal epithelial cells from 16 weeks to birth; it is dis­
charged 1, 2 or J days after birth. 155-157
Drugs and their metabolites a re passed into the amniotic
fluid from fetal urine, and meconium is though t to reflect
the swa llo wed a nd/o r sloughed sta ble metabolites. Analy­
sis of meconium may therefore reflect in utero exposure to
drugs. Therefore, the window of drug exposure that can be
detected by meconium testing is about 20 weeks , compared
to a 2- to J - day window of de tec tion for a urin e drug abuse
screen. 158
Pitfalls and limitations of drug screens I 267
Since meconium is often available in only sma ll amounts,
screening methods require the development of sensitive a na­
lytical procedures and cl ea n efficient extraction tec hniques
for the characterization of drugs in this particular matrix. 152
In this respect, screening and confirmatory procedures by
immunoassays and GC-MS respectively have been reported
for cocaine and its metabolites, 150, 151,158 cannab is,156 opiates,
including methado ne, 159 phencyciidine,157 cotinine, 160 oxy­
codone l61 a nd benzodiazepines. 162
Generally, the routine anal ysis of meconium fo r drugs of
abuse is recommended for cases in which urin e cannot be
obtained from the newborn or urinalysis is nega tive for the
substances despite a strong suspicion of matern al use of the
substances during pregnancy. 149 Liquid chromatography­
tandem mass spectromehy methods are also ava ilable for
the comprehensive analysis of xenobiotics in meconium. 163
Likewise, a similar method has been presen ted for the
analysis of 10 amphetamine- , methamphetamine- and
MDMA (ecstasy)-related an alytes in human meco nium. 164
Even at low intake, ethanol can cause adverse effects in
newborns and later problems in childhood. Fatty acid ethyl
esters (FAEEs) are fo rmed in the body by esterificatio n of
ethanol with free fatty acids and transesterification of
glycerides. They have been detected in human tissues dam­
aged by ethanol abuse, and also in blood a nd hair. They
can be used as ma rkers for ethanol use. In this respect,
FAEEs in meconium have bee n reported to be potential bio­
markers of feta l exposure to alcohol resulting from mater­
nal alcohol consumption during pregnancy.16 5 According
to Moore et al,166 a total FAEE concentration greater than
10 000 ng/g of meconium may indicate that the newborn
has been ex posed to significant a mounts of alcohol during
pregnancy.
Amniotic fl uid, umbilical cord t issue and cord blood
have also been suggested as alte rnative samples. 167-169
Amniotic fluid is present throug hou t gestation and is con­
stantly diluted owing to fetal urin ation . Fetal swa ll owing
of a mniotic fluid contributes also to the recircul ation of
drug metabolites through the fetu s. Many drugs have been
de tected in amniotic fluid. 17o The major disadvan tage of
amniotic fluid testing is the difficulty and invasiveness of
its collection.
PITFALLS AND LIMITATIONS OF DRUG
SCREENS
The rapid growth and developmen t of drug testing technol­
ogy has created a number of testing methodologies that
can assess a range of biological s pecimens, not only to pro­
vide evidence of recent drug use but also to indicate, as far
as it may be possibl e, the pattern of drug use (i.e. rou te, fre­
quency, dose and time of last use), the degree of impa ir­
ment or the exten t of drug dependence of the individu a l.
Many of these specimens, especially in paediatric forensic
26 8 I Recent advances in paediatric toxicology
medicine a nd pathology, must be used with s pecial atten­
tion bec ause of the potential for false-positive results
owing to the ir contamination during passive env ironmen­
tal exposure.
For the clinician or the pathologist, th e wide variation
in the accuracy of drug tests, coupled with t he confusing
rubric 'toxic or drug screen', often leads to considerable
confusion abo ut what, in fact, the test used is able to
detect. All toxic screen methods have in common, ho wever,
a basic desig n th at max imizes sensitiv ity while making
compromises in specificity.
Immuno assays ge nerally are less se nsitive than GC-MS
or LC-MS, a lthough all are capable of detecting very
low quan tities of drugs. The duratio n of a test positivity
depends not only, in part, on the sensitivity of the test,
but also on the circumstances of abuse. Multiple doses
of a drug taken chronically may redistribute it to deep
body compartments w ith slow release back into the blood
comp artment, s ig nificantly prolonging the pharmacokinet­
ics of elimination. This is particularly true for cannabi­
noids, methaqualone or phencycl idine. Therefore, issues
in dru g pharmacokinetics and test ing methodologies
can easily lea d to both false-positive and false-neg ative
tests, both with significant consequences. Addition all y,
forensic pathologists must bear in mind that after death
changes may occur due to post-mortem dru g redistribu tion
or degradation.
The main pitfalls in forensic toxi co logy and interpreta­
tion of ethanol analysis in post-mortem specimens as well
as ana lytical pitfalls in hair have been reviewed by
Rich ardso n, 40 Kugelberg and Jones,I71 and Musshoff and
Madea.172
Causes of False-positive Drug Screens
The two major causes of false - positive drug results concern
immunoassay tests and contaminations by envi ronmenta l
expos ure to drugs. All immuno assay tests are susceptible to
fa lse-pos itive results owing to a ntibody cross-reactivity
wi th substances bearing some structural simil arity to the
drugs to be detected . 173.174 Included among these are poppy
seeds in t he diet (bread), which may conta in opium con­
ge ners resulting in a drug screen that is pos itive for opiates.
Also com mon is the abi li ty of nasal decongestants , such as
ephedrine or phenylprop ano lam ine, to produce a uri ne
drug screen that is positive for amphetamin es. Newer
immunoassays for amph etam ines have begun to use mono­
clonal antibody assay that theoretically has a high er speci­
ficity for the amphetamine core. Even with the monoclonal
assay, however, cross-reactivity may still occur; phen­
metrazine and I-ephedrine may produce a posit ive result.
Dextromethorphan and diphenhydramine, commonly used
over-the-counter agents, also have the potential to p roduce
a positive urine test resul t for phencyclidine. Household
and commercial adulterants can produce fa lse- negative
results for some drugs a nd false-positive results for others
depending on the type of immunoassay that might be
used. 173 Finally, cross-reactive su bstances proved to pro­
du ce false-positive drug screen ing include: 66
• Opiates - chlorprom az in e, codeine, dextrom ethorphan,
dihydrocodeine, diph enoxylate, hydromorphone,
levorphanol, meperidi ne, oxycodone, d-prop oxyphene.
• Ampheta mines - cathinone, diethylpropion, dopamine,
ephedrine, p-hydroxya mpheta mine, isoxs upri ne,
labata lol , methy lphenid ate, l-meta mpheta min e,
ny lidrin, phenmetraz ine, phentermine, pheny lep hrine,
phenylpropanolamine, propylhex edrine,
pseudoephedrine, ranitidine, sergiline,
N-acetylprocain a mide, chloroquine, procainamide.
• Barbiturates - gJuthetimide, phenytoin.
• Cocaine (benzoylecgonine) - salicylates.
• Phencyclidine - am itripty line, chlorp romazine,
dextromethorpha n , diphenhydramine, doxylamine,
meperidine, thioridazine.
• Cannabinoids (THC-COOH) - ibuprofen, promethazin ,
ribo Ravin.
Although it may arguably represent genui ne exposure
and therefore not a fa lse-positive result, environmental
exposu re to drugs, such as marijuana smoke, may produce
a positive urin e specimen. Ho wever, it has been genera lly
co ncluded that passive inhalation exposure, in the absence
of extreme conditions of ambient exposure, does not result
in posi tive urine tests at conve n tiona l cut-off condi­
tions. 175 The problem of external contamination is more
crucial concerni ng hair testing. 172 ,)76 In situ ations in
which dru gs are know n to be present in the environment,
it is well known that passive exposure can produce posi­
tive hair analysis resu lts. In this respect, apprecia ble levels
of nicotine have been reported in th e unwashed hair of
non-smokers which ca me into contact with the hair of
smokers. 177 In contrast, cotinine (the ni cot in e metabolite)
does appear to be a marker of tobacco use in the smoker
population. 17B In a study condu cted by Smith and Kid­
well 179 on children living in a family in which cocaine
(crack) was used, and thus present as smoke in the envi­
ronment , it has been shown that 85 per cent of the chil ­
dren tested positive for cocaine and benzoyl ecgonine in
their hair. In th is study, it was ass umed that children aged
3-10 were unlike ly to be self-admi nisterin g cocaine, so
that any cocaine in their hair must have come from pas­
sive exposure. This was confirmed by the fact that skin
wipes obtained from the children by wiping their fore­
heads with a cotton swab were also positive for cocaine,
indicating extensive surface contact, whereas saliva test­
ing was negative in most of the ch ild ren, showing th at
in gestion of coca ine was not likely a so urce of the cocaine
in the hair.
The problem of hair contam in at io n by environmental
exp osure to a drug is more complex; experts disagree
about interpretation, with some claiming that external

contamination can be removed by washing through several
kinds of decontamination procedures, while others contend
that washing removes drugs from inside the hair shaft and does
not completely remove external contaminants. IBO Caution
until better decontamination procedures are established
should induce toxicologists to test only short hair lengths
from close to the scalp to limit environmental exposure.
Of course, such a limitation would reduce the potential
advantage of hair analysis to provide long-term drug use
history for an individual. Notwithstanding this contro­
versy, it seems reasonable to point out that not all positive
hair analysis results must be interpreted as due to passive
exposure. Certainly, most positive results are due to the
administration of drugs. Nevertheless, the above example
illustrates that interpretation of data must take passive
exposure, especially in young children, into consideration
to determine the source of drugs in any case of a positive
result. 176
Causes of False-negative Drug Screens
Equally problematic in drug testing are situations in which
a drug test fails to identify the adolescent who has, in fact,
been recently abusing psychoactive drugs and in whom a
drug test should be positive. The reasons for a false­
negative test can be divided into three general categories:
technological shOitcomings, pharmacokinetic characteris­
tics and intentional specimen alteration or adulteration.
TECHNOLOGICAL SHORTCOMINGS
One of the most common reasons for a negative toxic screen
result, despite obvious drug abuse, is that the clinician or the
pathologist fails to recognize that the particular screen being
used is incapable of detecting the substance in question. 52 For
example, some chemicals abused by adolescents, notably sol­
vents and other inhalants, cannot be detected by routine
screening of either blood or urine by commercially available
immunoassays. Without prior knowledge of what the test
specifically seeks, the diagnosis may be missed. Similarly,
although drug screens commonly identify all members of a
drug family, structural differences in specific drugs within the
family may limit their detectability (e.g. fentanyl and its ana­
logues in an opiate screen, or clonazepam in a benzodi­
azepine screen). Finally, drug tests can be falsely negative
because of poor quality control of the laboratory. Inaccuracy
rates (false positivity rates as high as 1.7 per cent; false neg­
ativity rates ranging from 17 per cent to 31 per cent) have
been reported in some selies.IBI-IB3
PHARMACOKINETIC CHARACTERISTICS
The pharmacokinetic characteristics of drugs have an
important influence on drug detectability, particularly if
Pitfalls and limitations of drug screens I 269
the wrong biological specimen is examined or if the wrong
technology is used. For example, drugs that have a large
volume of distribution have correspondingly low serum or
blood concentrations (cannabinoids or LS~). Immunological
drug tests designed for urine specimens are therefore gen­
erally incapable of detecting these drugs in serum
or blood.
Another pharmacokinetic characteristic, elimination
half-life, also has important implications for drug detection
capability. With drugs that have a short elimination half­
life (e.g. cocaine) the parent compound may be undetectable
in blood within 8 hours; however, cocaine metabolites
may be detected in urine for several days after a significant
exposure because of the drug excretory pattern. Another
factor to be pointed out is that the pharmacokinetics of cer­
tain drugs may vary according to age. In younger children
the elimination half-lives are shorter and clearances faster.
For example, clearance of morphine reaches adult levels by
6 months of age. IB4 Chlorpromazine is metabolized 2-5
times more rapidly by children under 5 years old than by
adults. IBs Subsequently, as a general rule, urine remains the
best biological specimen for drug testing because it takes
advantage of the fact that kidneys are the primary excretory
organs for most drugs.
INTENTIONAL SPECIMEN ALTERATION OR ADULTERATION
Biological specimens, palticularly urine, can also be inten­
tionally altered or adulterated to produce a false-negative
result. IB6 Methods available to accomplish this are numer­
ous and varied, ranging from simple dilution of the col­
lected specimen to substitution of urine that was produced
before any drug use occurred or a fluid that resembles
urine. 90 ,IB7 Another commonly used method consists of
drinking a large amount of fluid, or even using a diuretic,
in order to reduce the concentration of drug in urine so
that it falls below the detection threshold of the assay and
will produce a negative urine. However, adulteration of the
collected specimen with chemical agents is the method
chosen by many users because it requires little sophistica­
tion and can be easily accomplished in unobserved collec­
tion conditions . Such substances include vinegar, lemon
juice, bleach, ammonia-based cleaner, crystalline drain
cleaner, non-ionic liquid hand soap, methanol, sodium
chloride, toilet bowl cleaner, ionic detergents and even
whole blood anticoagulated with EOTA IBB Although well­
designed collection procedures can minimize the OppOItu­
nity for sample adulteration and laboratory tests exist that
may detect certain types of adulterants (temperature, spe­
cific gravity, pH measurement of the sample and measure­
ment of urinary creatinine), no system is absolutely
fool-proof. IB9
Commercial adulterants are also available: some con­
tain the fixative glutaraldehyde (UrinAid, ClearChoice). 190
others contain strong inorganic acid (Amber- 13), such as
nitrite and chromate (Urine Luck), or they may contain
270 I Recent advances in paediatric tox 'icoiogy
enzymes, e.g. peroxidase or protease (papain).'91.192
On-site adulterant detection devices that assess the integlity
of urine specimens are also commercially available
(e.g. Adultacheck 4).193
Post-mortem Changes
POST-MORTEM REDISTRIBUTION OF DRUGS
The difference in drug concentration in blood samples from
different sites represents a major concern for the forensic
toxicologist and, in general, blood concentrations after
death cannot be interpreted from a knowledge of concen­
trations in life. This is because, in life, drugs are distributed
to different tissues in the body and differences in concen­
tration between tissues can be built up and maintained.
Active processes, which are responsible for maintaining
concentration differences, cease after death and concentra­
tions can change dramatically. The changes in blood are
not necessarily uniform. Concentrations in venous blood
after death are expected to vary according to the tissues
whose blood they drain. For example, tricyclic antidepres­
sants, narcotic analgesics, local anaesthetics and antihista­
mines all accumulate in myocardial tissue and may be
partially responsible for the observed rise and subsequent
high levels found in heart blood post mortem. 194 In this
respect, blood in the right ventricle drains from the liver
and other body tissues and is likely to differ substantially
from the blood in the left ventricle, which drains the
lungs. The site and time-dependent variability of post­
mortem blood and tissues sampling, and the phenomenon
of post-mortem redistribution have been documented and
reviewed in the recent literaturel95-200 after having been
described as 'a toxicological nightmare' by Pounder and
Jones in 1990. 201
Several controlled animal studies as well as case reports
have been published that address these issues. Studies with
acetaminophen (paracetamol) in rabbits demonstrated that
post-mortem drug concentrations in blood increased sig­
nificantly with time for central sampling sites compared
with peripheral blood. 202 The post-mortem redistribution of
amitriptyline was demonstrated in rats, which suggested the
post-mortem drug release from lungs and other drug-rich
tissues into the blood. 203 Post-mortem redistribution has also
been demonstrated in rats following administration of mor­
phine, 204 digoxin 205 and secobarbital. 206 Conversely, rats to
which cimetidine was given did not show statistically signi­
ficant differences in post-mortem changes in cimetidine con­
centrations in tissues compared with blood. 207
Several human case reports have described similar find­
ings. Site-dependent post-mortem changes in blood and tis­
sues concentrations of digoxin, 20s cocaine,209 methadone 21O
and methamphetamine 2J' have been reported. Post-mortem
redistribution of chloroquine,212 dothiepin,213 amitripty­
line,214 fluoxetine and norfluoxetine,2J5 tranylcypromine 216
and co-proxamol 217 have been reported. In contrast, other
case reports have determined that there is little evidence that
post-mortem redistribution occurs with tricyclic antidepres­
sants,2lS trazadone 219 and zopiclone. 220
Similarly, drugs can be expected to diffuse from gastric
residue into blood, liver, lung and other nearby organs.
Rats killed and then subjected to gastric instillation of
amitriptyline showed drug diffusion into the liver from
5 hours post administration at room temperature. 221 Over
time, highest concentrations were reached in liver lobes adja­
cent to the stomach, with significantly lower levels in the
right lobe. Similar studies on human cadavers used amitripty­
line, paracetamol (acetaminophen) , and lithium in quantities
representing 10 tablets of each drug. 222 After 48 hours at
room temperature, drug diffusion affected the left lobe of the
liver and, to a lesser extent, the caudate lobe and variably the
light lobe posteriorly (with the cadaver supine). In the lung,
the left was more affected than the right and the base more
than the apex, although this was not always true.
Post-mortem diffusion of ethanol, facilitated by its high
hydrosolubility, from the stomach into the blood and the
nearby organs has been shown to follow a similar pattern
to that observed with drugs 223
It is apparent that drug diffusion (including ethanol) in
a cadaver is an important and complex phenomenon that
can affect a wide range of organs as well as blood . There­
fore, knowledge of sampling site and sampling technique is
a prerequisite for a valuable interpretation of analytical
results. To facilitate this, pathologists should formalize
their protocols for toxicological sampling at autopsy. Blood
samples should be obtained by needle puncture of the
external iliac or femoral vein. This may prove difficult in
babies. A liver sample should be obtained from deep within
the right lobe and any lung sampling should be taken from
the apex rather than the basal lobes.
POST-MORTEM DEGRADATION OF DRUGS
Post-mortem degradation of drugs and pOisons is a process
that is little understood but may also significantly affect the
interpretation of post-mortem toxicological results. First,
there is the delay that occurs before the sample has been
taken, and then between sampling and initial presumptive
drug screening or blood alcohol determination. Second, the
subsequent confirmation may not be performed until the
case goes to court for trial or may be done many days or
weeks after the blood has been taken, especially when toxi­
cological analyses are carried out by separate institutions or
at several laboratories each carrying out isolated tests. Usu­
ally additives and preservatives have been added but not
always, and samples have been stored at 4°C or at -20°C.
Degradation before analysis may occur as a result of chem­
ical or physical decomposition owing to the instability of
the drug, which lacks the protective effect of being bound to
plasma proteins. 224 For example, diazepam 225 and dilti­
azem 226 are susceptible to hydrolysis and flunitrazepam 227

and LSD 228 were reported to significantly degrade within
only 24 hours when exposed to sunlight. Chemical instabil­
ity has been also attributed to the time-dependent loss of
drugs, such as chlorpromazine or dothiepin, after death 229
Other drugs, such as paracetamol (acetaminophen), salicy­
lates, benzoylecgonine, cannabinoids, morphine, buprenor­
phine and amitriptyline, would be more or less stable. 230-232
Degradation may also derive from enzyme activities
linked to bacterial development that takes place after death
and may continue in an unpreserved sample after collec­
tion. 233 Enzyme activities in the blood of healthy living
persons are thought to be mainly restricted to esterases,
which can affect cocaine, procaine and other esters; this is
the reason why an enzymatic inhibitor must be added as a
preservative to the blood sample. 224
After death, intestinal bacteria penetrate the intestinal
wall, enter the blood and lymph vessels and migrate further
throughout the body. At 25°C this transmigration can
occur as soon as 5 hours after death. The possible role of
bacterial degradation has been assumed to explain, at least
partially, the continuous decrease in concentration in
blood after death of many drugs such as cocaine or
methamphetamine. Bacteria have been shown also to be
responsible for the post-mortem metabolic changes of
many drugs, including the nitrobenzodiazepines (clon­
azepam, nitrazepam and flunitrazepam) metabolized to their
7 -ami no meta bolites. 224 ,234, 235
POST-MORTEM FORMATION OF DRUGS
In contrast to the degradation of drugs post mortem, cer­
tain compounds may appear to increase in concentration
after death or during storage in unpreserved containers,9]
These include cyanide, which is generated by the degrada­
tion of haemoglobin and ethanol. Ethanol, which tops the
list of psychoactive substances encountered in post­
mortem toxicology, can be produced by post-mortem fer­
mentation of glucose, though this rarely ex ceeds 0.5 gOfc.
Ethanol can, of course, be both formed and destroyed by
microbial activity236 A recent comprehensive review of
issues relating to the interpretation of ethanol concentra­
tions in post-mortem specimens was presented by Kugel­
berg and Jones. llI A chemical preservative, such as sodium
fluoride (1-2 per cent w/v), should be added to specimens
intended for determination of ethanol, which should pre­
vent any further production of ethanol. The addition of an
insufficient amount of fluoride (such as 2-3 mg/mL in
commercially available tubes) may result in very particular
cases in the in vitro formation of ethanol, reaching a con­
centration as high as 3.5 giL. 237 Distinction between
ingested ethanol from that formed post mortem can be
made by the detection of ethylglucuronide, ethylphosphate
and ethylsulphate, three speciflC markers of ethanol
intake,2J8,2]9
If available, vitreous humour, which is more protected
from bacterial contamination, can be used instead of blood ,
Specific applications I 271
even though the predictive value of vitreous humour in
estimating blood alcohol concentration appears somewhat
questiona ble. 24o ,241
SPECIFIC APPLICATIONS
Drug Screening for Doping Agents
The misuse of doping agents in sport is concerned with
substances registered on an annually updated list, 'The pro­
hibited list. International sta ndard' (World Anti-doping
Code, World Anti-Doping Agency (WADA)) . Depending on
the survey methods used, it has been estimated that doping
may involve 3-5 per cent of adolescents. For instance,
studies carried out in eastern France with 1501 athletes
aged 15-19 years found that 4 per cent had used banned
substances at least once in their Iife,242 The same study
design and setting, i.e. self-questionnaire survey, were used
in a 4-year follow-up of doping prevalence among pre­
adolescent athletes,24J This study disclosed that 1.2 per cent
of the respondents entering the cohort and 3.0 per cent
4 years later reported that they had used doping substances
at least once during the preceding 6 months. The main sub­
stances used as doping agents were salbutamol, cortico­
steroids and cannabis. Cannabis is one of the drugs that is
most frequently detected in the context of doping.244
With the ever-increasing number of dietary supple­
ments used by athletes and the availability of numerous
synthetic steroids and recombinant peptide hormones and
modulators, detection of performance-enhancing com­
pounds is increasingly difficult,245,246 and needs more and
more sophisticated detection methods such as (fast) chro­
matography coupled with tandem mass spectrometry247 or
carbon isotope ratio determinat ion ,248
The paediatrician might suspect such use when the young
athlete presents with recent increase in weight or muscle
mass, oedema, gynaecomastia in males, deepening of the
voice in females, new acneiform rashes, changes in behav­
iour (initability, labile temperament, depression), or a new
onset of chemical hepatitis. Careful and sensitive question­
ing, in a non-judgemental but knowledgeable manner, can
sometimes induce the adolescent to confess steroid use so
that counselling can take place. 52 According to Dawson,249
the patients who should concern the physician the most are
not elite athletes but the yo uth who are being increasingly
drawn to the use of performance-enhancing drugs.
Many health problems may arise because of use and mis­
use of doping agents, such as intake of counterfeit products,
drug-drug interaction, dermatological disorders, and cardio­
vascular, hepatic, genitourinary, musculoskeletal, neuropsy­
chiatric, endocrine and haematological adverse side-effects,
not to mention infection resulting from the use of contami­
nated needl es,249,250
Over 100 an abolic steroids are widely available on
the black market in oral or injectable form. 251 Extensive
272 I Recent advances in paediatric toxicology
metabolic pathways as well as natural occurrence make the
detection of these substances difficult to interpret. Positive
tests are usually confirmed by GC-MS. Oral steroids can be
detected as long as 4 weeks after last use. Normally, testos­
terone exists in the body in balance with epitestosterone
with ratios of between 1 and 2. Epitestosterone is a natu­
rally occurring epimer of testosterone, thought to be of
extratesticular origin and having no biological activity.
A testosterone-epitestosterone (TIE) ratio greater than 6 (the
threshold ratio used by the International Olympic Commit­
tee) is considered as evidence of exogenous androgenic
steroid use. To further foil ingenious athletes from taking
exogenous epitestosterone to reduce their TIE ratios below
6, a maximum allowable urine concentration of 150 ng/mL
epistestosterone has been established. 252
When to use urine testing remains controversial.
Although unannounced steroid testing at high-school ath­
letic events would perhaps diminish such substance use, this
testing is prohibitively expensive. Additionally, it would
require that a chain of custody be established, and such
testing might engender an issue of privacy and consent.
Furthermore, besides anabolic agents, peptide hormones,
including growth hormone, erythropoietin, adrenocorti­
cotropin (ACTH), chorionic gonadotrophin and their releas­
ing factors, are also widely used among athletes. Although
detection of these peptides by immunoassay is well estab­
lished in the clinical laboratory, the acceptability of this
method of analysis as definitive proof of administration has
yet to be established in forensic toxicologY.246 However, the
unequivocal identification of peptide and protein analytes
used for doping will very likely rely on complex hyphenated
mass spectrometry methods following very specific extrac­
tion and purification methods. In this respect, immuno­
affinity purification will precede the analysis with tandem
mass spectrometers, such as quadrupole-linear ion trap or
linear ion trap-orbitrap analysers. 253
Forensic Applications of Hair Analysis In
Children and Teenagers
Owing to its ability to provide evidence of drug exposure as
well as drug use, hair testing for drugs of abuse has been
available for several years in courts and accepted according
to the legal conditions of the country or state where trials are
pending. Of course, toxicologists are aware that no sample
technique or specimen can provide answers to all toxicolog­
ical questions. More information is obtained with a variety
of analytical approaches and different biological specimens.
However, hair analysis offers 'a unique perspective on
human drug use'254 by providing a wider window of drug
detection and may offer advantages over other drug testing
methods in terms of accessibility, resistance to post-mortem
degradation and stability of the entrapped substances in it.
The following examples are given to illustrate the potential
value of hair drug test results in forensic practice.
Sachs 103 reported the case of a child who had been
brought to a clinic with symptoms of intoxication attributed
to clozapine. As clozapine was not freely available, the
mother of the child was suspected of having intoxicated her
child with the drug prescribed to her depressive sister. One
year before, a l-year-old boy belonging to the same family
had been found dead. Owing to the absence of any obvious
cause of death, the case was ruled as sudden infant death
syndrome. After exhumation, the child's hair, the only tis­
sue available for toxicological analysis, tested positive for
clozapine, providing evidence of repeated administration of
the drug to the child when he was alive. However, only the
qualitative result was of importance to the judge; as the
child could not have been under antidepressive treatment,
the mother was sentenced to 4 years only because she even­
tually admitted giving the drug to the child.
Lewis et al 255 reported the usefulness of hair testing of
children living with suspected drug users (crack) and con­
sidered to be at risk. Using the hair of the children to deter­
mine smoke exposure provided 'extra credibility' to the
child protective services and allowed them to remove chil­
dren from dangerous households. Similar hair testing
applications have been reported by Smith and Kidwell. 179
Strano-Rossi et al 256 described the application of hair
analysis in a judicial case to document coercive heroin
administration to a 5-year-old child who was admitted with
overdose symptoms to an intensive care unit. The results of
toxicological hair analysis were accepted by the court with­
out criticism; only the possibility of the consumption of
antitussive drugs was raised by the defendant's lawyers.
However, the presence of monoacetylmorphine (MAMl. the
specific metabolite of heroin, allowed the toxicologist to
refute this objection. The relative of the child was found
guilty and sentenced accordingly.
In the same way, Huestis 254 quoted several American
court decisions ordering that parents' hair be collected and
tested for drugs of abuse (cocaine and marijuana cases) to
determine the fitness of the custodial parent in child cus­
tody disputes or to evaluate their degree of drug addiction
in order to determine the best interests of the child in adop­
tion cases.
Ketamine is a rapid-acting dissociative anaesthetic. It
has been reported to be abused by an increasing number of
young people as a 'club drug', and is often distributed at
'raves' and parties. Teenagers are the major abusers. Keta­
mine concentrations were found to range between 0.8 and
92.3 nglmg in hair collected in entertainment places from
15 ketamine abusers. 257 Ketamine metabolites were also
detected, indicating ingestion of the drug and excluding
the hypothesis of external contamination.
Methylphenidate (sold as Ritaline®, Medikinet® or Con­
certa®) is used in the treatment of childhood ADHD.
Methylphenidate is hydrolysed in the body into ritalinic
acid. Both the parent drug and its metabolites can be
detected in blood. However, only the parent compound is
present in hair from treated individuals. As hair samples in

children can be easily collected, in contrast to invasive
blood sampling, determination of methylphenidate in hair
should be an alternative to check compliance in a wider
time window than if using blood. Methylphenidate can be
measured in hair by LC-MS. 258
Hair analysis can be very useful also in the case of sus­
pected drug-facilitated crime or alleged sexual abuse of
children and teenagers. Drugs generally involved in
such crimes are sedatives, hypnotics and anaesthetics. Alco­
hol or drugs of abuse, such as cannabis, LSD, gamma­
hydroxybutyrate (GHB) or ecstasy, are also used. lOB Because
the majority of cases are reported to the police after a few
days, the hair is very often the only specimen suitable for
drug detection. In cases of child abuse, the drug can be
administered once or chronically. A repeated administration
is easier to identify because drug concentrations are gener­
ally much higher than after single exposure. The analysis is
more challenging with low-dosed drugs, such as buprenor­
phine. 259 In the case of a single application, very sensitive
methods are required for drug detection, in concentrations
down to the sub-pg/mg hair range. This was made possible
through the introduction of new analytical techniques 2 60
For instance, bromazepam (0.8-28 pg/mg) was measured by
tandem mass spectrometry methods in three volunteers
after a single ingestion of 6 mg261.262 Recently, a sex
offender has been charged with oral genital contact with an
8-year-old girl. The events were reported by the girl, who
experienced insomnia and speech disorders. Hairs were
sampled 1 month later and bromazepam was detected in the
hair segment corresponding to the timing of the reported
facts. 263
The usefulness of hair testing was demonstrated in
another drug-facilitated sexual assault case of a 9-year-old
girl. In that case, diphenhydramine, an antihistamine agent
with sedative effects, was discovered in successive hair
segments, indicating repeated surreptitious administration.
The perpetrator admitted charges at court relating to the
abuse of young girls. 264 The same group reported the
administration of a phenothiazine derivative with sedative
properties (trimeprazine or alimemazine) to two children.
The stepmother, who was the perpetrator, did not challenge
in court the suggested lise of trimeprazine as a sedative
drug. 26 5
Two deaths of babies sedated with methadone by their
mothers were reported by Kintz et al 266 Both mothers were
former heroin addicts under methadone substitution ther­
apy. At autopsy, no evidence of violence was noticed. Tox­
icological investigations by GC-MS in hair and blood
demonstrated recent and repeated methadone exposure.
Hair (6 cm) from the first 14-month-old baby tested posi­
(ive at 1.9 and 0.8 ng/mL for methadone and EDDP
(methadone metabolite) respectively. In another case, a
comatose 5-month-old girl was taken to hospital, where
she was declared dead 1I days later. Hair analysis (5 cm)
revealed the presence of methadone at a concentration of
J.O and 21.3 ng/mL in the root and end hair segments
Specific applications I 273
respectively. Many other cases have gone to court having
similar circumstances of a baby dying suddenly and
unexpectedly.
Gamma-hydroxybutyric acid is presumed to be one of
the most frequently used substances in drug-facilitated
crimes. Because of its very short half-life, GHB is particu­
larly difficult to detect in blood and urine. Furthermore,
GHB is also an endogenous substance found in low con­
centrations in body fluids. Nevertheless, detection of a
single GHB exposure has been reported to be possible. 267
Determination of Fetal Exposure to Drugs of
Abuse
Maternal self-reported drug history has been shown to be
unreliable, as many women who deny use during preg­
nancy exhibit drug metabolites in their urine. 26B On the
other hand, systematic urinalysis during pregnancy is
hampered by the sholt elimination half-life of the drugs.
A negative result may be a result of deliberate abstinence for
several days before biomedical screening. A positive result
reflects only exposure during the preceding 1-3 days.
A common method of estimating the amount of fetal expo­
sure to a drug could be provided by measuring the concen­
tration in the umbilical blood at birth. 169 The problem is
tha t the presen ce of drug in umbilical blood reflects only a
very recent exposure and lack of drug does not rule out
fetal exposure for most of the pregnancy. Measurement of
drugs in neonatal urine presents the same disadvantages.
Evaluation of drug concentrations in the amniotic fluid
measured during pregnancy or at delivery cannot provide
information on the duration and degree of fetal expo­
sure. 167 The same disadvantages are noted with the analy­
sis of meconium, which is only a qualitative test at the
moment of delivery.269
Neonatal hair begins to form at approximately 6 months
gestational age; a positive result indicates use during the
last trimester. Hair testing can be collected during the first
trimester of life, after which time infant hair replaces
neonatal hair. Drugs in neonatal hair could originate from
deposition from fetal blood or from contamination of hair
by amniotic fluid . 170 Hair from the mother can be collected
also for drug analysis and results can be compared with
those obtained from neonatal hair. Hair analysis may
therefore remedy the disadvantages previously mentioned
with the other methods, with a wide window of detection
ranging from weeks to months, and may provide informa­
tion concerning the severity and pattern of an individual's
drug use when a maternal drug history is not available or
is in doubt.
Regarding the correlation of drugs in maternal and
neonatal hair, concentrations of nicotine and cotinine in
paired maternal and neonatal specimens were found to be
well correlatedYo Mothers and infants in the smoking
groups, both active and passive, had significantly higher
274 I Recent advances in paediatric toxicology
concentrations of nicotine and cotin in e than did those in
the control, non-smoking group. Th is study documents
the impoltance of hair ana lysis as a tool for measuring
exposure to cigarette smoke. Indeed, passive smoking has
been shown to adversely affect the health of infants and
children .271
Prenatal cocaine and methamphetamine exposure is
now frequently tested because of their relatively high
prevalence of use during pregnancy, especially in th e USA.
Gestational exposure has been associated with placental
abruption and premature labour, as well as with an
increase in low birth weight, microcephaly, congenital
anomalies and necrotizing enterocolitis. Neonatal hair was
found to be a suitable biological marker to demonstrate in
utero cocaine and methamphetamine ex posure. 272 ,27J Inter­
estingly, median cocaine concentratio n was lO-fold higher
in the hair of the mothers than in the neonates. 274
In utero alcohol exposure also can be detected by hair
an alysis. The retrospective detection of alcohol consu mp­
tion during pregnancy can be performed also by measuring
specific markers of alcohol abuse in the mother's hair. Fatty
acid ethyl esters and ethylglucuronide, which are minor
metabolites of a lcohol , are suitable markers for the detec­
tion of heavy alcohol consumption. They may be used in
instances of fetal alco hol syndrome or suspicion of regular
alcohol intake during pregnancy to objectiv ate alcohol
abuse,275
Recently, analysis of vernix caseosa has been proposed
as an alternative specimen to hair when hair is not avail­
able or is too difficult to collect. 276 Vernix caseosa is a
thick, white lipid and cell mixture that covers the fetus,
starting at about 24 weeks' gestational age. 171 Vernix can
eaSily be removed from a newborn's skin with a piece of
gauze. Cocaine and/o r three of its metabolites were
detected in a 3-cm section of gauze containing verni x
obtain ed from three out of five neonates.
THE IMPORTANCE OF PAEDIATRIC
TOXICOLOGY IN SPECIFIC CASES
Deliberate Poisoning by Health Professionals
Beverley Gail Allitt, dubbed 'the Angel of Death', was an
English State Enrolled Nurse (SEN) who was convicted of
killing four children and injuring five others, in 1991, on
the children 's ward of Grantham and Kesteven Hospital,
Lincolnshire, w here she worked. 277 She has since become
one of Britain's most notorious female serial killers. Her
main method of murder was to inject the children under her
care, particularly those in intensive care, with insulin or
potassium to ca use a cardiac arrest. 278 In an environment
such as a children 's wa rd, where staff are dedicated to pro­
viding the best possible care, the threshold for recogni zing
covert acts of excess and inappropriate ad ministration of
therapeutic substances is high. It is not, therefore, surprising
that several deaths had occurred before that thresh old was
reached in this particular instance. Each individual unex­
plained death occurring in a hospital environment requires
careful investigation, and the careful and thorough exclu­
sion of poisoning279 (see Chapter 18).
Fabricated or induced illness by carers (FlI) was first
described in 1977 by Meadows, a paediatrici a n in Leeds .
The condition was, at the time, referred to as Munchausen's
syndrome by proxy280 Among other manifestations, ill­
nesses are induced by carers of children (mostly females) by
ad ministering substances, smothering, withholding nutri­
ents and medicines, and by other means, e.g. introducing
infectious material into the gut or bloodstream, Commonly
used poisons include insulin, salt, antidepressants, anti­
coagulants and bleach. These chi ldren may present as acute
life-threatening events, as unexplained deaths or as baffling
clinical problems; the assistance of the tox icologist is of
great value in these instances.281 Occasionally, similar prob­
lems arise when others close to the child but not actual car­
ers, for a variety of potential reasons, induce illness in a
baby by administering therapeutic substances.
Toxicology and Sudden Infant Death Syndrome
Among the many hypotheses put forw a rd as explanat ions
for sudden infant death syndrome (SIDS), a number
involved poisoning, e.g. antimony leaching from cot mat­
tresses. 282 Another hypothesis that the primalY cause of
SIDS is poisoning by toxic gases generated in cot mattresses
was first publicized in the media in 1989 by Richardson, an
independent consultant on biodeterioration of materials. It
was first formally published in the Lancet in March 1990. 283
The theory proposed was that a fungus (Scopulariopsis
brevicaulis) sometimes found in the domestic environment
could degrade the chemical compounds of phosp horus,
arsenic and antimony that may be present in fire ret ardants
or plasticizers in PVC cot mattress covers and other cot fur­
nishings; this wou ld subsequently release the toxic gases
phosphine, arsine and stibine 284 The hypothesis was inves­
tigated by two Independent Expert Groups285,286 that found
that Richardson's conclusions could not be substantiated by
independent researchers, and thus concluded in 1991 and
1998 that the hypothesis was unfound ed.
CONCLUSIONS AND FUTURE
CONSIDERATIONS IN FORENSIC PAEDIATRIC
TOXICOLOGY
Use of hyphenated mass spectrometlY methods has permit­
ted the simultaneous measurement of a large spectru m of
potentially toxic drugs. Increase in sensiti vity allows a
decrease in the size of the sampled specimens. The use of
alternative biologica l matrices, such as neonatal hair,
meconium, amniotic fluid or oral fluid, extends the field

and time window of investigation. Howeve r, many short­
co mings ex ist, especially regardi ng the post-mortem redis­
tribution and stability of drugs in the fetus, neonate and
in fant. The neoformation of ethanol in paediatric subjects
must be studied and the usefulness of drug exposure mark­
ers for this particular popul ation should be better charac­
terized. Basic research is still needed to better characterize
the pharmacokinetic and pharmacodynamic aspects of
dnlgs in paediatric patients and victims. The know led ge of
the developmental patterns of dru g metabolizing activities
will facilitate the determin ation of the optimal therapeutic
dose ofa drug and its pharm aco logica l consequences in the
neon ates. If the mechanisms of drugs deposition in adult
hair and other alternative matrices are partially elucida ted,
almost eveJY1:hing remains to be done regarding neonates.
For instance, the influence of pigmentation or hair type
may differ from that known for teenagers and adults.
Finally, a la rge database that compiles t herapeutic, to x ic
and comatose fatal blood concentrations of drugs from
newborns and children should be readily available. The
creation of such a database will make easier the interpreta­
tion of blood concentrations measured in clinical and
forensic toxicology.
REFERENCES
I Goepp J. Pediatric poisonings. Clill Chem 1996; 42: 1356-60.
2 Ell enhorn MJ, Barceloux DG. Med ica l toxico logy. In Diagllosis
alld Treatment oj Human Poisoning. New York: Elsevier, 1988.
3 Fine J, Goldfrank L. Update in medical tox icology. Pediatr
28 Korkia P, Stimson Gv. Anabolic steroid use. Lancet 1993;
Ciin North Am 1992; 39:103J-51.
341:1407.
4 Burda A, Burda N. Controversies in tox ico logy. The na ti on's
first poison co ntro l center taking a sta nd aga inst acc idental
childh ood poisoning in Chicago. Vet HUll! Toxicol 1997;
39:115-19.
5 Hanzlick R. Nationa l association of medical exam iner's
stero id use in the United States. JAtvIA 1993;
Pediatric Toxicology (PedTox) regis try. Toxicology 1996;
2 70: 1217-21.
107: 153- 8.
6 Liebelt EL, Shannon MW. Small doses, bi g problems.
aspects of doping. J Ciill Forensic Med 1.998; 5:1-7.
A selected review of highly toxic common medications.
32 Buck ley WE, Yesa lis CE, Friedl KE, et at. Estimated
Pedia tr Emerg Care 1993; 9:292-7.
prevalence of anabolic steroid use amo ng male high schoo l
7 Berlin CM. Advances in pediatric pharmacology an d
toxicology. Adv Pet/illtr 1993; 40:545-74 .
8 Berlin CM. Advances in ped iatric pharmacology and
tox ico logy. Adll Pediatr 1995 ; 4 2 :593-629.
9 Berli n CM. Advances in pediatric pharmacology and
toxicology. Adll Pediatr 1997; 44:545-74.
10 Criddle LN!. An ove rvi ew of pediatric poisonin gs. AACN Ad/!
Crit Care 2007; 18:109 -18.
11 Eldridge DL, Van Eyk J, Kornegay e. Pedia tri c toxicology.
Emerg Med Ciin North Arn 2007; 25:283- 308; abstract
vii- viii.
12 Chasn off IJ, Burns WJ, Sc hn o ll SH, Burn s KA. Cocaine use
in pregnancy. N Engl J Med 1985 ; 313:666-9.
13 Rubin JD, Ferencz C. Loffredo C. Use of prescription and
non-prescription drugs in pregnancy. J Clill Epidemiol 1993;
46:581-9.
14 Chasnoff IJ, Lewis DE, Griffith DR, Willey SD. Cocaine and
pregnancy. Clinical and toxicological implicati o ns for the
neonate. Ciin Chem 1989; 35: 1276-8.
References I 275
15 Burkett G, Yasin SY, PaIO\'11 D, et al. Patterns of cocaine
binging. Effect on pregnancy. Am J Obstet Gynecol 1994;
171 :372-9.
16 Das G. Cocaine a buse and reprodu ction. Int J Ciin Pha/111aCol
Ther 1994 ; 32: 7- 11.
17 Sweet AY. Narcotic withdrawal syndrome in the newborn.
Pediatr ReI! 1982; 3; 285-91.
18 Bergman A, Weisner L. Relation of passive cigarette smoking
to sud den infant dea th syndrome. J Pediatr 1976; 58:665-8.
19 Wa rd SLD, Ba u tista D, Chan L, et al. Sudd en infant death
sy nd ro me in infants of substance-abu sin g mothers. J Pediatr
1990 ; 117: 8 76-81.
20 Meyer MB, Jonas BS , Tomascia JA. Perinatal events
associa ted with maternal smoking durin g pregnancy.
Am J Epidemio l 19 76; 103 :464-9.
21 Bar-Oz B, Bu lk owstein M, Benyamini L, et al. Use of
antibio tic and analgesic drugs during lactation. Drug Saj
2003; 26:925-35.
22 Koren G, Cairns J, Chitayat D, et al. Pharmacogenetics of
morphin e poisoning in a breastfed neon ate of a codei ne­
prescribed mother. Lancet 2006; 368:704.
23 Johnson TN. The development of drug metabolizing enzymes
and their influence on the susceptibility to adverse drug
reactions in children. Tox icology 2003; 192:37-48.
24 Ito S. Drug therapy for breast-feeding women. N Engl J Med
2000; 343:118- 26 .
25 Lopez-Munoz F, Alamo C, Quintero-Gutierrez FJ, Garcia­
Garcia P. A bibliometric study o f international scientific
productivity in attenti on -d eficit hy pera ctivity disorder
coverin g the period 1980- 2005. Eur Child Adolesc
Psychiatry, 2008 [epub a head of printJ.
26 Klein-Schwartz W. Abuse and toxicity of methylphenidate.
CUlT Opin Pediatr 2002; 14: 2 19-23 .
27 Bailey B, Letarte A, Abran Me. Methylphenidate
unintentional ingestion in preschool children. Ther Drug
Monir 2005; 27:284- 6.
29 DuRant RH, Escobedo LG , Heath Gw. Anabolic stero id use,
strength training and multiple drug use among adolescents
in the United States. Pediatrics 1995; 96:23-8.
30 Yesa lis C, Kennedy N, Kopstein A. Anabolic-androgenic
31 Madea B, Grellner W, Musshoff F, Dettmeyer R. Medico-l ega l
seniors. JAMA 1988; 260:344 1-5.
JJ Terney R, McLain LG. Th e use of an abo li c steroids in high
sc hoo l students. Am J Dis CI1i1d 1990; 144: 99- 103.
34 Hoffman RJ, Nelson L. Rational use of toxico logy testing in
children. Curr Opin Pediatr 2001; 13:183 - 8.
35 Madden MA. Pediatric poisonings: recogn ition, assessment,
a nd management. Crit Care Nurs Ciin North Am 2005;
17 :3 95 -404, xi.
36 Belson MG, Simon HK, Sullivan K, Geller RJ. The utility of
toxicologic a nalysis in children with suspected in ges tions.
Pedintr Emerg Care 1999 ; J 5:383-7.
37 Kind T. Carbon monoxide. Pedia tr Rei) 2005; 26:150-1.
38 Whi te ML, Liebelt EL. Update o n a ntid otes for pediatric
poisoning. Pedjatr Emerg Care 2006; 22: 740 - 6; quiz 7-9.
39 Heimann G. Enteral absorption and bioavailability in
children in relation to age. Eur J Ciin Pharmacol 1980;
18:43- 50.
40 Richardson T. Pitfalls in forensic toxicology. Al1n C/in
Biochem 2000; 37 (Pt 1):20-44.
276 I Recent advances in paediatric toxicology
41 Rane A, Tomson G. Prenatal and neonatal drug metabolism
in man. Eur J Clin Pharmacol1980; 18:9-15.
42 Hines RN. The ontogeny of drug metabolism enzymes a nd
implications for adverse drug events. Phannacol Ther 2008;
118:250-67.
43 So ldin SJ, Steele BW. Mini-rev iew: therapeutic drug
monitoring in pediatrics. Clin Biochem 2000; 33 :333 -5.
44 Sold in OP, Soldin SJ. Review : therapeutic drug monitoring
in pediatrics. Th er Drug Monit 2002 ; 24: 1-8.
45 Alco rn J, McNamara PJ. Pharmacokinetics in the newborn.
Ad/) Dru g Deliv Rev 2003; 55:667-86.
46 Allega ert K, Verbesselt R, Naulaers G, et a J. Developmental
pharmacology: neonates are not just small ad ults. Acta Clin
Belg 2008; 63:16-24.
47 Ward RM, Benitz WE, Benjamin DK, Jr. , et al. Criteria
sup portin g the study of drugs in the newborn . Clin Th e!"
2006; 28: 1385-98.
48 Johnson TN, Rostami-Hodjegan A, Goddard JM, et a J.
Co ntribution of midazolam and its I-hydroxy metabolite to
preoperative sedation in chi ldren : a pharmacokinetic­
pharmacodynamic ana lysis . Br J Anoesth 2002; 89:428-37.
49 Cuzzolin L, Atzei A, Fan os V. Off-l abel and unlicensed
prescribing for newborns a nd chi ldren in different settings: a
review of the literature and a consideration about drug
sa fety. Expert Opill Drug Safety 200 6; 5 :703- 18.
50 Shamoo AE, Tauer CA. Ethica lly q ues tion able research with
ch ildren: the fenfluramine study. Account Res 2002;
9: 143- 66.
51 Sharav VH. Children in clillical research: a conflict of moral
values. A m J Bioeth 2003; 3:W 12-W59.
52 Woolf AD, Shannon MW. Clinical toxicology for the
pediatrician. Pediatr Clill North Am 1995 ; 42:317-33.
53 George S. Posi tio n of immunological techn iques in screening
in clinical toxicology. Clin Chem Lab Med 2004;
42: J 288- 309.
54 Miyaish i S. lmmunoassays : theolY and in strumenta tion,
techniques, applicatio ns. In Townshend A (ed.) Encyclopedia
Analytical Sciences. Volume 5. London: Academic Press,
1995, pp. 2049-108.
55 Montagne M, Pugh CB, Fink. JL. Testing for drug use, pan 1:
analytical methods. Am J Hosp Pharm 1988; 45:1297-305.
56 Baselt RC. Urin e drug screenin g by immunoass ay :
interpretation of results. In Baselt RC (ed.) Adl1allces in
Analytical Toxicology. Foster City, CA: Biomedical
Publications, 1984, pp. 81-123.
57 Kolbrich EA, Kim I, Barnes AJ, et aJ. Cozart RapiScan Oral
Fluid Drug Testing System: an evaluation of sensitivity,
specificity, and effic iency for cocaine detection com pared
with ELISA and GC-MS following controlled cocai ne
administrati on. J Anal ToxicoI2003; 27:407-11.
58 Gjerde H, Christophersen AS, Skuterud B, et al. Screening for
drugs in forensic blood samples using EMJT urine assays.
Forensic Sci Int 1990; 44: 179-185.
59 Pichon V. Selective sample treatment using molecularly
imprinted polymers. J Chromatogr A 2007; 1152:41-53.
60 Souv era in S, Rudaz S, Veuthey JL. Restric ted access
ma teria ls and large particle suppo rts for o n-line sa mple
preparation: an attractive approach for biological fluids
an alys is. J Chro matogr BAnal Technol Biomed Life Sci
2004; 801:141-56.
61 Wille SM, Lamben WE. Recent developments in extraction
procedures relevant to analyti cal toxi cology. Ana l Bioana l
Chern 2007; 388:JJ8J-9 1.
62 Chen Y, Gu o Z, Wang X, Qiu C. Sample preparation.
J Chromatogr A 2008; 1184: 19 1-21 9.
63 Herzler M, Herre S, Pragst F. Selecti vity of substa nce
id entifi cati on by HPLC-DAD in toxicological ana lysis using
a UV spectra library of 2682 co mpo und s. J Anal Toxicol
2003 ; 27 :233 -42.
64 Pragst F, Herzler M, Erxleben BT. Systematic toxicological
ana lysis by high-performance liquid chrom atograp hy with
diode array detectio n (HPLC- DAD). Clin Chem Lab Med
2004; 42:1325-40.
65 Grobosch T, Lemm-Ahlers U. Immun oassay screenin g of
lysergic acid die thylamid e (LSD) and its confirmation by
HPLC a nd fluorescence detection following LSD ImmunElute
extraction. J Anal Toxicol 2002; 26: 181-6.
66 Moore CM. Drug test ing in the 90s. Tn Leikin SB, Paloucek FP
(eds) Poisoning alld Toxicology Handbook, 2nd edn.
Chicago, IL:Lexi-Comp inc; Hudson (Cleveland) and
American Pharmaceutical Associ atio n, 1996, pp. 11 62 -75.
67 Giro ud C, Michaud K, Spo rkert F, et a l. A fata l overdose of
cocain e asso ci ated with coin ges tion of marijuana,
bup ren orph ine, and fiu oxeti ne. Body flui d an d tissue
distribution of coca ine and its metaboli tes determined by
hydrophilic interact ion chro matograp hy- mass sp ectrometry
(HILlC-MS). J Anal Toxicol2004; 28:464-74.
68 Dienes-Nagy A, Rivier L, Giroud C, et al. Method for
quantification of morphine and its 3- and 6- glu curonid es,
codeine, codeine glucuronide and 6-m onoacetyJmorphine in
human blood by liquid chromatography- electrospray mass
spectrome try for routin e analysis in forens ic toxico logy.
J Chromatogr A 1999 ; 854:109-18.
69 Swartz ME . UPLC: An introduction and review. Journal of
Liquid Chromatography Et Rela ted Teclmologies 2005;
28: 1253-63.
70 Churchwell MI, Twaddle NC, Meeker LR, Doerge DR.
Improving LC-MS sensitiv ity thro ugh increases in
chromatographic performance: co mparisons of UPLC­
ES/MS/MS to HPLC-ES/MS/MS. J Chromatogr BAnal
Technol Biomed Life Sci 2005; 825 :JJ4-43.
71 Bugey A, Staub C. Appli ca tion of mon o lithic SUppOit S to
online extra ctio n and LC-MS analysis of benzodiazepines in
whole blood samples. J Sep Sci 2007; 30:2967 -78.
72 Segura J, Ventura R, Jurado C. Derivatization procedures for
gas chromatographic-mass sp ectro met ric determination of
xenob iot ics in biological samples, with spec ial attentio n to
drugs of abuse and dop ing agents. J Chromatogr B BioMed
Sci Applications 1998; 713:61- 90.
73 Gunnar T, Ariniemi K, Lillsunde P. Validated toxicological
determination of 30 drugs of abuse as optimized derivatives
in ora l fluid by long co lumn fast gas chro matogra phy/
electron impact mass spect rometry. J Mass Spectrom 2005;
40: 739-53.
74 Gun nar T, Ariniemi K, Lillsunde P. Fast gas chromatography­
negative- ion chemica l ionization mass spectrometry with
microscale volume sample preparation for the determination of
benzodiazepines and alph a-hydroxy metabolites, zaleplon and
zopiclone in whole blood. J Mass Spectrom 2006; 41:741-54.
75 Th omas A, Widmer C, Hopfgartner G, Stau b C. Fast gas
chromatography and negative-ion chemical ionization tand em
mass spectrometry for forensic a nalysis of cannabinoids in
whole blood. J Pharl17 Biomed A nal 2007 ; 45:495- 503.
76 Thormann W, Thormann W. Progress of capillary
electrophoresis in therapeu tic drug monitoring a nd clinical
and forensic toxicology. Th e?' Drug Manit 2002; 24: 222-3 1.
77 Thormann W, Zhang C, Schmutz A. Capillary electrophoresis
for drug analysis in body fl uids. Ther Drug Monit 1996;
18:506-520.
78 Vo n Heeren F, Thormann W. Capi llary electrophoresis in clinica l
and forensic analysis. Electrophoresis J997; 18 :241 5- 26.
79 Lloyd DK. Cap illary electroph oresis ana lysis of biofluid s with
a focus on less commonly ana lyzed matrices. J Chromatogr
BAnal Technol Bi0111ed Life Sci 2008; 866:154-66.

80 Smith ML, Vorce SP, Holler JM, et al. Modern instrumental
methods in forensic toxicology. J Anal Toxicol 2007;
31 :237-53, 8A-9A.
81 Maurer HH, Peters F1. Toward high-throughput drug
screening using mass spectrometry. Ther Drug Monit 2005;
27:686- 8.
82 Maurer HH. Hyphenated mass spectrometric techniques ­
indispen sable tools in clinical and forensic toxicology and
Cocaine and metabo lites in the hair of ancient coca leaf
in doping control. J Mass Spectrom 2006; 41: 1399- 4 13.
chewers. Forensic Sci Int 1993; 63:269-275.
83 Pa vlic M, Libi se ller K, Oberacher H. Combined use of E51­
QqTOF- MS and E51-QqTOF-MS/MS w ith mass-spectral
library sea rch for qualitative analysis of drugs. Anal Biaanal
C/wn 2006; 386:69-82.
84 Ojanpera S, Pelander A, Pelzing M, et al. Isotopic pattern and
accurate mass determination in urine drug screening by
liquid chromatography/time-of-flight mass spectrometry.
Rapid Commul1 iVIass Spectrom2006; 20:1161-7.
85 Hartmann C, Smeyers - Verbeke J, Massali DL, McDowall RD.
Validation of bioanalytical chromatographic methods.
J Pharm Bio111ed Anal 1998; 17:193-218.
86 Peters fl, Drummer OH, Musshoff F. Validation of new
and metabolites into hair: role of pigmentation. Drug Metab
methods. Forensic Sci Int 2007; 165:216-24.
Dispos 1996 ; 24:495-501.
87 Pichini S, Ventura M, Pujadas M, et al. HAIRVEQ: an
external quality control scheme for drugs of abuse analysis
in hair. Forensic Sci Int 2004; 145:109-15.
88 Wille SM, Lambeli WE. Volatile substance abuse - post­
III Mieczkowski 1. Assessing the potential of a 'color effect' for
mOliem diagnosis. Forensic Sci Int 2004; 142:135-56.
hair analysis of 11-nor-9-carboxy-d el ta(9)­
89 Gallardo E, Queiroz JA. The role of alternative specimens
tetrahydrocannabinol: analysis of a large sample of hair
in toxicological analysis. Biomed Chromatogr 2008; 22:
specimens. Life Sci 2003; 74:463-9.
795-821.
112 Mieczkowski T, Kruger M. Interpret ing the color effect of
90 Warner A. Interference of common household chemicals in
immunoassay methods for drugs of abuse. Ciill Chem 1989;
35:648-51.
91 Jaffee WB, Trucco E, Levy S, Weiss RD. Is this urine really
negative? A systematic review of tampering methods in
urine drug screening and testing. J Subst Abuse Treat 2007;
33:33-42.
92 Fraser AD, WOlih D. Urinary excretion profiles of 1I-nor-9­
carboxy-delta9-tetrahydrocannabinol. Study III. A delta9­
THC-COOH to creatinine ratio study. Forensic Sci Int 2003;
137:196-202.
93 Gottshalk LA, Cravey RH. Toxicological and Pathological
quantitation of cocaine and its major metabolites in human
Studies on PsychoactilJe Dnlg-inIJolved Deaths. Davis, CA:
hair by gas chromatography/chemical ionizat ion mass
Biomedical Publications, 1980.
spect rometry. J Anal Toxicol 1991; 15:260-65.
94 Druid H, Holmgren P. Compilations of therapeutic, toxic, and
fatal concentrations of drugs. J Toxicol Clin TO.ricol 1998;
36:133-4; author reply 5-6.
95 Schulz M, Schmoldt A. Therapeutic and toxic blood
117 Henderson GL Mechanisms of drug in corporatio n into hair.
concentrations of more than 800 drugs a nd o ther
Forensic Sci Int 1993; 63:19-29.
xenobiotics. Pharmazie 2003; 58:447-74.
118 Henderson GL, Harkey MR, Zhou C, et a l. Incorporation of
96 Blake MJ, Castro L, Leeder J5, Kearns G L Ontogeny of drug
metabolizing enzymes in the neo nate. Semill Fetal Neonatal
Med 2005; 10:123-38.
97 Blake MJ, Gaedigk A, Pearce RE, et al. Ontogeny of
dextromethorphan 0- and N-demethylation in the first year
of life. Clin Pharmacol Ther 2007; 81 :510-6.
98 Johnson TN. Modelling approach es to dose estimation in
children. Br J C/in Plwrmacol 2005; 59:663 - 9.
99 Johnson TN. The problems in scaling adult drug doses to
children. Arch Dis Child 2008; 93:207-11.
100 Walson PD. Therapeutic drug monitoring in special
populations. Clin Chem 1998; 44:415-9.
101 Moeller ME, Fey P, Sachs H. Hair anal.ysis as evidence in
forensic cases. Forensic Sci In! 1993 ; 63:43-53.
102 Kintz P, Tracqui A, Mangin P. Detection of dnlgs in human
hair for clinical and forensic applications. hIt J Legal Med
1992 ; 105:1-4.
References I 277
103 Sachs H. Forens ic applications of hair analysis. In Kintz P.
(ed.) Drug Testing in Hair. Boca Raton, FL: CRC Press. 1996;
211 - 222.
104 Kintz P, Villain M, Cirimele V. Determination of
trimeprazine-facilitated sedation in children by hair analysis.
J Anal Toxicol 2006; 30:400-2.
105 Springfield AC, Cartmell LW, Aufderheide AC, et a1.
106 Nakahara Y, Kikura R. Hair analysis for drugs of abuse. VII.
The incorporation rates of cocaine, benzoylecgonine and
ecgonine methyl ester into rat hair and hydrolysis of cocaine
in rat hair. ArchilJes Toxicol1994; 68:54-59.
107 Rollins DE, Wilkins DG, Gygi SP, et al. Testing for drugs of
abuse in hair. Experimental observations and indications for
future research. Forensic Sci Rev 1997; 9:23-36.
108 Pragst F, Balikova MA. State of the art in hair analysis for
detection of drug and alcohol abuse. Clin Chim Acta 2006;
370:17-49.
109 Gygi SP, Joseph RE, Cone EJ, et al. Incorporation of codeine
110 Wilkins DG , Haughey HM , Krueger GG, Rollins DE.
Disposition of codeine in female human hair after multiple
dose administration. J A/wi Taxieol 1995; 19:492-8.
melanin on cocaine and benzoylecgonine assays for hair
analysis: brown and black samp les compared. J Forensic Leg
Med 2007; 14:7-15.
113 Cone EJ , Joseph RE. The potential for bias in hair testing for
drugs of abuse. In Kintz P (eel.) Drug Testing in Hair. Boca
Raton, FL: CRC Press, 1996, pp. 69- 93.
114 Goldberger BA, Caplan YH, Maguire T, Cone EJ. Testing
human hair for drugs of abuse. N. Ident ificatio n of heroin
and 6-mo noacetylmorphine as indicators of heroine use.
J Allal Toxicol 1991; 15:226-31.
115 Harkey MR, Henderson GL, Zhou C. Simultaneous
lJ6 Kintz P, Cirimele V, Mangin P. Testing human hair for
cannabis. II. Identification ofTHC-COOH by GC-MS-NCI as a
unique proof. J Forensic Sci J995; 40:619-22.
isotopically-labeled cocaine and metabo lites into human
hair. 1. Dose-response relationships. J Analytical Toxicol
199 6; 20:1-12.
119 Nakahara Y, K.i.k.ura R, Takahashi F. Hair analysis for drugs of
abuse, X. Effect of physicochemical properties on incorporation
rates into hair. Bioi Pharm Bull 1995; 18:1223-7.
120 Naka hara Y, Kikura R. Hair analysis for drugs of abuse. XIII.
Effect of struc tural factors on incorporation of drugs into
hair: the incorporation rates of amphetamine analogs. Arch
Toxicol 1996; 70:841-9.
121 Blank DL, Kidwell DA. External contamination of hair by
cocaine: an issue in forensic interpretation. Forensic Sci Int
1993 ; 63: 145-56.
122 Cirimele V, Kintz P, Mangin. P. Drug concentrations in
human hair after bleaching. J Anal To.ricol 1995; 19:331-2.
123 Jo seph R, Su TP, Cone EJ. In lIirro binding studies of drugs
to hair: influence of melanin and lipids on cocaine binding
278 I Recent advances in paediatric toxicology
to caucasoid and africoid hair. J Anal Toxicol 1996;
20:338-44.
124 Hold KJV!, Wilkins DG, Crouch DJ, et al. Derection of
stanozolol in hair by negative ion chemical ionization mass
spectrometry. J Anal To.ricol 1996; 20:345-49.
125 Chiarotti M, Strano-Rossi S. Preparation of hair samp les for
drug ana lysis. Forensi c Sci Rev 1996; 8:111-2 8.
126 Irving RC, Dickson SJ. The detection of sedatives in Ilair and
nail sa mpl es using tandem LC-MS-MS. Forensic Sci lilt
2007; 166:58-67.
127 Mangin P, Kintz P. Variability of opiares concentrations in
human hair according to their anatomical origin: head,
axillary and pubic reg ions. Forensic Sci Jnt 1993; 63:77-83.
128 Offidani C, Strano-Rossi S, Chiarotti M. Dru g distribution in
the head, axillary and pubic hair of chronic addicts. Forensic
Sci lnt 1993; 63:1105-0.
129 Mangin P. Orug analysis in nonhead hair. In Kintz P (ed.)
Drug testing il1 hair. Boca Raton, FL: CRC Press, 1996,
pp. 279- 87.
130 Moeller MR, Eser HP. The ana lytical tools for Ilair resting. In
Kintz P (ed.) Drug testing ill hair. Boca RalOn, FL: CRC Press,
1996, pp. 95-120.
131 Hegstad S, Khiabani HZ, KJ'istoffersen L, et al. Drug
sc ree ning of hair by liquid chromatography-tandem mass
spectro merry. J Anal Toxical 2001l; 32:364-72.
132 Tracqui A. Unusu al drugs in hair. In Kintz P (eel.) Drug Testing
in Hair. Boca Raton, fL: CRC Press, 1996, pp. 191-210.
13 3 Idowu OR, Caddy B. A review of the use of sal iva in the
forensic detection of drugs and other chemicals. J Forensic
Soc 1982; 22:1 23 -235.
134 Caddy B. Saliva as a spec imen for drug analysis. In Baselt RC
(ed.) Ad/J in Analytical To.ricology. Vol 1. fo ster City. CA:
Biomedical Publi cations, 1984, pp. 198-254.
135 Schramm W, Smith RH, Craig PA, Kidwell DA. Dnlgs of
ab use in saliva: a review. J Anal Toxicol 1992; 16:1-9.
136 Dyer KR, Wilkinson C. The detection of illicit drugs in oral
fluid: another potential strategy to reduce illicit drug-related
harm. Drug Alcohol Re/J 2008; 27:99-107.
137 Dams R, Choo RE, Lambert WE, et al. Oral fluid as an
alternative matrix to monilO r opiate and cocaine use in
substance-abu se trea tment patients. Drug Alcohol Depend
2007; 87:258-67.
138 Samyn N, Laloup M, De Boeck G. Bioanalytical procedures
for determination of drugs of abuse in oral fluid. Anal
Bioanal Chem 2007; 388: 1437-53.
139 Cone EJ. Saliva test ing for drugs of abuse. Ann NY ACllri Sci
1993 ; 694:91-127.
140 Kintz P, Cirimele V, Lue\es B. Codeine tesring in sweat and
saliva with the drugwipe. lnt J Legal Med 1998; 111 :82- 4.
141 Dams R, Hu estis MA, Lambert WE, et al. Matrix effect in
bio-analysis of illicit drugs with LC-MSjMS: influence of
ionization type, sample preparati on, and biofluid. J Am Soc
Mass Sp ectrom 2003; 14: 1290-4.
142 Kintz P, Tracqui A, Jamey C, Mangin P. Detection of code ine
and phenobarbital in swea r collected with a sweat patch.
J Anal Toxicol1996; 20:197-201.
143 Cone EJ, Hillsgrove MJ, Jenk.ins AJ, et al. Sweat testing for
heroin, cocaine and metabolites. J Alla l Toxicol 1994;
18:298-305.
144 Tracqui A, Kintz P, Ludes B, et al. The detection of opiate
drugs in non traditional specimens (clothing): a report of ten
cases. J Forensic Sci 1995 ; 40:263-5.
145 Dolan K, Rouen D, Kimber J. An overview of rhe use of
urine, hair, sweat and saliva to detect drug use. Drug Alcohol
Rev 2004; 23:213-7.
146 Fogerson R, Sc hoendorfer D, Fay J, Spiehler V. Qualitative
detection of opiates in sweat by RIA and GC-MS.
J Analytical Toxicol 1997; 21 :451-8.
147 Burns M, Basel! RC. Mon itoring drug use with a sweat patch:
an experiment with coca ine. J Anal Toxicol 1995; 19:41-8.
148 Kintz P. Brenneisen R, Bundeli P, Mangin P. Swea t testing
for heroin and metabolites in a heroin maintenance
program. Clin Ch em 1997; 43:736- 9.
149 Moriya F, Chan KJV!, Noguchi n, WU PYK. Testing for drugs
of abuse in meconium of newborn infants. J AHal Toxicol
1994; 18:41- 5.
150 Lewis DE, Moore CM, Leikin JB, Koller A. M.econ ium
analysis for cocai ne: a va lidarion slUdy a nd comparison with
paired urin e ana lysis. J Anal Toxicol 1995; 19:148-50.
151 Oyler J, Darwin WD, Preston KL, et al. Coca ine disposition
in meconium from newborns of cocaine-abusing mothers
and urine of adult drug users. J Anlll Toxicol 1996;
20:453-62.
152 Moore C. Negrusz A, Lewis D. Determination of drugs of
abuse in meco nium. J Chrol11lltogr B Biol11ed Sci Appl 1998;
713: 137-46.
153 Gareri J, Klein J, Koren G. Drugs of abuse testing in
meconium. Clill Chim Acta 2006; 366:101 -11.
154 Gareri J, Klein J, Koren G, et al. Drugs of abuse testing in
meconium. Cli n Chim Acta 2006; 366:101-11.
J 55 Browne S, Moore C. Negrusz A, et al. Detection of cocaine,
l1orcocaine and cocaethylene in the meconium of premature
neonates. J Forensic Sci 1994; 39:1515-19.
156 Moore C, Lewis D, Becker J, Leikin J. The determinarion of
II-nor-69-tetrahydrocannabinol-9-carboxylic acid (THC­
COOH) in meconium. J Anal Toxicol 1996; 20:50-1.
157 Moore CM, Lewis DE, Leikin JB. The determination of
phen cyclidine in meconium using ion trap mass
spectromelly. J Forensic Sci 1996; 41: 1057-9.
150 Clark GD , Rosenzweig TB, Ra isys VA, et al. The analysis of
cocaine and benzoyJecgonine in meconium. J Anal TO.ricol
1992 ; 16:261-3.
159 Stolk LM, Coenradie SM, Smit BJ. Analysis of methado ne
and its primary metabolite in meconium. J Anal Toxicol
1997 ; 21:154- 9.
160 Dempsey D, Moore C, Deitermann D, et al. The detection of
cotinin e in hydrolyzed meconium samples. Forensic Sci Jnt
1999 ; 102:167-71.
161 Le NL, Reiter A, Toml illson K, er al. The detection of oxycoclo ne
in meconium specimens. J Anlll To.ricoI2005; 29: 54-7.
162 Williamson S, Jackson L, Skeoch C. et al. Determination of
the prevalence of drug misuse by meconium ana lysis. Arch
Dis Ch ild Fetal Neonatal Ed 2006; 91 :F291-2.
163 Xia Y, Wang P, Banlett MG, er al. An LC-MS - MS method for
the co mprehensive analysis of cocaine and cocaine
metabolites in meconi um . Anal Chem 2000; 72:764-71.
164 Kelly T, Gray TR, Hu estis MA. Development and va lidation
of a liquid chromatography-atmospheric pressure chemical
ionization-tandem mass spectrometry merhod for simultaneous
ana lysis of 10 amphetamine-, methamphetamine- and
3,4-methylenedioxym ethamp hetamine-related (MDMA)
ana lytes in human meconium. J Chrol11a togr BAnal Technol
Biomed Life Sci 2001l; 867: 194-204.
165 Moo re CM, Lewis D. fatty acid ethyl esters in meconium:
biomarkers for the detect ion of alcohol exposure in
neonates. Clin Chim Acta 2001; 312:235-7.
166 Moore C, Jones J, Lewis D, Buchi K. Prevalence of fatty acid
ethy l esters in meconium specimens. Clin Che l11 2003;
49: 133-6.
167 Ripple M, Goldberger EA, Caplan YH, et al. Detection of
cocaine and its metabolites in human amniotic fluid . J A lla l
To.ricoI1992; 16: 328 -31.
168 Moore C, Browne S, Tebbett I, et al. Determination of
coca ine and benzoylecgonine in human amn iotic fluid using
high flow solid-phase exrraction columns and HPLC.
Forensic Sci Jnt 1992; 56:177-81.

169 Winecker RE, Goldberger BA, Tebbett 1, et al. Detection of
cocaine and its metabolites in amniotic f1uid and umbilical
cord tissue. J Ana! To.rico!1997; 21:97~104.
170 Gray T, Huestis M. Bioanalytical procedures for monitoring in
litera drug exposure. Ana! Bioanal Chem 2007; 388:1455~65.
171 Kugelberg FC, Jones AW. Interpreting results of ethanol
analysis in postmortem specimens: a review of the literature.
Forensic Sci Int 2007; 165:1O~29.
172 Musshoff F, Madea B. Analytical pitfalls in hair testing. Anal
Bioana! Chem 2007; 388:1475~94.
173 Liu RH. Comparison of common immunoassay kits for
effective application in workplace drug analysis. Forensic Sci
Rev 1994; 6:20~57.
174 Green K, lsenschmid DS. Medical review officer interpretation
of urine drug test results. Forensic Sci Rev 1995; 7 :42~60.
175 Cone EJ, Johnson RE, Darwin WD, et al. Passive inhalation
of marijuana smoke: urinalysis and room air levels of delta-
9-tetrahydrocannabinol. J Anal To.rico! 1987; 11 :89~96.
176 Kidwell DA, Blank DL. Environmental exposure ~ the
stumbling block of hair testing. In Kintz P (ed.) Drug testing
in hair. Boca Raton, FL: CRC Press, 1996, pp. 17~68.
177 Haley NJ, Hoffmann D. Analysis for nicotine and cotinine in
hair to determine cigarette smoker status. Clin Chem 1985;
31:1598~1600.
178 Kintz P, Ludes B, Mangin P. Evaluation of nicotine and
cotinine in human hair. J Forensic Sci 1992; 37:72~6.
179 Smith FP, Kidwell DA. Cocaine in hair, saliva, skin swabs
and urine of cocaine users' children. Forensic Sci lilt 1996;
83:179~189.
180 Blank DL, Kidwell DA. Decontamination procedures for
drugs of abuse in hair: are they sufficient? Forensic Sci Int
1995; 70:13~38.
181 Davis KH, Hawks RL, Blanke RV. Assessment of laboratory
quality in urine drug testing ~ a proficiency testing pilot
study. JAIVIA 1988; 260: 1749~54.
182 Kelly K. The accuracy and reliability of tests for drugs of
abuse in urine samples. Pharmacotherapy 1988; 8:263~75.
183 Badia R, Segura J, Artola A, de la Torre R. Survey on drugs­
of-abuse testing in the European Union. J Ana! To.ricol
1998; 22: 117~26.
184 McRorie TI, Lynn AM, Nespeca MK, et al. The maturation of
morphine clearance and metabolism. Ani J Dis Child 1992;
146:972~6.
185 Furlanut M, Benetello P, Baraldo M, et al. Chlorpromazine
disposition in relation to age in children. Clin
Pharmacokinetics 1990; 18:329~3J.
186 Mikkelsen SL, Ash KD. Adulterants causing false negatives
in illicit drug testing. Clin Chem 1988; 34:2333~6.
187 Cody JI, Schwarzhoff H. Impact of adulterants on RIA
analysis of urine for drugs of abuse. J Ana! Toxico! 1989;
13:277~284.
188 Schwarzhoff R, Cody JT. The effects of adulterating agents
on FPlA analysis of urine for drugs of abuse. J Anal Toxicol
1993; 17:14~17.
189 Soper JW, Canfield V. Description of a screening assay
system to detect adulteration of urine samples. J Anal
Toxicol 1996; 20:77.
190 Goldberger BA, Caplan YH. Effect of glutaraldehyde
(UrinAid) on detection of abused drugs in urine by
immunoassay. Clin Chem 1994; 40: I 605.
191 Cody JT, Valtier S. Effects of Stealth adulterant on
immunoassay testing for drugs of abuse. J Anal To.ricol
2001; 25:466~70.
192 Burrows DL, Nicolaides A, Rice P J, et al. Papain: a novel
urine adulterant. J Anal To.ricoI2005; 29:275~95.
193 Pearce MR, Tarnai.LD. Performance evaluation of three
on-site adulterant detection devices for urine specimens.
J Anal Toxico! 2002; 26:464~70.
References I 279
194 Shepherd MF, Lake KD, Kamps MA. Postmortem changes and
pharmacokinetics: review of the literature and case report.
Ann Pharmacother 1992; 26:51O~14.
195 Hilberg T, Ragde S, Morland J. Postmortem drug
redistribution ~ human cases related to results in
experimental animals. J Forensic Sci 1999; 44:3~9.
196 Pelissier-Alicot AL, Gaulier JM, Champsaur P, Marquet P.
Mechanisms underlying postmortem redistribution of drugs:
a review. J Anal To.ricol 2003; 2 7 :533~44.
198 Drummer OH. Postmortem toxicology of drugs of abuse.
Forensic Sci Int 2004; 142:!oI~J3.
198 Skopp G. Preanalytic aspects in postmortem toxicology.
Forensic Sci Int 2004; 142:75~!OO.
199 Yarema MC, Becker CEo Key concepts in postmortem drug
redistribution. C!il1 Toxicol (Phila) 2005; 43:235~41.
200 Drummer OH. Post-mortem toxicology. Forensic Sci Int
2007; 165:199~203.
201 Pounder DJ, Jones GR. Post-mortem drug redistribution ~ a
toxicological nightmare. Forensic Sci Int 1990; 45:253~63.
202 Gomez HF, McKinney P, Philipps S, et al. Postmortem
acetaminophen pharmacokinetics: an experimental study of
site and time-dependent concentration changes. J Forensic
Sci 1995; 40:980~2.
203 Hilberg T, Morland J, Bjorneboe A. Postmortem release of
amitriptyline from the lungs: a mechanism of postmortem
redistribution. Forensic Sci Int 1994; 64:47~55.
204 Koren G, KJein J. Postmortem redistribution of morphine in
rats. Ther Drug Manit 1992; 14:461~3.
205 Koren G, MacLeod SM. Postmortem redistribution of digoxin
in rats. J Forensic Sci 1985; 30:92~6.
206 Quatrehomme G, Bourre F, Liao Z, Ollier A. An experimental
methodology for the study of postmortem changes in toxic
concentrations of drugs using secobarbital as an example.
J Forensic Sci 1994; 39:1300~4.
207 Imamura T, Nagat T, Kimura K, et al. Pharmacokinetics and
postmortem changes of cimetidine in body tissues. Nippon
Hoigaku Zasski 1994; 48:75~78.
208 Aderjan R, Mattern R. Validity of digoxin concentrations in
blood determined postmortem. ZeitschriJl Jiir Reci1tsmedizin
1980; 86: !3~20.
209 Hearn WL, Keran EE, Wei HA, Hime G. Site-dependent
postmortem changes in blood cocaine concentrations.
J Forensic Sci 1991; 36:673~84.
210 Levine 8, Wu SC, Dixon A, Smialek JE. Site dependence of
postmortem blood methadone concentrations. Am J Forensic
Med Patho11995; 16:97~!oO.
211 Miyazaki T, Kojima J, Yashiki M, et al. Site dependence of
methamphetamine concentrations in blood samples collected
from cadavers of people who had been methamphetamine
users. Am J Forensic Med Patho11993; 14:121~4.
212 Kuhlman JJ, Mayers RW, Levine B, et al. Chloroquine
distribution in postmortem cases. J Forensic Sci 1991;
36: 1572~9.
213 Pounder DJ, Hartley AK, Watmough P J. Postmortem
redistribution and degradation of dothiepin: human case
studies and an animal model. Am J Forensic jl;Ied Pathol
1994; 15:231~5.
214 Pounder DJ, Owen V, Quigley C. Postmortem changes in
blood amitriptyline concentration. Am J Patho! 1994;
15:224~30.
215 Pohland RC, Bernhard NR. Postmoliem serum and tissue
redistribution of f1uoxetine and norf1uoxetine in dogs
following oral administration of f1uoxetine hydrochloride
(Prozac"). J Forensic Sci 1997; 42 :812~ 16.
2J 6 Yonemitsu K, Pounder DJ. Postmortem changes in blood
tranylcypromine concentration: competing redistribution
and degradation effects. Forensic Sci Int 1993; 59:
177~84.
280 I Recent advances in paediatric toxicology
217 Yonemitsu K, Pounder DJ. Postmortem toxicokinetics of
co-proxamol. Jnt J Legal Med 1992; 104:347-53.
218 Hanzlick R. Postmortem tricyclic antid epressa nt
concentrations: letha l versus nonlethal levels. Am J Forensic
Med Pathol 1989; 10:326-9.
219 Martin A, Pounder DJ. Postmortem toxicokinetics of
trazadone. Forensic Sci ll7t 1992; 56:201-7.
220 Pounder DJ, Davies JI. Zopiclone poisoning: tissue
distribution and potentia l for postmoltem diffusion. Forensic
Sci Jnt 1994; 65:177-83.
22 1 Hilberg T, Bugge A, Beykich KJV!, et at. Diffusion as a
mechanism of postmortem drug redistribution: an
experimental study in rats. lut J Leg Med 1992; 105:87-91.
222 Pounder DJ, Fuke C, Cox DE, et al. Postmortem diffusion of
drugs from gastric residue. Am J Forensic Med Pat/IOI 1996;
17: 1-7.
223 Pounder DJ, Smith RW. Postmortem diffusion of alcohol
from the stomach. Am J Forensic Med Pathol 1995;
16:89-96.
224 Skopp G, Potsch L, Kon ig I, Mattern R. A preliminalY study
on the stab ili ty of benzod iazepi nes in blood and plasma
stored at 4' C.lnr J Leg Med 1998; 111:1-5.
225 Mayer W, Erbe S, WolfG, Voigt R. Beitrage zur Analytik und
Stabilitat einiger pharmazeutisch interessanter 1,4­
Benzodiazepine. Pharl1lazie 1974; 29: 700-7.
226 Koves EM, Lawrence K, Mayer JM. Stability of diltiazem in
w ho le blood: forensic implications. J Forensic Sci 1997;
43:587- 97.
227 Benhamou-Batut F, Demotes-Mainard F, Labat L, Vincon G.
Determination of flunitrazepam in plasma by liquid
chromatography. J Pharm Biomed Anal 1994 ; 12:93 1- 6.
228 Augsburger M, Mangin P. LSD, Ie phenix des hallucinogenes.
Toxicorama 1998; 10:61-6.
229 Ferner RE. Forensic Pharmacology. Oxford, NF: Oxford
University Press, 1996.
230 Paterson S. Drugs and decomposition. Med, Sci and Law
1993; 33: 103- 109.
231 Hadidi FA, Ol iver JS. Stability of morphine and buprenorphine
in whole blood.lnt J Leg Med 1998; 111:165-7.
232 McCurdy HH, Callahan LS, Williams RD. Studies on the
stability and detection of cocaine, benzoylecgonine and
I I -nor-69-tetrahydrocannabinol-9-carboxylic acid in
who le blood using Abuscreen radioimmunoassay. J Forensic
Sci 1989; 34:858-70.
233 Robertson MD , Drummer OH. Postmortem drug metabolism
by bacteria. J Forensic Sci 1995 ; 40:382-6.
234 Robertson MD, Drummer OH. Stability of
nitrobenzodiazepines in postmortem blood. J Forensic Sci
1998; 43 :5- 8.
235 Robertson MD, Drummer OH. Postmortem distribution and
redistribution of nitrobenzodiazepines in man. J Forensic Sci
1990; 43:9-13.
236 Corry JEL. Possible sources of ethanol ante- and
postmortem: its relationship to the biochemi stry and
microbiology of decomposition. J App! Bacterial 1978;
44: 1-56.
237 Hoiseth G, Kristoffersen L, Larssen B, et al. In uitro
formation of ethanol in autopsy samples containing fluoride
ions. lnt J Legal Med 2008; 122:63-6.
238 Schm itt G, Adeljan R, Keller T, Wu M. Ethy l g lu curonide: an
unusual ethan ol metabolite in hum ans. Synthesis, analytical
data, a nd determination in serum and urine. J Anal Toxicol
1995 ; 19:9 /-4.
239 Bicker W, Lammerhofer M, Keller T, et a l. Validated method
for the determination of the ethanol consumption markers
ethyl glucuronide, ethyl phosphate, and ethyl sulfate in
human urine by reversed-phase/weak anion exchange liquid
chromatography-tandem mass spectrometly. Anal Chem
2006; 78:5 884-92.
240 Cap lan YH, Levine B. Vitreous humor in the eva luation of
postmoltem blood ethanol concentrations. J Anal To.ricol
1990; 14:305-7.
241 Pounder DJ, Kuroda N. Vitreous alcohol is of limi ted value
in predicting blood alcohol. Forensic Sci lnt 1994; 65:73-80.
242 Laure P, Lecerf T, Friser A, Binsinger C. Drugs, recreational
drug use and attitudes towards doping of high school
athl etes. Jnt J Sports lVIed 2004; 25:133 - 8.
243 Laure P, Binsinger C. Doping prevalence among
preadolescent athletes: a 4-year follow-up. Br J Sports Med
2007 ; 41 :660-3; discussion 3.
244 Sa ugy M, Avois L, Saudan C, et al. Cannabis and sport. Br
J Sports Med 2006; 40 (Suppl 1):i13-5.
245 Bowers LD. Analytical advances in detection of
performance-enhancing compounds. Clin Chem 1997;
43: 1299-304.
246 Cowan DA, Kickman AT. Doping in sport: misuse, analytical
tests, and legal aspects. Clin Chem 1997; 43:1110-13.
247 Thevis M, Schanzer W. Mass spectrometry in sports drug
testing. Structure characterization and analytica l assays.
Mass Spectrom Rev 2007; 26:79-107.
248 Saudan C, Baume N, Robinson N, et al. Testosterone and
doping control. Br J Sports Med 2006; 40 (Suppl l):i21-4.
249 Dawson RT. Drugs in sport - the role of the physician.
J Endocl'illo/ 2001; 170:5 5-61.
250 Sjoqvist F, Garle M, Rane A. Use of dop in g agents ,
palticularly anabolic steroids, in sports and society. Lancet
2008; 37 1:1872-82.
251 Musshoff F, Daldrup T, Ritsch M. Black market in a nabolic
steroids - analysis of illegally distributed products.
J Forensic Sci 1997; 42:1119 - 25.
252 Catlin DH, Hatton CK, Starcevic SH. Issues in detecting
abuse of xenobiotic anabolic steroids and testosterone by
analysis of athletes' urine. Clin Chem 1997; 43: 1200-8.
253 Thevis M, Thomas A, Schanzer W. Mass spectrometric
determination of insulin s and their degradation products in
sports drug testing. Mass Spectrom Rell 2008; 27:35-50.
254 Huestis MA. Technical and legal aspects of drugs of abuse
testing in hair. In Nntz P (ed.) Drug testing ill hair. Boca
Raton, FL: CRC Press , 1996 , pp. 5-15.
255 Lewis D, Moore C, Morrissey p, Leikin J. Determination of
drug exposure using hair: application to child protective
cases. Forensic Sci Jnt 1997; 84:123-8.
256 Strano-Rossi S, Offidani C, Chiaroll i M. Application of hair
analysis to document coercive heroin administra tion to a
child. J Ana/ Toxico/1998; 22: 75-7 .
257 Xiang P, Shen M, ZlllI o X. Hair analysis for ketam ine and its
metabolites. Forensic Sci lnt 2006; 162: 131-4.
258 Marchei E, Munoz JA, Garcia-AJgar 0, et al. Development
and validation of a liquid chromatography-mass
spectrometry assay for hair analySis of methylphenidate.
Forensic Sci Tnt 2008; 176:42-6.
259 Kintz P, Villain M, Tracqui A, et al. Buprenorphine in drug­
facilitated sexual abuse: a fatal case involving a 14-year-old
boy. J Anal Toxicol2003; 27:527-9.
260 Negrusz A, Gaensslen RE. Analytica l developments in
toxicological investigation of drug-facilitated sexual assault.
Anal Bioanal Chem 2003; 376:1192-7.
261 Cheze M, Villain M, Pepin G. Determination of bromazepam,
c!onazepam and metabolites after a single intake in urine
and ha ir by LC-MS/ MS. Application to forensic cases of drug
facilitated crimes. Forensic Sci Jil t 2004; 145:123-30.
262 Deveaux M, Cheze M, Pepin G. The role of liquid
chromatography-tandem mass spectro metry (LC-MS/MS) to
test blood and urine samp les for the tox icologica l

in ves tigation of drug-facilitated crimes. Ther Drug Manit
2008; 30:225-8.
263 Rey-Sa lmon C, pepin G. (Drug-facilitated crime and sexual
abuse: a pediatric observation]. Arch Pediatr 2007;
14: 13J 8-20.
264 Kintz P, Evans J, Vill a in M, et a1. Hair analysis for
diphenhyd ramine after surreptitious admin istration to a
child. Forensic Sci 1nt 2007; 173: 171-4.
26 5 Kintz P, Villain M, Cirimele V. Hair ana lysis for drug
detection. Thn Drug Manit 2006 ; 28:442-6.
266 Kin tz P, Vi llain M, Dumestre-Toulet V, et a J. Methadone as a
chemical weapo n: two fatal cases involv ing babies. TIler
Drug Manit 2005; 27:741-3.
267 Kintz p, Cirimele V, Jamey C, Ludes B. Testing for GHB in
hair by GCjMSjMS after a single ex posure. Application to
document sexual assaul t.] Forensic Sci 2003; 48:19 5-200.
268 Ostrea EM, Chavez CS. Perinatal problems (excluding
neonatal withdrawal) in materna l drug addiction : a study of
830 cases.] Pediatr 1979 ; 94:292 - 5.
269 Ostrea EM, Brady M, Gause S, et a J. Drug sc reen in g in
newbo rns by meconium ana lysis: a large-scale prospective,
epid emiologic study. ] Pediatr 1992; 89: 107 - 13.
270 Klein J, Koren G. Hair analysis - a biological marker for
passive smok ing in pregnan cy and childhood. Hum Exp
Toxicol 1999; 18 :279-8 2.
271 Kukla L, Hruba D, Tyrlik M. Trends in respiratory morbidi ty
of children in relation to their passive smoking exposure.
Cent Eur] Public Health 2006; 14:180- 5.
272 Bar-Oz B, Klein J , Karasko v T, Koren G. Comparison of
meco nium and neonatal hair ana lysis for detection of
gestational exposure to drugs of abuse. Arch Dis Ci1ild Fetal
Neonata l Ed 2003; 88:F98-FIOO.
273 Garcia-BoUl'nissen F, Rokach B, Karaskov T, Koren G.
Methamphetamine detection in maternal and neonatal hair:
impli cations for fetal safety. Arch Dis Child Fetal Neol1a tal
Ed 2007; 92.F35 1-5.
References I 281
274 Garcia-Bournissen F, Rokach B, Karaskov T, Koren G.
Cocaine detection in maternal and neonatal hair:
implic ations to fetal toxicology. Ther Drug Manit 2007;
29:71- 6.
27 5 Pragst F, Yegles M. Determination of fatty ac id ethyl esters
(FAEE) and ethyl glucuronide (EtG) in hair: a promising way
for retrospective detection of alcohol abuse during
pregnan cy? Ther Drug Manit 2008; 30:2 55-63.
276 Moore C, Dempsey D, Deitermann D, et aJ. Feta l cocaine
exposure: analysis of vernix caseosa. ] Anal Toxicol 1996;
20:509-11.
277 Marks V, Richmond C. Beverly Allitt: the nurse who killed
babies.] R Soc Med 20 08; 101 :110-5.
278 Dyer C. Children's nurse convicted of murd er. EM] 1993;
306:1 43 1-2.
279 Appleyard WJ. Murder in the NHS. EM] 1994; 308:287-8.
280 Repper J. Munchausen syndrome by proxy in health care
workers.] Adll Nlirs 1995; 2 1 :299-304.
28 1 Jones DPH, Boots CN. Factitious lJIness by Proxy. In David TJ
(ed.) Recel1t AdIJances in Paediatrics J 7. London : Churchi ll
Li vingstone, 1999, pp. 57-71
282 Jenkins RO, Craig P J , Goessler W, Irgolic KJ. Antimony
leaching from cot mattresses and sudden infant death
syndrome (SJDS). Hum E.rp Toxicol 1998; 17 :138- 9.
283 Richardson BA. Cot mattress biodeterioration and SJDS.
Lal1cet 19 90; 335:670.
284 Richardson BA. Sudden infant death syndrome: a possible
primary cause. ] Forensic Sci Soc 1994; 34: 199- 204.
285 Department of Health. Expert Group to In ves tigate Cot Delllh
Theories: To.ric Gas Hypothesis. Chairman, Lady Limerick.
Final report. London: Department of Health, 1998.
286 Report of the expert working group enquiring into the
hypothesis that toxic gases evolved from matress covers and
cot matresses are a ca use of SIDS. Sudden infallt death
syndrome (STD S) London: Her Majesty's Sta tionery Office,
1991.
I CHAPTER 14 I
HEAD AND NECK INJURIES
Robert A Minns and TY Milly Lo
Definition 282
Epidemiology 283
Non-accidental head injury 294
Traumatic birth injury 300
Primary mechanisms of injury to the brain 302
DEFINITION
Yager et al 9 assessed the interobserver variability for six
Head injury can be defined as the injury resulting from an
external force to the head by penetrating, compression or
impact forces, causing damage to the scalp, skull or brain
(traumatic brain injury). If brain injury results, there is usu­
ally an impairment of consciousness and a period of retro­
grade amnesia. Not all cases of brain damage are associated;
however, with a loss of consciousness (e.g. some frontal lobe
injuries, extradural haematomas and some penetrating an d
compression injuri es) but, conversely, all cases of loss of
consciousness indi cate brain injury.
Traumatic brain injury in children may result from
traumatic birth injury, accidental head injury and non­
accidental head injury (NAHI).
The Gl asg ow Co ma Scale (GCS) developed by Teasdale
and Jennett i in 1974 has become a widely used clinical
tool for assessment of t he level of co nsciousness in adults.
Furthermore, the GCS enab les a basa l initial sco re to be
estab lished and thus allows for comparison between suc­
cessive examinations of the level of consciousness identi­
fying any clinical deterioration. Many of the responses
assessed in the GCS requ ire an adult level of neurodevelop­
mental function and ca nnot be easily graded in children
who are under 10 years of age. 2 Modifications of the GCS
or other coma scales that include age- adjusted verbal and
motor responses have been described and used to assess the
level of consciousness in .chi ldren with neurodevelopm en­
tal levels who are und er 10 years of age. 2- 8
Secondary mecha nisms of brain injury 307
Injury to the cervical spinal cord 311
Genetic influence on recovery from
traumatic brain injury 312
References 313
different coma scales used in children an d found tha t
the paediatric coma scale had the highest interobserver
agreement. In an attempt to standardi ze the way coma is
measured in children, the British Paediatric Neurology
Association (BPNA) has recommended the Jam es ad apta­
tion of the GCS as it takes into account the developmental
immaturity in small chi ldren, uses the sa me number of
points irrespective of the child 's age, and is simple to use.
As the co nsciousn ess level decreases, vital mechanisms
such as airway protective reflexes (gag, cough and swa llow
reflexes) may be compromised, m aki ng the patient vulner­
able to aspiration 10 and secondary hypoxic ischaem ic
injury. The Ad va nced Paediatric Life Suppo rt Guideline
suggests that all head-injured chi ldren with a coma score
of less than 8 should have endotracheal intubation (to pro­
tect their airways) and ventilatory suppOl1. Children wi th
mild (G CS 13-15) to moderate (GCS 9-12) head injury may
have impaired gag reflexes and wou ld also be at risk of
developing aspi ration, and may also require intubation and
mechanical ventilation.
In intubated children, a grimace score has been sug­
gested to replace the ve rb al component of the GCS assess­
ment. II This modified paediatric coma sca le for intensive
care practice origin ated from the Sharples' adaptation of
the James ' adaptation of the GCS. Non-verbal communica­
tion and verbal language are not totally independent, and
an important part of non-verbal commu nication invo lves
facial expression and grim ace. Intero bserver reliability was

moderate to good, with all components assessed with the
grimace score being better than the verbal score in all of
the 73 children studied; this suggested that the grimace
score may be more useful in intubated patients. II
The most common cla ssification for se1!erity of head
i njLI1Y in adults is by means of the GCS, I in which a score
of 13-15 is designated mild, a score of 9-12 is designa ted
moderate and a sco re of 3-8 is designated severe. Modifica­
tions of the GCS for children are similarly used to assign
severity. The duration of unconsciousness or post-traumatic
amnesia have also been used to classify severi ty. Post­
traumatic amnesia may be four times as long as the
duration of coma . Severity based on imaging is useful only
for severe injury. Therefore, a GCS, which is reall y a mini­
neurological exa mination , remains the most useful index
of severity of brain injury.
EPIDEMIOLOGY
Accidental Head Injury
Limitations to comparing literature data on head injury
include:
1. The interna tional Classification of Diseases (lCD), 9th
edition, covers head injury with 10 codes that are not
mutually exclusive.
2. The severity of inju ry is difficult to identify from ICD
codes with no reference to impairment of
consciousness.
3. Application of these codes may be different in different
hospitals a nd countries, and coding errors are frequent.
4. Head-injured patients a re frequently transferred to
other hospitals, causing difficulties in defining a
catchment population.
5. Reports from pathologists or from trauma units do not
usually allow denominator identifica tion , thus
calcul ation of incidence varies depending on whether
the numerator is deaths, admissions or atte ndances.
Admission and triage policies vary widely from
hospital to hospital.
Deaths from head injury at all ages have been falling
since 1968 12 in the United Kingdom and in the Uni ted
States. 13 In most countries, the peak incid ence is in males
in the 15- to 30-year age group. Age-specific admiss ion
rates are similar in Britain and in the USA and an overall
estimate is 200-300 per 100000 ; 80 per cent are catego­
rized as mild and 5-10 per cent as severe.
The literature perta ining to the incid ence of accid ental
head i~UJy in children is seen in Table 14.1, which sum­
marizes th e nature of the study, the study period, t he age
group, inclusion criteria, mea n incidence and mortality
rate, along with the main causes for traumatic brain injury
(TBI). In Scotland, the incidence of severe TB[ is between
14.5 and 29 per 100000 child ren per year. There are more
Epidemiology I 283
ARE presentations due to head injury in children less than
10 years than in any other age group. I 8
Head injury is the co mmon est ca use of death in children
aged 1-15 years in England and Wales. 41 The mean mortal­
ity from head injury in children calculated from 18 pub­
lished studies was 5.1 ± 2.39 per 100000 children per
year.9.14-1 6, 18,19,21,24-32,34-39 Although accidental lIlJury
mortality rates have fallen in the UK, groups of children
who are at greater relative risk remain within the popula­
tion, such as children residing in less affluent areas. 31,41 [n
Scotland, accidents to pedestrians were the leadin g cause
of head injury mortality in children.
Sample statistics for children with head injury attending
ARE at the Royal Hospital for Sick Children in Edinburgh,
over two I-year periods, are seen in Fig. 14.1. 42 The peak age
group was between I and 2 years of age, with a steady decline
in attendance thereafter up to 13 years. Falls accounted for
53 per cent of all attendances, with fails of more than I metre
in height accou nting for 12 per cent. Road traffic accidents
were responsible for 4.6 per cent of attendances; the remain­
der (42 per cent) was due to miscellaneous causes.
Non-accidental Head Injury
Non-accidental brain injury refers to inflicted TBI (traumatic
en cephalopathy), usually by an adult, to infants and young
children, sustained as a resul t of deliberate impact (accelera­
tion or deceleration), head compression, penetrating head
injury, repetitive rotational injury (shaken baby syndrome),
rotation and impact (shaken impact syndrome) or whiplash
(cervicomedullary syndrome). Anyone mechanism or com­
bination of mechanisms may be present in the individual
child. The term 'shaken baby syndrome' has often been used
as a generic term for NAHI, and this has caused confusion in
lega l proceedings as it implies shaldng as the cause of all
NAHls. The term 'non-accidental head injury' is to be pre­
ferred, as no mechanism of injury is thereby infened.
The limitatio ns to obtaining accurate data on NAHI are
greater than those for accidental hea d injury data, because
of the follo wing factors : diagnostic difficulties, protracted
legal processes and outcome, and the mobile nature of this
population, making it difficult to asce rtain true, proven,
numbers of cases of NAHI. It is essential that information
pertaining to suspected cases is not accepted as proven
unless its no n -accidental origin has been acknowledged or
ha s resulted in a criminal conviction 43
From July 1998 until December 1999, 19 new cases of
suspected NAHI were identified in Scotland (12 boys a nd
seven girls), giving an annual incidence of 24. 6 per 100000
children under than 1 year of age (95 per cent confidence
intervals, 14.9- 38.5). The median age at acute admission was
2.2 months (range 4 weeks to 8.8 months). No child was older
than 1 year and 75 per cent of the children were admitted
during the autumn and winter months. A higher incidence
was found in the urba n areas of greater Glasgow and
 Ta bl e 14. 1 Studies of the incidence of accidental traumatic brain injury (TBI) in children
Reference Country/region Study Study period Age group Inclusion criteria Mean incidence Mortality Main causes Main ca use of TBI
for TBI ofTBI rate ofTBI (Ofo) mortality
---- -­
Annegers et al Olmsted County. Retrospective 1965-1974 Al l ages, Evidence of brain Males 270 Overall 35.0 <5 years - falls
1980 14 Minneso ta, USA Medical age- trauma Females 116 (age-adjusted) Gi rls 5-14 yea rs ­
linkage system adju sted Co ncussion with From graph: all From graph: horse riding,
LaC, PTA, 0-5 years, 180; 0-5 years, 8.0; cycling
neurologica l signs boys 5-15 years, 5-15 yea rs, 8.0 All boys 21
Sku II fracture 310; All girl s 11
girls 5-15 years, 160 Main ly ages 5-14 years
Rantaka lJ io, Finland Prospective 0-14 yea rs ~ 161years
von Wend t, 1985 15 birth cohort 235 cumulative
Kraus et aI California, USA Consecutive Children TBI 180
1986 16 Review of several
cross-sectional
studies

Cha n et al Sydney, Australia Retrospective January 198 5 0-15 years Admitted to hospital 13 per cent 'serious 2 Not given Pedestrians, MVC
1989 " (W Metropolitan Paediatric to December with trauma. GCS, TS, injuries' with ISS ;;;' 16 occupants, cyclists
Region) Trauma Study 1986 ISS noted 90 per cent
Se rious: TS ". 12 involved cranium
Serious TB I 17
Brookes et al Scotland, UK Retrospective 198411985 Adults, Evidence of brain Admission rate 4011 Children constitu te: 40­ MVC 71%
1990 18 Case record s Children damage: altered Admission with brain 50% of ARE attenders,
for 23 Scottish < 5 yea rs, consciousness on injury 290 31 % of admissions,
ARE 5-11 yea rs, arriva l or history Most seve re 5-10 25% of neurosurgical
departments 12-14 years of altered per ce nt unit tra nsfers, 200/0 of
consciousness and 14.5-29 death s on neurosurgical
am nesia on arrival un it, 11 0/0 of operations
for ICH, 750/0 of severe
injuri es in children
are MVC
Yager et al 19909 Virginia, USA 1978 All ages 175 all 5.3 chi ldren
Sha"rpl es et al N Engl and, UK Retrospective 1979-1986. < 16 yea rs All deaths from head 5.3 HI was major sing le
1990 19 review of HI in c. injury in child re n ICD Mortality cause of deaths in
deaths 9 codes N800- NB04, significa ntly related children (1 5% ) 53 % of 25 5 fatal
N850-N854 to depriva tion index: 250/0 deaths 5-1 5 yea rs HI accidents
14.0 with greatest 49% di ed before admission occurred during
deprivation, 0.9 with Most deaths (76% ) due play
least deprivation to MVC: 53% pedestrians,
(P < 0.00001) 140/0 cyclists, 100/0
passengers, 120/135
pedestrian deaths from
unsafe child beh avio ur
lepd> cL al U)A Helrospeclive Not given National Pediatri c 44Ofo ha ve HI Mortality was Fall s
1990 20 Trauma database Trauma Regi ster, lower in children MVC, ped estrian /
of children identified head injury occupant 3.10/0 of 10098
alone, head plu s TBI admissions
extracranial inju ry, only
and extracranial injury
Hansen et al Denmark, 0-14 yea rs Severe TBI 18
1991" Ringkobin g county

Levin et al USA: Retrospective, 14 April 1983 ,,"15 years Severe traumati c head 6 months po st­ 0-4 years: 31 % fall/ju mp; Not give n
1992 22 San Diego, affluent hospital-based t o 18 April banded : inju ry with in 48 injury mortality 26 0/0 pedestrian; 20%
urban area; 1988 0-4 years hours of approximatel y motor vehicle occupant;
Virginia, inner-city, 5-10years admission, 52 0/0 0-4 years; 17 0/0 assault
suburban , rural ; 11-15 GCS"" 8 180/0 5-10 years; 5-10 years: 49 0/0 motor
Texas, large urban, years Gunshot wounds 22Ofo /1-15 years veh icle occupant; 26%
rural beach resort; excluded pedestrian; 130/0 fall/jump;
Houston, urba n; 9 0/0 cyclist
New York, inner-city 11 ­ 15 years: 55 0/0 MVC
occupant; 14% pedestrian ;
12 010 motorcycle; 11%
bi cyc le; 4% fall/jump

Ald rich et al USA Retrospective January 1984 ""1 6 years Severe traumati c 53% died
1992" to September HI with diffu se brain
1987 swell ing

Berney et al Switzerland ;;' 1 hour Lo C 15.2 6.8

19942 • Cli nica I/i magi ng evidence
of brain contusion
Exclud es: concussion,
mild HI

Arnarson, Iceland 1987 -1991 0-14 years Hospita liza ti on 170

Hal dorsson fo r HI, all severities
199525
Jennett Scotlan d, UK Retrospective 1985 A&E attenders Attend ers 4011, 5.3 UK age-specific death
199626 admissions 400, and adm ission rates
evidence of brain peak at 15-30 yea rs
damage 290 A&E attenders peak
at < 10 years

Gabella et al Colorado, USA Surveillance 1991-1992 All ag es TBI defined from ICD Girls Age -adjusted Causes by age not given MVC mortality :
1997 27 study, residents 9 codes: 800 - 80 1, 0-4 years: 70; mortality rate s urban 14.8, rural
only 803-804 (skull 5-14 yea rs: provided 25.5
fracture), 850-854 60 all areas
(intracranial injury) Boys
Data from hosp ital 0-4 years:
discharge records 20 remote ru ra I,
and death ce rtificates 100 all other
5-14 yea rs:
90 metropolitan,
160 rural areas
(Con tinued)
 Table 14.1 (Continued)

Reference Country/region Study Study period Age group Inclusion criteria Mean incidence Mortality Main causes Main cause of TBI
for TBI ofTBI rate ofTBI (%J mortality
L­ ____ . _
--- _._­ -- - - ­

Eman uelson, SW Sweden Retrospective 1987-1991 0-1 7 yea rs ~1 hour of LoC 12 2.6 MVC 60<¥0 MVC 75%
von Wendt clinica l, Falls 22%
. 1997 28 neu rophysiol og ica I, Sport 7%
imaging evidence of Missile 3%
brain contusion Head trauma 20/0
(excludes: con cussion, Unknown 60/0
mild HI)
Suom in en Helsi nki, Finl and Retrospective 1985- 1994 < 16 yea rs Severe multiple trauma, All severe All severe trauma MVC (pedestrians Pedestrian s 21%,
et al 1998 29 inc. marked impaired trauma 14-185.6% 4.8 versus vehicle; MVC 20<¥0
consciousness, head injury 72% bicycle versus Cycl ists 12.7%
admission to ICU, to neurosurgical vehicle)
or auto psy ICU
Durkin et al New York - Retro spective 1983-1995 < 17 years Major MVC 9.3 < 17 years 2.9 < 17 years Falls 24%
1999 30 N Manhattan inc. HI = LoC ~ 1 hour, MVC 16%
with lacerati on :!: (2/3 pedestrians)
haemorrhage
Morrison Scot land, UK Retrospective 1981-1995 0-14 years Morta lity figures Injury mortality MVC mortality 4.8 Injury 140/0 of
et al 1999 31 Regi st rar Genera l, 11.6 Assau lt 0.7 total child
Data on Childhood Falls 0.6 mortality
Injury Mortality

Thurman, USA Annual surveys 1980-1 995 o to > 65 Hospitalization rates 1994-1995 In-hospital In cidence of mild and Not given
Guerrero years stratified : for TBI, from National TBI-related mortality rate, moderate TBI decreased
199932 0-4 years, Hospital Discharge admissions: all ages, remained over study period, but
5-14 years, Survey, CDC, PN CHS. 0-4 yea rs, 105 constant at 5.3 seve re TBI in creased
15-24 years, ICD 9 codes (range 68-142); from 10 to 19
25-44 yea rs, 800.0- 801.9, 5-14 years, 75
45-64 years, 803.0-804.9, or (range 50-100)
;:065 years 850.0-854.1
Benoit R N Virgi nia Retrospec ti ve 1991-1 999 0- 14 years Falls from windows, 3.9<¥0 mortality. 660/0 of 2322 falls resulted
et al 200033 admitted Trauma Centre, All deaths < 4 years in head trauma; children
or died fell onto concrete ";; 4 years more head inju ry
but less extremity injury than
10-14 years
rhurman et al USA, National Retrospecti ve 1997 0-14 years ICD 9 codes: Hospitali zed lBI: Pre-hospital deaths Causes different in Not given
2000 34 Centre Injury 800-801, 803-804, 0-4 ye ars = 2, at all ages different age groups:
Prevention, 850-854, and 873 59 Fatal outcome: 0-4 years: fall s >
Control for fatalities 5-9 years = 7.50/0,0-4 years transport related>
50 and 5-9 years; assault> firearms
10-14 years = 9.20f0, 10-14 years 5-9 and70-years:
51 transport > falls >
assaults/firearm s

Reid et al Minnesota, USA Retrospective 1993 0-19 years Children living in 73.5 ~ 19 years 9.3 ~ 19 years < 7 year: fall s, Non-metropolitan
2001 35 Minnesota with TBI Incidence peaks were Metropolitan 6.5, battering/assau It MVA
re sulting in in groups < 1 year, non-metropolitan 7-4 and 70- 74 years:
hospitalization or 6-7 yea rs and 15.4 fall s; MVC
death 13-19 years 5-9 years: fall s,
ICD 9 codes: Incidence in MVC (pedestrian)
800.00-801.99, Metropolitan 72.4, 75-79 years: MVC, falls
803.00-804.99, non-metropol itan
850.00-854.19 areas 76.1 TBI decrea sed as
and cases identified median hou se hold
from death certificates income and percentage
of high school
graduates in census
block increa sed

Masson et al Aquitaine, France Prospective 1996 All ages All admissions via Overall, 17.3 Overall mortal i ty 5.2 MVC (occupants 53 0/0, Ove rall highest
36 (IS-year emergency service severe TBI Mortality: pedestrians ISO/a, cyclists mortality: firearm s
2001
bands) with a diagnosis of 0-14 years, 8.4 0-9 years, 1.3; 10-19 12%, motorcyclists (but very few
severe head inj ury severe TBI years, 2.1 200/0) most common cases), pedestrians,
Excluded deaths Low-level falls, motorcyclists, falls
at scene high-level fall s on one level,
At 0-14 years, cyclists, high - level
mortality from falls, MVC
MVC-related lBI occupants
3.9; fall s 3.1

Adekoya et al USA Surveillance 1989-1998, o to >75 Mortality from lBI Not given Rates reduced Racial differences: 0-4 years: Falls

2002 37 summary by year years (5-year compiled from death graduall y At ages 0-19 years, 0.3, MVC 3.6,
Mortality only age bands) certificates, all States, 1998 mortality: MVA highest firearms 0.3, other 2.5
using ICD 9 codes 0-4 yea rs, 6.1 ; among American 5-9 years:
5-9 years, 3.3; Indian/Alaska Natives Falls 0.1, MVC 2.8,
10-14years, 5.5; Firearms lBI deaths firearms 0.4, other 0.5
15-19 years, 24.6 highest in black 70-74 years:
Americans Falls 0.1, MVC 3.4,
Fall-related TBI deaths firearms 2.0, other 0.6
highest in black 75- 79 years:
children at 0- 4 years; Falls 0.3, MVC
no difference s at 5-14 15.3, firearms
years; highest in white 12.5, other 1.6
chi ldren at 15-19
years old
(Continued)
Table 14.1 (Continued)

Reference Country/region Study Study period Age group Inclusion criteria Mean incidence Mortal ity Main causes Main cause of TBI
for TBI ofTBI rate of TBI (0/0) mortality

Williamso n et al Scotland. UK HI mortality 1986- 1995 0-14 years National mortality Not give n 290 child deaths MVC (ped estrian) MVC (pedestrian)
2002 38 review inc. data, Registrar from HI over 10 other bo th sexes 1.2
Gene ral for Scotland years M VC (occupants declined over study
ICD 9 codes for Mean annu al HI and cyclists) period
external cause of morta lity 3.0 ; fell Other MVC
inju ry and from 4.1 in 1986 (occupants,
skull fracture, to 1.8 in 1995 cyclists) 0.9
intracrania l inju ry Highest rate 5. 1 Falls 0.3
in most deprived areas; 0.3 Assaults 0.3
lowest 2. 1 in most Other, 0.3
affluen t areas assau lts in infants

Masson et al Aqui taine, Prospective 1996 All ages, Hospita I-admitted, Overa l1 8.5, All ages, 51.6 0/0 died All ages MVC 5.2
2003 39 SW France Traum at ic 0-14 years mas t severe bra i n 0- 14 years 3.2 0-15 yea rs, 42.9<\'0 0- 14 years: MVC
coma inj uries (GCS 8 or died 76 0Al, gunshot wounds
le ss), comatose for 10 0Al, falls 40Al,
> 24 hours, or died other 10%
without waking

Andersson et al W Swede n Prospective 1992- 1993 All ages, All grades or fatal Peak incidence All ages, 4.0 All ages: Morta lity for seve re
2003 40 and lO-year TBI, ide ntified from 0- 9 years Fal l at same level TBI ha s redu ced from
retrospecti ve age band s AEtE discharge register 0-9 years 35% 310/0, fall from height 40% in early 1980s
and death registers 10-19 years 20% 27%, MVC 160/0, hi t to - 10% in 1999
ICD 9 cod es 850-854 Overall, 2.50/0 by object 15<\'0
and 800-804 mode rate or severe
TBI, only 1% admitted
to hospital

CDCP, NCHS, Centers for Di sease Control and Prevention, National Hospital Discharge Survey; GCS, Glasgow Coma Score; HI, head injury; lCD, International Classification of Diseases; ICU, intensive care unit;

in c., inclusive; ISS, injury severity score ; LaC. loss o f con sc iousness; M VC, motor vehicle collision; PTA, post-traum atic amnesia ; TS, trauma score.

All rate s 100000 chil dren /year.

 Epidemiology I 289

350 r

300 r-
V>
c
.<::>
V>
250
V>

E
"0 200
~
'"
0
150
~
.D
E
::l 100

J J J J I~ I
Z

50
I I I I I I I
0
2 3 4 5 6 7 8 9 10 11 12 13
Age (years)

10 1998-99 • 2002-03 1
Figure 14.1 Admissions to the Accident and Emergency Department, Royal Hosp ita l for Sick Ch ild ren, Edinburgh, with head injury
1998-1999 and 2002-2003, by age. (From ref. 42, wit h permission.)

Incidence of suspected NAHI in Scotland hour. 46 Because birth trauma was often associated with a sec­
per 100000 chi ldren, aged 0 -12 months
ondary asphyxia, its precise incidence was difficult to obtain .
35.0
Routine magnetic resonance scanning of newborn
30.0
infants has demonstrated that clinically silent subdural
0

0 25.0 /' ~
haemorrhage does occur in infants in 6 per cent of normal
....--­ ~
0
S2 20.0 vaginal deJiveries. However, these haemorrhages resolve
"
C>­

"cu 15.0 within 4 weeks and remain clinically silent, and the infants
u" 10.0 develop normally.47 Instrumental delivery was associated
"g
5.0 with a higher incidence of subdural haemorrhage. 47 It
0.0
remains to be established whether these haemorrhages may
1998 1999 2000 2001 2002 explain minor neurological deficits in children. In a fUlther
95% (119.38.41 (17.5,481) (16.9,48.1) (106.37.8) (121.411)
ConL int.

sma ll study, Holden et al 48 detected small subdural haem­

Number 57945 56629 54028 52027
orrhages in the falx or tentorium in four out of eight oth­
Figure 14.2 Incidence of non-accidental injury in Scotland/
100000 children aged 0-12 months 43
Edinburgh. 44 The annual incidence of suspected NAHl in Scot­
land from 1998 to 2002 has remained constant in a declining
population of 57 945 in 1998 to 51046 in 2002, and is shown
in the accompanying figure (Fig. 14.2) .43 The incidence of
subdural haematoma in infancy in South West England and
South Wales was found to be 21 per 100000 children (95 per
cent confidence intervals, 7.5-34.4). of which it was estimated
that NAHI accounted for 82 per cent. 45
Traumatic Birth Injury
The incidence of traumatic birth injury has declined signifi­
cantly over the last 30 years, from approximately 6 per 1000
live biIths to approximately 0.2 per 1000 live births. This
decrease commenced in the 1970s as a result of changes in
obstetrics and neonatal care, which adva nced new manage­
ment principles such as not allowing labour to proceed
beyond 24 hours or the second stage to proceed beyond 1
51 046
erwise normal term infants who were vaginally delivered .
Mild Accidental Head Injury
----------------
The majority of head injuries in children are mild and these
include injuries in children who have had a brief interrup­
tion of consciousness, with or without vomiting. Clear
guidelines have been established to recommend which
patients should be admitted, and in all mild head injuries the
most important aspect is to recognize those who deteriorate.
The most frequent causes are accidents in the home - from
falling over, or sholt/high falls. Falls from bunk beds are
relatively common, as well as motor vehicle (Fig. 14.3) or
pedestrian accidents, non-accidental injury (NA!), SpOlt and
playground accidents, and bicycles or equestlian accidents.
Local policy reco mmends imagi ng by computerized
tomography (CT) scan for patient who have experienced loss
of consciousness of 5 minutes or more, together with persist­
ent vomiting or changed behaviour. The commo nest finding
is a linear parietal skull fracture (22 per cent) and small
intracranial lesions such as contusions; small extradural or
small intraventlicular haemorrhages are seen in fewer than
3 per cent of patients and are usually associated with an
290 I Head and neck injuries
100
90
80
-R 70
.s
c:
:::>
60
:~ 50
V\
:::>
0 40
·c
Q.)
Vl 30
20
10
0
3 6 9 20 25 30 35 40 45
(a) Speed at impact (mph)
Figure 14.3 (a and b) Relationship between veh icle speed and li ke lihood of serious injury or survival in Scotland.
overlying fracture and require no surgical action. Plain skull
films remain the most common investigative method in most
cases of mild head injuries.
Hospital policies for admission differ, but most will admit
children who have suffered a significant loss of conscious­
ness (more than 5 minutes) or amnesia, the presence of
abnormal neurological symptoms or signs, or skull fractures.
Indications Jor CT scanning include loss of consciousness or
arrmesia for more than 10 minutes, a deteriorating conscious
level, an open skull fracture, seizures and confusion, a signifi­
cant fall (from higher than 1 metre) or suspected non-accidental
aetiology. Continued altered consciousness or deterioration of
conscious state, depressed skull fractures, or presence of blood
or cerebrospinal fluid (CSF) leaks warrant neurosurgical referral.
Mild head injury may cause ischaemic stroke in a sma ll
number of children and most will make a good recovery.
The vessels involved are the functional end arteries in the
striatocapsuJar region (lenticulostriate branches of the mid­
dle cerebral artery [MCAl).49 The postulated mechanism is a
mechanical disruption of flow in the perfo rating branches
of the MCA with intimal trauma, and subsequent thrombo­
sis or arterial spasm. Genetic and other environmental fac­
tors (chickenpox) may be implicated. 50
Delayed cerebral oedema and fatal coma has been
reported after mild or trivial head injUly occurring after a
lucid interval. Attacks of familial hemiplegic migraine
(FHM) may be triggered by minor head injury and can be
accompanied by coma. Mutations in the CACNAIA cal­
cium channel subunit gene on chromosome 19 are associ­
ated with episodic symptoms, including co ma. It was,
therefore, postulated that the novel S218L mutation in the
CACNA lA calcium channel subunit gene is involved in
FHM and fatal cerebral oedema after mild head injury.51
Patients with mild head injury and persistent post-con­
cussive symptoms have been found to have a high inci­
dence of temporal lobe injury on single photon emission
computed tomography (SPECT) and positron emission
tomography (PET) scanning, which may explain the fre­
quent memory disorders. 52
100
90
80
~ 70
co
.2:> 60
~ 50
~
0
<U
v
40
I::
co
.<:; 30
u
20
10
0
50 20 30 40
(b) Vehicle speed (mph)
Physiological Monitoring in the Neurointensive
Care Unit
Modern intensive care management of head injury aims to
avoid secondalY physiological insults by optimizing cere­
bral oxygenation and perfusion while the brain recovers.
To ena ble early recognition of these damaging insults and
to institute prompt intervention, intensive physiological
monitoring is required.
Sta ndard physiological monitoring of the critically ill
head-injured child includes continuous electrocardiography,
oxygen saturation monitoring and measurem ent of arteria l
blood pressure via an intra-arterial catheter, as well as the
core and peripheral temperatures. Central venous pressure
and urine output are monitored continuously to guide ther­
apy for maintaining euvolaemia in these patients. End-tidal
CO 2 levels are measured using a capnometer to provide a
trend of the CO 2 removal, but frequent arterial blood gas
analyses are also carried out to allow monitoring of arterial
oxygenation and maintenance of normocapnia, which is
vital for control of intracranial pressure (fCP) .
Patients with documented hypoxaemia at any stage of
head injury management have a statistica lly poorer out­
comeY-57 The use of pulse oximetry (Sa0 2) has ena bl ed pre­
vention and early treatment of hypoxaemia. It has become
a widely used non-invasive method of monitoring oxy­
genation in modern head injUly management. Oxygenation
is most accurately determined by arteria l blood gas stud ies,
which require arterial puncture or an indwelling alierialline
and therefore can be performed only in a hospital setti ng.
Over the past few years, the use of a fibreoptic sensor
that is capable of measuring the oxygen tension, CO 2 ten­
sion, pH and temperature in the blood has gained consider­
able interest and has been adapted for use in the brain
tissue. 58 This multipara meter sensor can be inserted into
the brain tissue, together with the standard ventricu­
lostomy catheter and a microdialysis probe, through a
triple-lumen intrac ran ial bolt.

Many earlier studies of severe head injury have concen­
trated on the importance of raised ICP, current evidence
moves toward emphasising the importance of cerebral perfu­
sion pressure (CPP), the principal determinant of cerebral
blood flo w (CBF). Thus, maintenance of an adequate CPP is
essential to maintain cerebral oxygen delivery and prevent
cerebral ischaemia. The only way to reliably determine CPP in
the brain-injured patient is to continuously measure alierial
blood pressure and ICP (cerebral perfusion pressure = mean
arterial blood pressure - intracranial pressure).
Although ICP monitoring is not universally used , virtu­
ally all major adult head injury centres in the USA use ICP
monitoring in guiding management and it is an integral
pali of intensive care in these centres. The practice varies
in Britain, with rou t in e ICP monitoring occurring in 48-57
per cent of intensive care units (ICUs) that regularly treat
head-injured ad ults. 59 ,GO Similarly, evidence suggests the
monitoring and managemen t of children with head injury
who are at risk of developing intracranial hype rtension
varies considerably in the UK and worldwide. Segal and
colleagues G1 conducted a postal survey of the use of ICP
monitoring in children with acute neurological illness. All
British !CUs that admit five or more children each year with
acute neurological illness were contacted. Only one-half
of the 70 per cent who responded reported the use of ICP
monitoring. In t his survey, children admitted to paediatric
intensive care units (PICUs) were more likely to have ICP
monitoring performed than those admitted to adult ICUs. 61
There are no published guidelines for the use of ICP
monitoring in children, but the major determinan ts in t he
decision to initiate lCP monitoring appear to be the admis- .
sion GCS score and brain CT scan appearance. GI When ICP
monitoring is clinically indicated in a patient, a decision
on the type of monitoring device to be used is required.
Intracranial pressure monitors that are avaiJable at present
consist of a pressure transduction device that can be placed
in the following intracranial compartments: ep idural,
subdural, subarachnoid, parenchymal or ventricular space.
Pressure transduction can be undeliaken t hrough external
strain ga uge, catheter tip strain ga uge or catheter tip fibre­
optic technology.
The Camino fibreoptic pressure monitoring device is the
most fre quently used system in paediatric neurointensive
care. Initial reporting and validation of this type of trans­
ducing system have indicated its ab ility to measure the
brain tissue pressure directly with a rapid response rate to
ICP changes. Gambardella's group62 from Italy examined
its use in a group of children with severe craniocerebral
tra uma, aged 2-16 years, and demonstrated that this
method of recording ICP was minimally invasive and that
it correlated very closely with the ventriculostomy method,
Complications associated with ICP monitOIing include
bacterial colonization, infection, haemorrhage, malfunction,
obstruction or malposition of the device. Long-telID morbid­
ity is rarely associated with these complications but they may
affect the accuracy of ICP readings and influence the acute
Epidemiology I 291
management of these critically ill patients. HaemolTh age
associated with an ICP device is uncommon, with an overall
reported incidence of 1.4 per cent in the adult population 63 ­ 7o
Fewer patients (0.5 per cent) develop significant haematomas
requiring surgical evacuation. 65,67,69 Pople and co-workers 71
examined 303 children who required ICP monitoring and
found that only one patient (0.3 per cent) developed intra­
cranial haemolThage after insertion of the ICP moni tor; this
patient was known to have a low platelet count prior to the
procedure.
Malfunction or obstruction has been repOIied to occur in
all types of ICP monitors. The mechanical failure rate of the
fibreoptic device used in children with severe brain trauma
was reported to be 13 per cent in one series,72 but in another
study t he malfunction rate was much lower at 2,6 per cent. 71
Cerebral pe~fusiol1 pressure is the difference between the
mean arterial pressure (MAP) and intracra nial pressure
(ICP), i.e. CPP = MAP - ICP. The low systolic blood pres­
sure in infants means that the perfusion pressure is more
easily deranged than in adults. The normal cerebral perfu­
sion pressure is 60-70 mmH g in adults and, when reduced
there is a progressive fall in the CBF to the brain. When the
CBF falls below 40 mmHg there is absolute reduction in
cerebral perfusion and ischaemia a nd infarction will result
at flows below 18- 20 mL!100g of brain per minute.
Cerebrovascular autoregulatory mechanisms ensure that
there is an adequate CBF to meet metabolic demands by
changing the cerebrovascular resistance in response to
blood pressure changes over the range of 60-160 mmHg.
The arterioles constrict when the blood pressure rises and
dilate when it falls. 73
Cerebral blood flow also alters in response to the meta­
bolic demands focally or generally within the brain. Both
of these autoregulatory mechanisms are regulated by the
small arteries and arterioles. Major blood vessel dimensions
are reduced by vasospasm, which can be sufficient to cause
infarction. Small vessels have sympathetic innervation,
possibly from the locus coeruleus, and respond to meta­
bolic demands of a paliicular part of the brain (including
during sleep). These local flow changes are all influenced
by pH , nitric oxide and lactate production.
The Cushing phenomenon is a progressive response of
systemic arterial pressure in response to an increasing fCP
as it increases to levels approaching the diastolic blood
pressure. It is mediated via brainstem reflexes, which
depend on blood flow to the locus coeruleus. The cardiac
output and peripheral resistance are increased, the latter by
sympathetically mediated peripheral vasoconstriction and
an increase in vasomotor tone. Thus there is a sequence of
comp ensatory events to raised ICP from the initial cerebral
arteriolar dilatation followed by elevation of the MAP, fol­
lowing which both MAP and CPP fall in respons e to con­
tinued ICP elevation. 74
Continuous body temperature monitoring in patients with
head injury is a routine practice, as abnormalities in body
temperature, paliicufarly pyrex ia, occur frequen t ly after
292 I Head and neck injuries
brain injury.75,76 Conventional indicators of core temperature
include rectal and bladder temperature, but recent advances
in intracerebral monitoring have enabled brain temperature
to be measured; this is done by using a thermocouple embed­
ded in the tip of an intraventricular catheter used for rcp
monitOling (or a triple bolt inserted into the frontal region
encompassing an ICP monitor) , a microdialysis catheter and
a Para trend multiparameter (oxygen, carbon dioxide, pH and
temperature) sensor. 75 ,77,78
The precise influence of pyrexia on the outcome after
brain trauma remains unclear. Pyrexia has been postulated
to exacerbate brain injury by increasing cerebral metabolic
requirements for oxygen and excitatory neurotransmitter
circulation.
The clinical use of hypothermia was pioneered in
1938 79,80 and since then several laboratory studies 81 - 84 and
small clinical trials 85.86 have suggested its association with
an improved recovery. In 1994, a multi centre randomized
controlled study was conducted, in the hope of providing
definitive evidence of the efficacy of hypothermia in adults
with severe brain trauma. The study was halted in 1998 after
enrolment of 392 of the planned 500 patients when induced
hypothermia, bladder temperature of 33°C, initiated within
6 hours of injury and maintained for 48 hours, failed to
improve clinical outcome.87 Furthermore, patients who were
in the hypothermic group and more than 45 years old had a
higher incidence of poor outcome and had more complica­
tions than those who were kept nOlIDothermic. 87 Neutropenia
and coagulopathy may complicate induced hypothermia.
Thus, fuliher studies are required to determine the influence
of brain temperature on outcome after head injury. It is not
resolved whether hypothermia induced by selective head
cooling in newborns merely reduces the extracranial blood
flow without alteration of the intracranial circulation.
ELECTROPHYSIOLOGICAL MONITORING
Head-injured children requiring intensive care often need
to be pharmacologically sedated and paralysed; this masks
the motor and other clinical manifestations of seizures and
makes the assessment of cerebral function by conventional
clinical neurological examination impossible, apart from
pupillaly reactions. Furthermore, some children develop
abnormal movements without electrical seizure activity
after head injUlY and without electrophysiological moni­
toring they can receive unnecessary treatment. In addition
to providing valuable information on the degree of cerebral
insult in the critically ill child with an al tered level of
consciousness, electrophysiological monitoring of brain
activity has been recognized to have prognostic value in
children with severe head injury.aa-90
Multichannel electroencephalography (EEG) recording is
the gold standard in measuring surface brain activity.91-94
It allows distinction to be made between generalized, local­
ized, focal or multifocal seizures and to provide an aetio­
logical and prognostic guide in relation to single, repeated
or multifactorial insults. Tasker and co-workers 95 subse­
quently used the cerebral function analysing monitor
(CFAM) in conjunction with serial multichannel EEG to
monitor cerebral function in 54 critically ill comatose chil­
dren , and demonstrated that CFAM traces could be used to:
(1) identify effects of acute or cumulative cerebral insults;
(2) recognize unstable patients exquisitely sensitive to
standard care procedures; and (3) evaluate seizure control.
Although maintaining an adequate CPP is important as
it provides a pressure gradient governing the CBF, the ulti­
mate determinants of the cerebral metabolism are CBF and
the oxygen content in the blood. Monitoring of the ICP and
CPP alone offers no information on the brain's oxygen
delivery and usage. At present, methods to measure cere­
bral blood flow such as PET, Xenon clearance or SPECT
are too cumbersome for use in an ICU, however, jugular
venous bulb oximetry and transcranial Doppler ultrasound
enable a better understanding of the state of the cerebral
circulation and oxygen consumption .
CEREBRAL BLOOD FLOW
The CBF depends upon: (I) cerebral perfusion pressure;
(2) cerebral vascular resistance (radius of vessel, length of
vessel) inversely; (3) viscosity ; (4) venous pressure; and
(5) ICP. (CPP = CBF/CVR, CBF = CPP X CVR.)
The normal CBF is around 50 mL per 100 g of brain tissue
per minute. 96 The CBF is a higher percentage of the cardiac
output in small infants than in adults. It is also highest in
grey matter (100 mL per 100 g) and decreases, sequentially,
through the inferior colliculus, sensory cerebral cortex,
motor cerebral cortex, geniculate bodies, superior colliculus,
caudate nucleus, thalamus, cerebellum to the cerebral white
matter, where the rate of flow is 20 mL per 100 g of brain tis­
sue per minute. Impairment of CBF due to raised ICP is clin­
ically seen as distension of scalp and retinal veins, pulsation
of the fontanelle and a loud systolic cranial bruit, which
appears when the ICP exceeds the diastolic blood pressure
and increases in pitch with slowing of the cerebral circula­
tion, and disappears with a 'carotid stop'.
Non-invasive methods of assessing the flow velocities of
the intracranial cerebral arteries using transcranial Doppler
ultrasound (TCD) were first described in the early 1980s. 97
Flow velocities can be measured in the middle, anterior and
posterior cerebral arteries by placing a probe in the temporal
area, just above the zygomatic arch. As the cross-sectional
area of the arteries cannot be measured directly, flow can­
not be measured from velocity but the Doppler shift meas­
ured is inversely proportional to the diameter of the vessel.
Thus, providing that all of the other factors remain con­
stant, the mean flow volume (MFV) will approximate CBF.
Vasospasm will result in an increase in flow velocity.
The resistive index (RI) or Pourcelot index (S - DIS) is an
alternative ratio for assessing flow velocity and is normally
of the order of 0.7. It increases with increased ICP and is
usually reliable unless there is distortion of the course of the

MCA. Using transcranial Doppler, MCA spasm was identi­
fied in 40 per cent of patients after traumatic brain injury
and could begin as early as 48 hours post injury.98 Maximal
MCA spasm was found between 5 and 7 days after the
initial trauma. 98 In another study of head-injured adults,
severe MCA spasm identified by transcranial Doppler was
confirmed by angiography.99
Daily measurements of the MCA velocity from 121
patients with varying severity of head injury showed an
inverse correlation between the severity of head injury and
the MCA velocity. 100 In addition, a significantly lower MCA
velocity on admission was demonstrated among the head­
injured patients than among the normal control subjects. 100
High ICP and low CBF were responsible for the low velocities
in the intracranial circulation after head injury. FUlihermore,
an admission MCA velocity of less than 28 cm per second
correctly predicted 80 per cent of the early deaths. 100
MICRODIALYSIS
As the composition of the cerebral interstitial fluid reflects
the biochemistry of the neurones and glial cells in the
brain, attempts have been made over the past decade to
develop the use of cerebral microdialysis to monitor neuro­
transmitter release and energy metabolism in the brain. By
mimicking the function of a blood capilJary, the micro dial­
ysis catheter enables monitoring of the chemical changes
in the interstitial fluid. 101 Because of a concentration gradi­
ent across the semi-permeable microdialysis membrane,
chemical substances within the interstitial fluid diffuse
across the dialysis membrane and into the perfusion fluid
inside the catheter. 99 Samples are collected in microvials
and brought to a bedside analyser as often as necessary. 101
The microdialysis catheter can be implanted during sur­
gery or when the patient is in the ICU. If implanted during sur­
gery, the catheter is tunnelated under the scalp and through a
small incision through the dura. Ihe membrane of the catheter
is then positioned in the penumbra region, usually 1 cm from
the border of the evacuated lesion. A second catheter may be
placed in the surrounding 'normal' brain tissue through a sep­
arately drilled hole in front of the intraventricular pressure
catheter. Ihis evaluates any further degradation of normal tis­
sue physiology due to secondary damage. In those patients
who do not require surgery the microdialysis catheter is
inserted through an intracranial bolt, but this makes it more
difficult to reach a position within the brain that is relevant
for the interpretation of the brain chemistry. Ihe position of
the catheter may be visualized on CT scans.
Ihe dialysate lactate, lactate-pyruvate ratio, glycerol,
glutamate, glucose and adenosine levels have been investi­
gated in head-injured patients. 102-108 Ihe lactate-pyruvate
ratio has been shown to be a better marker of ischaemia
than lactate alone, as an increase in lactate may be the
result of hypoxia, ischaemia or hypermetabolism. 102, 103 But,
during ischaemia, the neurones become dependent upon
the anaerobic production of AlP from glucose, resulting in
Epidemiology I 293
an increased production of lactate while the production of
pyruvate decreases; this causes an increase in the lactate­
pyruvate ratio. 102 ,103 Glycerol is an integral component
of the cell membrane and has been use, therefore, as a
marker of cell membrane damage. 104 ,105 Although gluta­
mate has been proposed as an indirect marker of cell dam­
age, it is often hard to interpret the dialysate level as the
neuronally released glutamate is mixed with the large
metabolic pool of glutamate. A reduction in the dialysate
glucose is often associated with a decrease in brain oxygen
tension suggestive of reduced cerebral perfusion. 106 A
marked increase in the adenosine level of the cerebral
interstitial fluid has been shown to occur during jugular
venous oxygen desaturations suggesting a potential role
for adenosine during the periods of secondary insults after
brain trauma. 109
IMAGING OF ACCIDENTAL HEAD INJURY
A non-contrast CT scan of the head and upper cervical spine
and plain cervical spine radiographs should be the first-line
imaging in children with accidental head injury, but it is
important to recognize that up to 60 per cent of spinal
injuries in children may occur without evident bony injury. 110
Computerized tomography scanning of head-injured patients
is recommended because it is sensitive to the presence of
haemorrhage and best demonstrates bony integrity. It is the
usual imaging modality on presentation and also for subse­
quent scans that are undertaken during the acute admission.
The usefulness of magnetic resonance imaging (MRI) for
accidental brain injury in children has yet to be defined but,
in general, MRI and CI are complementary modalities. Mag­
netic resonance imaging is better able to depict extra-axial,
suboccipital, subtemporal, subfrontal or tentorial haemor­
rhage and provides a better view of the posterior fossa and
brainstem. It is useful in locating shear injury and at follow­
up, the extent of parenchymal damage.
Ihe Marshall CI Score lll is used to classify the CI
injuries following head trauma (see Iab.le 14.2).
Specific CI scan findings after head injury include dif­
fuse axonal injury, contusions, focal ischaemia, hydro­
cephalus, diffuse brain swelling and brain oedema. Diffuse
brain swelling is more common after head injury in chil­
dren than in adults and may be due to hyperaemia, vascu­
lar congestion, an increased water content or brain
(cellular) oedema. It is associated with at least two CI scan
appearances: (1) a low-density pattern, suggestive of
ischaemia from raised ICP or additional hypoxia in coma­
tose children and (2) a normal or hyperdense pattern that is
more amenable to pressure-reducing measures.
EPIDURAL HAEMATOMAS
Epidural haematomas in adults are usually arterial in
origin but may be of venous origin in children. About one­
third of cases in children are associated with coma, due to
294 I Head and neck injuries

Table 14.2 The Marshall CTscore

Diffuse injury I (no visible pathology) No visible pathology seen on CT scan

Diffuse injury II
Diffuse injury III (swelling)
Diffuse injury IV (shift)
Evacuate mass lesion
Non-evacuated mass lesion
Cisterns are present with midline shift of 0-5 mm and/or:
lesi on densities present, no high- or mixed-density lesion > 25 cc,
may include bone fragments and foreign bodies
Cistern compressed or absent with midline shift of 0-5 mm, no high- or
mixed-density lesion >25 cc
Midline sh ift >5 mm, no high- or mixed-density lesion >25 cc
Any lesion su rgically evacuated
High- or mixed-de nsity lesion >25 cc, not surgically evacuated

concomitant parenchymal injury. This improves after surgical
remova l of the clot without opening the dura. La rge volumes
of blood may be lost, leading to shock and secondary brain
injury or death. The risk factors for deterioration include a
fracture crossing the middle meningeal artelY, vein or sinus,
or an increase in the size of the epid ural haemorrhage.
SUBDURAL HAEMATOMAS
Thin subdura l haematomas assoc iated wit h diffuse brain
injury or swelling are most common. The indications for
decompression include midline shift, persistent raised [CP
and no associa ted severe brain swelling. Subdural haemor­
rhages in the interhemispheric fissure may be tapped in
infants via the anterior fontan elle.
CONTUSIONS
Focal contusions may not be immediately visible on CT
scans but are evident in 50 per cent at follow-up.1I2 They
exert a mass effect and contribute to the overa II raised [CP.
in the intensive thera py unit (ITU), paediatric neurology
department and by the child protection team. The inappro­
priateness and inconsistency of these multiple accounts
may be an important indicator of an inflicted injury. A
traumatic explanation is offered in 10-20 per cent of cases
to account for the child's condition.
The symptoms at presentation to hospital with thei r rel­
ative frequencies were:
• irritability and crying, 41 per cent;
• bruising and su perficia l injury, 41 per cent;
• vomiting and a nore x ia , 38 per cent ;
• apnoea and respiratolY difficulties, 38 per cent ;
• extensor stiffness, due to fit, anoxic rigidity or raised
intracranial pressure, 34 per cent;
• probable fits, 21 per cent;
• pallor or cyanosis, 29 per cent;
• history of a potential traumatic event, 26 per cent;
• depressed conscious state, 21 per cent.
ON EXAMINATION

SKULL FRACTURES Scalp Injury

Depressed sku II fractures require repair of the dura to prevent
raised pressure, which leads to brain herniation with venous
infarction of the cortex. They also predispose to infection.
Fractures of the skull base, with leakage of CSF throu gh
the nasal cribriform plate or t he ear, result in a compens a­
tory mechanism that keeps the ICP low until the leakage
ceases. The defi nitive rep a ir is undertaken after the acute
injUly period.
Penetrating injuries (gun shot, knife, darts, needles) to
the brain in children are rare in the UK. In those cases when
there is a preservation of consciousness, the object is left in
position until after sca nning. Debridement is then under­
taken with pressure monitoring and antibiotic cover.
NON-ACCIDENTAL HEAD INJURY
A history of preceding events lea ding to hospital admission
is obtained on several occas ions in the AEtE departm ent,
Bruising and oedema of the sca lp , periorbital bruising or
Battle's sig n (bruising over the mastoid) (Fig. 14.4), usually
seen in accidenta l head injury, is rarely present, even in the
presence of severe intracranial damage.
Skeletal Injury
Skull fractures are seen in approximately one-quarter of cases
and evidence of impact is seen in more than one-half of cases.
Signs of impact, particularly of the head, indicate
inflicted trauma. As young infants are unable to injure
themselves accidentally, they may be an indicator of intent.
Clinical evidence of imp act is:
• cutaneous or subcutaneous bruising;
• subgaleal haemorrhage;
• skull fracture;
• extradural haemorrh age;
• focal subdural haemorrhage;
• focal cerebral contusion.

Figure 14.4 Battle's sign. Prominent bruising behind th e pinna
indicates a skull base fracture.
Overa ll, 53 per cent of cases ha ve no fracture on skeleta l
survey. Rib fractures are seen in one-third of cases and
fractures of the long bones seen in one-qu arter.
BRUISING AND SKIN TRAUMA
Our experience is that some bruising and superficial abra­
sions occur in more than 40 per cent of cases, often a
mixture of new and old bruises. Additionally, abras ions,
sc ratches, lacerat ions, hair loss, subconjunctival haemor­
rhages and bl eedin g from the ears and mouth are present in
a small number of cases.
A firm adult grasp is necessary in order to shake a child
and may result in a pattern of bruises. These must be care­
fully documented and photographed. Common methods of
grasping a chiJd (and subsequent injuries) are:
• by the chest - thumb marks at the sides of the nippl es,
fractured ribs and retinopathy;
• by the arms - spiral fracture of the humerus;
• by the legs - bucket handle metaphyseal fracture ;
• by the throat - carotid trauma (ipsilateral infarction
and loss of ve nous pulsation in t he retina, contralateral
hemiplegia, faci al and retina l petech iae);
• by the abdomen - finger bruising, retroperitoneal
bleeds and bowel bruising;
• by the shoulders - finger and th umb bruises;
• bruising around t he mouth suggests possible
suffocation to prevent the infant screami ng; thumb
marks should be carefully sought under the mandib le.
Non-accidental head injury I 295
NEUROLOGICAL PRESENTATIONS OF NON-ACCIDENTAL
HEAD INJURY
Fifty years ago, the whiplash element of the shake n baby
sy ndrome was highlighted by Caffey ll3 and Guthkelch 1l4
but the more obvious supratentorial rotational injuries, i.e.
su bdura l haematoma, are easier to see and since then have
dominated clinical descripti ons. We propose a classi fica­
tion based on the clinical pattern of presentation, which
should be considered as the tempora l classification of neu­
rological presentations:
I. hyperacute cervicomedullary syndrome (whiplash
shaking injury), 6 per cent;
II. acute encephalopathy of classical 'shaken baby
syndrome' with rotat ion plus or minus impact (fits,
coma, decerebration and central apnoea ­
ventil ation). 53 per cent ;
III. sub acute non - encep halopathic presentation (subdura l
haemorrhage, haemorrhagic retinopathy, fractures,
bruising). 19 per cent;
IV. reCUlTent encepha lopathy - none was recorded in t his
series;
V. chronic extracerebral presentation (isolated subdural
haemorrhage), 22 per cent. 115
Hyperacute Cervicomedullary Syndrome
This presentation results from severe shaking forces. Geddes
et al" 6 , 11 7 found localized axonal damage at the craniocervi­
cal junction, in the corticospinal tracts in the low er brain­
stem and the cervica l cord roots; additionally, t here is the
possibility of traum atic thrombosis of the vertebral arteries
as they wind through the foram ina of the cervica l vertebrae.
Damage to these areas is thought to result from hyperflex ion
and hyperextension movements, a 'whiplash' shaking stem
injury. These cases are infrequently or only briefl y seen by
the clinician (6 per cent) as they a re either dead on a rri val at
hospital or die shortly thereafter, but are more frequently
seen by pathologists. Th ese young infants, at about 2- 3
months of age, presen t with apnoea as part of this cervi­
comedullary syndrome, with only 'trivial' subdural bleeding.
The apnoea gives rise to severe seconda ry hypoxic brain
injury with cerebral oedema but minimal axonal shearing.
In this hyperacute presentation, all subdural haemorrhages
that are seen on imaging and post-mortem exa mination are
sma ll , as a result of ac ute cerebral oedema. Significant sub­
dural haemorrhages do not become evident until 2- 3 days
later. In summary, the presentation of the hyperacute pattern
is apnoea, acute respiratOlY fa ilure and death associated wi th
severe hypoxic ischaemic dam age found post mortem.
Acute Encephalopathic Presentation
This is characterized by seizures, decerebration, homeostatic
derangements, bilateral large subdural haematomas and
widesp read haemorrhagic retinopathy. Additional findings,
296 I Head and neck injuries

such as rib fractures, metaphyseal 'corner' or 'chip' fractures,

bruising, cuts and cigarette burns, may be present. This is the
commonest pattern seen by hospital paediatricians and has
been referred to as classical 'shaken baby (or shaking impact)
syndrome'. Apnoea and respiratory arrest, grunting respira­
tions, shallow respirations or choking represent direct
medulialY or vertebral artery trauma or the effects of raised
ICP. Following admission to hospital, two-thirds of children
presenting with 'shaken baby' syndrome develop epileptic
seizures, which are often severe and drug resistant, reaching
a climax at 24-48 hours post injury, but usually decreasing
and ceasing by the fifth day; it is likely that the fits are an
epiphenomenon that refiects the extensive brain injury
Also, more than two-thirds of these children have docu­
mented raised ICP. The increased ICP, together with shock
and hypotension, further reduce the CPP and increase the
risk of secondary ischaemic brain damage. Low cerebral
perfusion pressure correlates with long-term handicap. li S

Subdural Haemorrhage
Bleeding from torn bridging veins in to the subdural space
is the hallmark of non-accidental shaking injUly in the first
year of life. Almost 90 per cent of patients suffer subdural
haemorrhage, which is frequently bilateral, and in 20 per
cent of cases it is associated with subarachnoid haemor­
rhage. The subdural haemorrhage may be unilateral at
presentation. Subdural haemorrhage may be over the con­
vexity, interhemispheric, subtemporal, suboccipital or in
the posterior fossa. The subdural signal intensity on MRI
scanning may be different in different intracranial com­
partments and does not imply repeated bleeding from
repeated trauma. By the time of necropsy, only a thin film
of subdural haemorrhage may be present over cortical con­
vexities (Fig. 14.5).
Retinal Haemorrhages
In shaken baby syndrome, retinal haemorrhages occur in
80 per cent of cases and one-third are unilateral. I1 9 Levin's
groupl20 found them in 84 per cent of cases compared to
16 per cent in accidental head injuries - usually high­
velocity road traffic accidents with side impact. Retinal
haemorrhages were extensive and involved all of the reti­
nal layers. They are thought to be due to vitreous traction.
The vitreous humour is attached to the retina at its periph­
ery and with different inertias this is also the site where
shearing forces tear the vitreous attachments from the
retina (or ora serrata) and cause disruption of ocular and
orbital blood vessel integrity - similar to the shearing
forces which cause subdural bleeding and cortical tears.
Non-encephalopathic Subacute Presentation
With this presentation, there is no acute brain swelling or
diffuse cerebral hypodensities or fits, coma or decerebra­
tion. Subdural haemorrhage, retinal haemorrhage, rib and
other fractures and bruising occur in various combinations.
Figure 14.5 Infant, aged 3 weeks, at necropsy. A thin film
of subdural haemorrhage is present over cortical convexities,
following inflicted injury. Bilateral rib fractures and bruises were
present.
The child may present with a recurrent encephalopathic
presentation with 'odd turns' or fits, apnoeic attacks, cyan­
otic attacks, rigidity or coma.
Non-encephalopathic Isolated Chronic Subdural
This is a 'late' presentation in a child with an expanding
head circumference, subdural haemorrhage and often little
else. There may be vomiting and hypotonicity. Other causes
for subdural haemorrhage must be excluded.
DIAGNOSIS OF NON-ACCIDENTAL HEAD INJURY
The diagnosis of NARI is made on the history, which is fre­
quently inconsistent and does not explain the clinical find­
ings - ophthalmological and radiological features compatible
with, and sometimes characteristic of, non-accidental oligin;
and the social pathology in a significant number of cases
in the form of known risk factors (Table 14.3). Evidence that
shaking is responsible for many NARIs emanates from
numerous studies and reports of clinical experience, fre­
quently acknowledgement or confession with deSCription of
the mechanism by the perpetrator. Biomechanical and animal
studies have confirmed the forces generated by shaking and
other mechanisms. Injurious behaviour has been witnessed
by closed circuit TV. Consensus statements by scientific bod­
ies such as the American Academy of Pediatrics state the
extensive combined experience of professionals. 121 As the

Table 14.3 Risk factors for inflicted head injury
Single parent or cohabitating partner
Young parental age
Past history of abuse to children
Drug or alcohol abuse
Domestic abuse
Parental mental or psychological illness
Premature birth
Twin s
Re cu rrent medical consultations and hospital admissions
Past history of social work enquiry or involvement
Table 14.4 Features of the expanded syndrome of
non-accidental head injury
Acute encephalopathy
Subdural hematoma
Age usually less than 1 year
Acute cervicomedullary injury
Haemorrhagic retinopathy
Bruising
Sku II fractu re
Rib or limb fracture
Evidence of malicious injury (bites, cuts, cigarette burns,
whip marks)
A history that is incompatible or inconsistent with
clinical findings
Acute cerebral oedema or diffuse cerebral hypodensities
Early cerebral atrophy
Poor long -term prog nosis
trauma is almost always unwitnessed , it is important that the
clinician is careful about attributing a mechanism of injury in
a particular case and diagnoses 'inflicted ' or a 'non-accidental'
injury (which, however, may be consistent with a shaking
mechanism).
Although, theoretically, the features in Table 14.4 repre­
sent the most complete (expanded) syndro me of NAHI, in
practice common clinical presentations include a combina­
tion of several of these components.
The component palis do not carry equal weight nor are
they independent. The more components present, the more
secure the diagnosis. Although a single component, for
example subdural haematoma or retinopathy or isolated
haemorrhagic retinopathy in a yo ung infant, is still most
likely due to NAl, it may be due to other causes. Individual
signs and symptoms can have different causes and some
features are more characteristic of NAl and so the diagnosis
only becomes statistically secure when several components
are found in combination. The most secure diagnosis of
NAHI is either when there has been an admission from the
perpetrator or when the head it\iury is seen in combination
Non-accidental head injury I 297
with evidence of other malicious injUlY, such as multiple
fractures of different ages, skin incision , cigarette burns or
repeated beating.
Combinations of any two of the following three factors
are highly predictive of inflicted head injury (P < 0.001), 122
i.e. head injury plus any two of th e three following :
• inconsistent history/physical ex amination;
• retinal haemo rrhages;
• parental risk factors (alcohol or drug abuse, previous
soci al service intervention within the family or a past
histolY of child abuse or neglect).
Traumatic retinoschisis and retinal folds are thought to
be pathognomonic for 'shaken baby sy ndrome', although
extensive multilayer retinal haemorrhages are most
unlikely to be due to any other cause. It is important to dis­
tinguish between making a diagnosis of NAHI with raising
concern or suspicion of it. Kivlin 123 quotes 'the mere pres­
ence of any retinal haemorrhages adequate to raise concern
of "sh aken baby syndrome", the extent or type of the
haemorrhage is less clinically important.'
Subdural haemorrhage is more likely to be non-acci­
dental in origin if it is convex and bilateral, interhemi­
spheric, and is associated with cerebral tears and ruptured
bridging veins that are visible on imaging.
Rao and co-workers l24 concluded that hypoxic-ischaemic
encephalopathy (HIE) plus interhemispheric subdural haemor­
rhage was highly specific for the diagnosis of NAI. The inter­
hemispheric subdural haemolThage and HIE OCCUlTed in 89 per
cent of cases. Zimmennan's group 125 suggested that interhemi­
spheric subdural haemorrhage could be specific for NAI (61 per
cent of cases).
There is nothing characteristic about the traumatic
encephalopathy in NAHl, although post-traumatic seizures
occur velY much more frequently in inflicted head injury than
in accidental head injury and they abate by the fifth day.1 26
Duhaime et al 127 developed an algo ri thm for determining
the probability that an injury was inflicted. It results in a
diagnosis of 'presumptive inflicted ' injury or 'suspicious
inflicted'injury.
The presence of anyone of the features in Table 14.3 in
a child with injuries must raise concerns or suspic ion about
the possibility of child abuse and should be an indication
for investigation.
It is imp0!1ant to consider the questio n of intent, which can
sometimes be inferred from the clinical findings , for example
when the intent was malicious and the perpetrator cannot
have been in any doubt about the injury he was causing to the
child. The clinical findings supporting this intent include:
multiple bruises of different ages and patterns in different
sites, repeated admissions with physical injuries, multiple
fractures of different ages, and particular types of injuly pat­
tern, such as bite marks, cuts and cigarette bums.
Disciplinary injuries are not associated with the intent
to injure and there may be no intention of harming the
child at all. The intention to discipline is not premeditated
298 I Head and neck injuries
or sadistic. AI.though a physician may be sympathetic to
the stressful circumstances surrounding such injury, it
must be reali zed that without intervention , if the stress is
repeated then so may the abuse. Other possible explana­
tions for the injury include episodic dyscontrol, puerperal
psychosis, euthanasia, and Munchausen's syndrome by
proxy (fabricated or induced illness).
ACCIDENTAL SHORT FALLS
'Short falls' are sometimes offered as the explanation for a
child's injuries and most will report a 'short fall ' «3 feet)
from a bed, lounge, changing table or a parent's arm. Most
toddlers fall three to five times per week 128 but our own
direct observational study of more than 700 child-hours
found that 1- to 2-year-olds fell approximately once every
2 hours, hitting their heads in just over 10 per cent of falls
but without fracture or serious injury. 129 In infants who are
not yet weightbearing, any fall must be initiated by an
adult or equipment failure. Fatalities from short falls are
extremely rare but do occur: 0.14-0 .22 deaths per 100000
children between aged from 0 to 4 years. IJO
Documented fatalit ies and severe brain injuries have
been seen as a result of 'infant walker' incidents involving
stairs. 131.132 They a lso occur in an older group of children
as a result of falls from playground equipement,133 top
bunksl J4 a nd 'baby bouncers '. 131 For very short falls, a
short impact time is associated with low terminal velocity,
which is not able to generate sufficient rotational velocity
and thus subdural haematoma.
Imp act injuries from linear acceleration a nd deceleration
result in skull fracture and extradural ha ematoma without
concussion. Shearing, with retinal haemorrhages, subdural
haemorrhage, parenchymal inju ry and concussion, results
only if the impact is associated with rotational injury.
Skull fractures accompany 26 per cent of cases ofNAHI
a nd 22 per cent of accidental head injuries. They are clini­
cally important if depressed, if they extend through the
cribriform plate or the petro us temporal bone resulting in
CSF leaks and a risk of meningitis, or if they involve the
skull base with brainstem injury. They indicate an impact
force, which may be accidental or non-accidental in origin.
Skull fractures do not heal by callus formation, making
dating of an injury difficult; however, if the edges are round
and smooth then the injury is more than 2 weeks old. At
autopsy, bone edges are heaped , smooth and discoloured
by haemosiderin. A skull fracture normally heals in 2-3
months and disappears on radiograph by 6 months.1J5 In
small infants, the fracture site may not heal; rather it can
form a growing skull fracture as described below. There may
be no bruising at all over the skull, even with a severe
impact fracture; however, bruising of the aponeurosis may
be evident at surgery or autopsy. Equally, fractures may not
be seen on radiograph - what appears to be a simple disci­
plinary shaking injury may be negated at autopsy or sur­
gery if we ll-defined fractures are seen, suggesting more
severe impact injury. The type of skull fractures reported in
a study of 100 consecutive children 127 who were less than
2 years of age revealed that 27 were linear, eight were
depressed, three were multiple, one was stellate, three were
bilateral and four were basal.
BIOMECHANICS OF SKULL FRACTURE
Force is the product of mass multiplied by acceleration . The
force of approximately 35 foot-pounds is required to cause
an adult skull fracture. With constant mass, force can only
be increased by increasing the velocity at impact, i.e. fall
from a greater height or harder punch, etc. Falls from a bed
result in infrequent skull fractures in young children; those
that do occur are unilateral, narrow and do not cause seri­
ous injury to the child.1J6 A fracture is indicative of force
to the head but the force may have been linear or rota­
tional. Extracranial muscle, hair and skull thickness will
diminish the effect of the force on the brain.
The membranous skull of the young infant is more
elastic than adults and may deform without fr acture. 137 A
wider area of impact will dissipate the force and is less
injurious or likely to cause fracture. Soft surfaces result in
a prolonged impact time and a low resultant fo rce. More
elastic infant bones are likely to experience bounce than
adult ones and further dissipate force.
Skull fractures occur along the lines of leas t cranial rein­
forcement, i.e. the temporal and parietal regions. Bursting
fractures result from compression of the skull. Non-accidental
origin of a skull fracture is suggested by a fracture line that:
• bra nches;
• is stellate;
• crosses suture lines;
• is bilateral;
• is multiple;
• is wide at presentation and expands;
• is a growing skull fracture;
• is a depressed occipital fracture in a child under
3 years of age.
Growing skull fractures are seen in both accidental and
NAHl, but are thought to be more typical of a non-accidental
origin (Fig. 14.6). They occur during the rapid phase of
skull growth and most occur in infancy. They may result
from trapping of the dura between the fracture margins,
meningeal hernia with pulsation of the dura, or as the
result of pseudarthrosis. lJs
OTHER SKELETAL FRACTURES
The incidence of fractures in NAl varies widely, and is esti­
mated at between 11 per cen t and 55 per cent. 139 They are
commonest in children under 2 years of age. Fractures in
babies under 4 months old are nearly all due to abuse.
Radiological features suggestive of NAl and their differen­
tial diagnosis are discussed in Chapter 3.
 Non-accidental head injury I 299

Figure 14.6 Plain lateral radiograph of sku ll with growing

fracture caused by the interposition of tissue between bone edges,
seen after both accidental and inflicted injury.
Table 14.5 Investigations in suspected non -accidental head
injuries
Figure 14.7 Ei ght days post admission : T2 transve rse magnetic
resonance imaging sequence showing high intensity white matter
bilaterally at the parieto -occi pital and right temporal regions.
Other sequen ces confirm posterior and interhemispheric high­
signal subdural haemorrhages.

Brain imaging IMAGING OF NON-ACCIDENTAL HEAD INJURIES

MRI (FIGS 14.7-14.10)
CT
USS
Both CT and MR scanning have been used to image the
Dopp ler stud ies acute non-accidentally head-injured infant for which
Photogra phy subdural haematoma is the radiological hallmark. The
Electroen cepha log ra phy subdu ral haematomas are most likely to be posterior, i.e.
Spectrophotometry of cerebrospinal fluid
parieto - occipital or interhemispheri c in site. 141 In the early
Intracranial pre ssure measurement ac ute stage this may simply appear as a thin crescentic
Coagu lati on screen density on CT scan. As the haematoma breaks down by fib­
Inborn errors of metabolism investigations rinolysis and water is draw into the haematoma to form an
Child Protection Team effusion, there is marked expansion so that by 7 days it has
Paediatrician
undergone a significant enlargement and afte r 2 weeks it
Socia l wo rk
becomes isodense with brain and may be missed on CT or
Police

ultrasound scanning.142
Foren sic

Magnetic resonance imaging findings reflect the

INVESTIGATIONS FOR NON-ACCIDENTAL HEAD INJURIES
Table 14.5 lists the impol1ant investigations for all sus­
pected cases of NAH!. Additional investigations may be
necessary to exclude other diagnoses. The Royal College of
Radiologists recommends skeletal survey and mandatory
CT of the head in infants and children aged 0-2 years l40
and X-ray of the clinically suspic ious area in those aged
3-5 years . Skeletal survey is not generally indicated in
ch ildren over 3 years of age. Bone scintigraphy is indicated
in children over 2 years if the skeletal survey is equivocal
(see Chapte r 3) .
pathologica l consequences of a rotational acceleration(
deceleration injury. Barl ow et al 143 have identified subdural
haemo rrh age in all cases, which may be su btemporal, inter­
hemispheric, have torn surface veins, differing signal inten­
sity, compal1mentalized and suboccipital position. Brain
contusions, cerebral lacerations, petechial haemorrhage at
the grey-w hite matter junction and corpus callosum, focal
asphyxial insult, cerebral oedema, intraventricular and sub­
arach noid haemorrhages were also found .
Diffusion-weighted brain MR images were abnormal in
all suspected child abuse cases with subdural and retinal
haemo rrhages. 142
In CT(MRI comparative studies comparing inflicted with
non-i nflicted traumatic brain injuries it has become apparent
300 I Head and neck injuries

Figure 14.8 (a) Twenty-five days post admission: fluid-attenuated inversion recovery (FlAIR) coronal sequence showing posterior layering
within the subdural collection and high signal in the left parieto-occipital cortex. There is widespread encephalomalacia (with large subarachnoid
spaces, dilated ventricles and atrophy) and early gliosis. (b) Twenty-five days post admission: T2 transverse magnetic resonance imaging sequence
showing left temporal scalp haematoma, widespread high intensity white matter and encephalomalacia more marked on the right.

SPECTROPHOTOMETRY

Figure 14.11 shows absorbance of oxyhaemoglobin, at

413-415I1m, and bilirubin at 450-460nm, undertaken on
the subdural aspirate of a 2-month-old infant who pre­
sented with retinal haemorrhages, raised lCP and subdural
haemorrhage. The methaemoglobin absorbance, when
present, is identified at 405 nm. Although timing of the
bleed may be more specific with spectrophotometric analy­
sis of blood-stained CSF, the presence of bilirubin in sub­
dural aspirates is less precise and indicates that bleeding
occurred 24 hours to 3 days earlier.

TRAUMATIC BIRTH INJURY

Prenatal trauma in pregnancy may result in direct injury to

Figure 14.9 Four months post admission: fluid-attenuated
inversion recovery (FlAl R) corona I seq uence showi ng locu lations
in secondary subdural collections with rapidly evolving atrophy.
that there are more interhemispheric bleeds, more ventricu­
lomegaly, subdural hygromas and large extra-axial spaces in
the inflicted group, whereas axonal shearing injuries and
skull fractures are more common in accidental injuries.
the fetal brain or, more usually, indirect injUly as a result of
maternal circulatory or respiratory injury or retroplacental
haemorrhage. Direct injuries to the fetal brain are rare
because of the protective effects of the amniotic fluid but
when they occur they can result in fetal death or later
neuromotor disability, such as hemiplegia (see Chapter 9).
Birth injuries include: extracranial haematomata (caput
succedaneum, subgaleal haemorrhage and cephalo­
haematoma), skull fractures, intracranial haemorrhage ­
epidural, subdural, subarachnoid (less commonly
intra parenchymal or intraventricular), parenchymal contu­
sions, brainstem and spinal cord injury, and peripheral
nerve injury.
 Traumatic birth injury I 301

Figure 14.10 (a) Seven months post adm ission: fluid-attenuated inve rsion recovery (FLAIR) coronal seq uence showing mixed intensity
loculations bilaterally, suggestive of low-density clot within the collecti on . Overall brain atrophy. (b and c) Eigh teen months post
admission: FLAIR coronal and T2 transverse images demonstrating marked gliosis and residual cystic change particularly in the right
hemisphere. Loculations within the subd ural collections are not now visible.

Pericranial and intracranial birth injuri es are illus trated
in Chapter 8.
Caput succedaneum results from scalp oedema secondary
to pressure of the head on cervix. It resolves within a few days.
Subgaleal haemorrhage involves bleeding into the sub­
ga lea l space, often over the whole sca lp . It is an infrequent
complication of vacuum extraction or rotational forceps
delivery. A large volume of blo od may be lost, thus requir­
ing resuscitation.
Cephalohaematomata are haemorrhages beneath the
periosteum; they limited by its attachment to cran ial
sutures and are so metimes associated with skull frac tures
(10-2 5 per cent); 145 they are caused by compression to the
fetal head and usually resolve without sequ elae.
Skull fractures can result from instrumental delivery.
They are usu ally simple linear parietal fractures but,
if depressed, may impinge on the cerebral cortex, cerebel­
lum or brainstem.
302 I Head and neck Injuries

1.8 - direct;

1.6 - acceleration;

a 1.4 - deceleration;

2- 1.2 indirect;

Q.J
u
c: 1
co - linear acceleration/deceleration;

.D 0.8
a 0.6 - rotational acceleration/deceleration ;

'"
.D
<x: 0.4 • wh ipl ash/s haking injury ;
0.2 • combi natio ns of the above.
0
350 375 405 415 425 460 500 533 575
Wavelength (nm) PENETRATING HEAD INJURY

Figure 14.11 Spectrophotometric graph of a subdural aspirate.
The bilirubin peak indicates that blood has been present for
>24 hours.
Epidural haemorrhage is a rare bilih injury. It is usually
associated with a skull fracture and is usually extremely
small because of the tig ht application of the dura to the
skull bones in infancy.
Subdural haemorrhages may follow breech delivery,
cephalope lvic disproportion, vacuum extraction and pre­
cipitate delivery, which lacerate bridging veins or dural
venous sinuses. They may require surgical intervention or
develop as a subdural hygroma.
Subarachnoid haemorrhage occurs in full term and,
more frequently, in pre-term infants and is usually hypoxic
in origin. Choroid plexus haemorrhage or subependymal
venous haemorrhage extend into the ventricle system and
may subsequently extend into the subarach noid space,
around the poste rior fossa structures . There is a risk of sub­
sequent hydrocephalus about 2 weeks later.
Brain contusions resulting from forceps or compressive
damage to the head during bilih are rare but are associated
with skull fractures.
Brainstem and spinal cord injuries result from excessive
neck movements, particularly during breech delivelY or
other malpresentation. Fractures are not usually evident.
InjUlY results from traction or compression of the brain­
stem and spinal cord, or thrombosis of the vertebral or
anterior spinal artery. Lower cord injuries are rare but may
result from similar exaggerated spina l movements.
Peripheral nerve injuries result either from intrauterine
malposition with oligohydramnios, or traction on the
neck or upper limbs that produces Erb 's (with or without
Horner's syndrome), Klumpke's or other peripheral nerve
palsy.
PRIMARY MECHANISMS OF INJURY TO
THE BRAIN
There are several mechanisms by which head inju ry may be
sustained:
• penetrating injury;
• compression injury;
• impact injury;
Penetrati ng head injuries damage the brain by laceration,
haemorrhage and infection. Children may suffer penetrat­
ing head injuries as a result of gunshots, scissors, knives,
darts and pencils (particularly involving the tonsillar fossa
with carotoid artery injury). As they are non-rotational,
they a re non- concussive.
COMPRESSION INJURY
Compression is rarely a primary mechanism of NAH!.
The classical model for compress ion head injury is bilih
injury complicating malpresentation or cephalopelvic
disproportion.
The clinical patterns of traumatic birth injuries have
been outlined above . The mechanisms of intracranial dam­
age in compression injury are: (J) compression (impaired
cerebral blood flow, oedema, traumatic asphyxia, coning);
(2) moulding (l atera l moulding - sagittal sinus entrapment,
tearing of bridging veins, anteroposterior moulding ­
tearing of the vein of Galen, tentorium and the venous
sinuses); and (3) traction (cervical spine, brachial plexus,
vertebral arteries).
Compressional injuries are non-rotational and non­
shearing; therefore, they are non-concussive until cerebral
oedema, raised ICP and surface haemorrhages result (such
as subdural and subarachnoid haemorrhages).
IMPACT INJURIES
Direct acceleration impact injuries are exemplified by a
straight punch and direct deceleration/impact injuries dur­
ing a fall. Indirect linear acceleration/deceleration injuries
are typically from vehicular whiplash injuries an d indirect
or direct rotational acceleration/deceleration injuries that
commonly result from motor ve hicle collisions. Shaking is
a specific mechanism of injury that includes rotational
acceleration/deceleration components. In individual cases,
there is often more than one of the above mechanisms
operating.
Most accidental head injuries in children are impact
injuries. In acceleration injuries, the force is dissipated in
the resultant motion of the head. In deceleration injuries,
the force can only be ameliorated by a 'fo rgivin g' surface
or bounce. Deceleration injuri es produce the most severe
clinical features.

SHAKING AS A MECHANISM OF NON-ACCIDENTAL HEAD
INJURY IN INFANTS
There are five major components to the syndrome:
1. shaking causing rotation of the brain inside the sk ull ;
2. impact against a soft surface, causing acute
deceleration of the brain and vastly increasing the
rotational torque forces inside the skull (no external
evidence of impact);
3. hyperflexion and hyperextension injury from whiplash,
causing repetitive subluxation of the cervical
vertebrae; this results in injury to the cervical cord
(Fig. 14.12) and/or medulla;
4. shaking with hard impact, causing skull fractures and
contusions in addition to the rotational injuries;
5. primary hyp ox ic-ischaemic brain injury resulting from
apnoea due to brainstem injury.
PREDISPOSING FACTORS
A number of factors predispose the infant to injury from
shaking:
1. The young infant has a relatively large and heavy head
in relationship to the body size.
2. The neck muscles are weak and there is little head
control in early infancy. As the child gets older the
neck muscles become stronger so that the head moves
with the body; shaking whiplash injuries are rare,
therefore, after the second year of life.
3. The extracerebral space is relatively large
(physiological craniocerebral disproportion).
4. Th ere is physiological laxity of the meninges in
infancy compared with older children, with less
anchoring of the brain.
5. In comp lete myelination in the white matter w ith a full
complement of neuronal cell bodies in the grey matter
means a difference between the specific gravity of the
grey and the white matter and a gradient of tangential
accelerations that produ ces shearing.
BIOMECHANICAL DETERMINANTS OF SHAKII\IG INJURY
An external force has first of all to overcome the neck mus­
cles and induce anteropos terior movement with rotation.
This sets up shearing forces within the skull and within the
brain itsel f (rotational shearing forces). As with impact
injuries, the amount of damage will depend upon the rate
of acceleration/deceleration before the direction is
reversed. Soft impact brings the skull to a sudden halt with
maximum stretching of the bridging veins, thus the force
used is a major determinant of the injury sustained.
Sufficient force is necessary to Jift the weight of the infant
and accelerate it in space and pOSSibly add additional force by
active impaction of the child. The force with which the infant
hits the surface will be dependent upon the strength of the
individual; the angular forces (i.e. torque), which cause rota­
tion (see above) do not cause fractures but are of paramount
Primary mechanisms of injury to the brain I 303
importance in shaking injuries. The forces within the skull
resulting from shaking against a soft surface, such as a cot
mattress, are signi ficantly less than the forces arising from
being shaken against a hard surface, such as a wall.
Two biomechanical studies by Duhaime et al 146 and
Minns et al 147 found the angular accelerations to be of the
order of 1139 rads -2 (mean) and 200-892 rads -2 (range)
respectively.
The forces exerted by the shaker are dependent on the age,
height, gender and strength of the individual shaking, so that
young children (3 years) are unable to lift or shake a weight
equivalent to a 2-month-oJd child. Acceleration patterns
increase gradually to adult values by the age of 10 years.
The mean frequency for adult volunteers shaking
manikins was 3.5- 2.7 shakes per second; for children was
2.1-1.8 shakes per seco nd. Males shake significantly faster
than females (P < 0.0001) but the frequency is independ­
ent of the load.
In order to produce brain damage, a minimum duration
of shaking is required. It is likely that relatively fewer
shakes would be necessary to induce a whiplash type of
stem injulY than those required to produce damage from
repetitive rotation. There is also a maximum duration of
shaking that is possible before fatigue prevents further
shaking. Experimentally determined maximum possible
durations for shaking may not be exactly similar to dura­
tions or rates during situations associated with anger and
rage, but for ad ults shaking manikins that a re equivalent
in weight to a 2-month-old baby the median duration was
21.5 seconds; for a manikin equivalent to a 5-month-old
baby the median duration was 19.75 seconds. Duration
sustained was significantly longer in men than in women
and was related to the strength of the weakest arm and to
the height of the individual. It is, therefore, not possible
for adults to shake infants for prolonged uninterrupted
periods.
The centre of the rotational arc for shaking infants is
likely to be in the upper part of the cervical spine, between
the craniocervical junction and C2.
The posterior excursion of the head during shaking is
limited by the spinal joint ang les and ligamentous attach­
ments, which decelerate the movement more slowly than
the sudden deceleration from shaking with impact.
It is possible to shake fast with shOlter angles and slower
with greater angular excursions. The resultant brain dam­
age may be a product of the speed and degree of rotation.
The pericerebral space in infants can be up to 1 cm in
depth to allow for subsequent brain growth. This cranio­
cerebral disproportion with increased CSF spaces reduces
the inertia for any brain movement consequent on cranial
movement, and more swirling of the brain is possible with
stretching of the bridging veins. Measurements of the
sinocortical distance (from the superior sagittal sinus to the
cerebral cortex) were made during ultrasound scanning of
the brain, showing a change in this space as a result of a
change in the child's position, i.e. supine versus sitting.
 ~===7===A=C=C=ID=EN=T=A=L=H=EA=D=I=NJ=U=R=Y==~~~ll
MALPRESENTATION
or PEN ETR ATl ON
DISPROPORTION

IMPACT
BRAIN COMPRESSION PENETRATING INJURY
(AcceIera ti onl decel era ti on)

BRAIN DISTOR TI ON INDIRECT CONTUSION

CORTEX

Tracti on (stretch Cervical root

Overriding
and tea r) of falx pocket
parietal bones
and tentorium hae morrh ages

Tears of vein of
Galen and venous

Tearing bridg ing veins

Figure 14.12 Mechanisms of brain injury in different situation s. CBF, cerebral blood flow; CPP, cerebral perfusion pressure; EDH, extradural
 NON-ACCIDENTAL HEAD INJURY TYPES OF
HEAD INJURY

SHAKING WHIPLASHING MECHANISMS

OF INJURING

WHIPLASH
MECHANISMS OF
ROTATION BRAIN INJURY
SHAKIN G IN JURY

ANATOMICAL
CERVICOM EDU LLARY LEVEL
PERCU SS ION AND TRACTI ON

PRIMA RY
INJURY

SECONDARY
CHRONIC SDH ASPHYX IA
INJURY

INTRADURAL
HAEMORRHAGE
SPACE

OCCUPATION

RAISED ICP

Art eria l territory

(ACA. PCA. PICA
and pericallosa l)
infarct

(contd) haemorrhag e; ICP, intracrania l pressure; SAH, subarachnoid haemorrhage; SDH, subdural haemorrhag e.
306 I Head and neck injuries
PATHOPHYSIOLOGY OF SHAKING INJURIES
There are many intermedialY pathophysiological mecha­
nisms that will determine the final pattern of brain damage
(Fig. 14.13).
The esse ntial pathophysiological mechanisms are:
• surface shearing - cortical emissary veins - subdural
haemato ma;
• parenc hymal shearing;
- midbrain shearing injury;
- shearing of grey-white interface;
• cervicomedullary whiplash injury ;
- vascular injury: carotid compression and
ingress of water; (3) continued fresh bleeding; or (4) a true
vertebrals in whiplash;
inflammatory pachymeningitis elicited by red cell break down.
• ischaemia-hypoxia, secondary shock and cerebra l
oedema with raised rcp;
• contusion from impact injury.
SUBDURAL HAEMATOMA
A haematoma in the subdural space exerts pressure on the
underlying brain parenchyma with secondary ischaemia
to the immediately underlying white matter and, later, to
the grey matter. It also exerts pressure on the overlying
calvarium and may cause erosion of the inner table of the
skull.
An acute subdural haematoma occurs within 3 days of
injury a nd is usually associated with severe shock and
brain contusion as well as oedema. A subacute haematoma
occurs between 3 days and 3 weeks of injury, and a chronic
subdural haematoma appears more than 3 weeks from the
time of injury.14B
Figure 14.13 Primary brainstem haemorrhage at pontine level
in a 3-year-old girl, who fell 30 feet.
~ ---=-
CHRONIC SUBDURAL HAEMATOMA
A chronic subdural haematoma is often clinically silent
with only an inappropriately increas ing head circumfer­
ence. The causes of, and conditions predisposing to, a sub­
dural haem ato ma, are seen in Table 14.6.
A chronic subdural is not a true haematoma but a subdural
effusion in that it is mainly water with a haematocrit that is
usually less than 10 per cent. It is usually due to failure of
the acute haematoma to resolve. An acute haematoma may
become cill'onic because of: (1) its large volume; (2) a n
increased osmolarity of the subdural haematoma, causing
It has two components: a liquid low-haematocrit
'haematoma', with evidence of continued fresh bleeding, and
a membrane. The membrane, which is vascular and easily
bleeds, encapsulates the haematoma and binds it to the dura,
where it undergoes degradation and invasion by fibroblasts.
Calcification may be detected by 3 months. Incorporation of
the haematoma into the dura as a membrane is the basis of
the healing process. Brown staining of the dura, resulting
from deposition of haemosiderin during haemoglobin break­
down, persists for many months (see Fig. 11.1, p. 206).
Table 14.6 Conditions associated with, or predisposing to,
subdural haemorrhage or hygroma
Birth trauma
Accidental trauma
Non-accidental head injury
Aspergillosis
Kawasaki's disease
Osteogenesis imperfecta
Glutaric aciduria type 1
Meningitis (pneumococcal or group B streptococci)
Septicaemia
Haemophilia/factor 5 deficiency/anticoagulant treatment
Idiopathic thrombocytopenic purpura
Malignancy
Atrioventricular malformation or aneurysm
Post-cardiopulmonary bypass
Alagille's syndrome
Disseminated intravascu lar coagulation
Menkes' kinky hair disease
Slit ventricle syndrome
Hyperosmo lar dehydration from mannitol
Wide extra-axial space
Prema tu ri ty
Physiological
Familial 7Autosomal dominant
Renal dialysis
Severe dehydration
Congenital subarachnoid cyst
Chronic subdural haematoma
­ --
- - --

INTRADURAL HAEMORRHAGE
Intradural haemorrhages are not specific for NAHI ; they
are usually microscopic, venous/capillary in origin and
equivalent to intradural petechiae, and have been described
after fetal, perinatal and later neonatal death and have
been put fonrvard as a source of subdural haemorrhage in
infants.149 They are likely to be due to agonal anoxia in
many cases, although other causes of hypoxia may be
involved.
The role of intradural haemorrhage as a source of sub­
dural haemorrhage is an essential component of the 'uni­
fied theory' of causation of the triad of brain swelling,
subdural and retinal haemorrhages, which occurs as a
direct result of apnoea in cases when there is no pericranial
impact injury or injury elsewhere tha t might be interpreted
as inflicted. 149 A recent study of 82 infants with
hypoxic-ischaemic encephalopathy from a variety of
causes found no concomitant subdural haemorrhage in any
case, suggesting that other factors are necessary to produce
subdural haemorrhage in infants . 150
SUBARACHNOID HAEMORRHAGE
Approximately 20 per cent of infants with a NAHI have a
significant subarachnoid haemorrhage as well as a subdural
haemorrhage. The blood in the CSF may cause ischaemia to
the brain itself by inducing arterial vasospasm, and may
lead to secondary hydrocephalus requiring CSF diversion.
PARENCHYMAL SHEARING INJURIES
Midbrain Shearing Injury
The midbrain acts as pivo t upon which the cerebral hemi­
spheres can rotate, with the immediate effect of concus­
sion. In impact injuries associa ted with concussion, the
sudden coma is due to sudden rotation of the cerebral
hemispheres upon the more-anchored midbrain, i.e. a pri­
mary stem injury (Fig. 14. 14) .1 51 The clinical picture is of
dilated pupils, loss of vision, decorticate posture, akinetic
mutism and abnormalities of blood pressure, pulse and res­
piration. The midbrain may be damaged from the primary
injury described above or as part of a tentorial herniation
from raised ICP with brain stem haemorrhage (Duret haem­
orrhages). The reticular formation is very oxygen depend­
ent and consciousness is rapidly lost with lowered oxygen
tensions, i.e. hypoxia.
White Matter Shearing Injuries
The increase in brain weight over the first 4 years of life is
mainly due to an increase in white matter myelin. Grey
matter is firm and cellular, whereas poorly myelinated
white matter is more gelatinous and of slightly different
density; t his means that with 'shaking' the white and grey
matter swirl at different velocities, giving rise to character­
istic tears in the parenchyma for each type. 152
Secondary mechanisms of brain injury I 307
Strich l5J first described primary shearing of whi te mat­
ter. Long axons, such as those found in the commissural
fibres and long association tracts, can be avulsed by trac­
tion but axonal damage can also be seen in hypoxic
ischaemic damage. Further projection fibres from the cor­
tex may be sheared at the grey-white interface and be seen
as petechiae on MR imaging. 14J It is likely that both white
matter shearing and white matter oedema contribute inde­
pendently to the gross cerebral atrophy of the white matter
and ex vacuo ventricular dilatation.
The infant's axons may stretch more than fully myeli­
nated adult fibres l54 but, when sheared by traction forces,
they exude axoplasm, which appears as 'retraction balls'
on staining, followed by microglial scars after 2 weeks and
Wallerian degeneration after 6-8 weeks.1 55, 156
White matter shearing injuries are seen, in particular, in
the corpus callosum, the superior cerebellar peduncle and
in the midbrain.
It has alternatively been suggested that hypoxic
ischaemia, resulting in vascular axonal injury, may be
the main cause for the axonal damage in shaken baby
syndrome. 116,11 7
SECONDARY MECHANISMS OF BRAIN INJURY
In addition to primary brain injuries, patients with head
trauma also suffer from secondary brain damage, mostly of
ischaemic origin (Fig. 14. 15). These occur in response to the
initial trauma, resulting in secondary physiological deran<Je-
I~' b
ments such as hypOXia , low blood pressure, raised ICP,
reduced cerebral perfusion, pyrexia and disturbances of
global brain oxygen ex traction (Sj02)' These intracranial and
systemic secondary physiological insults may arise during
the initial resusci tation, or la ter in the reu, and cause inade­
quate delivery of oxygen and substrates (particularly glu­
cose), resulting in secondary brain ischaemia and further
damaging the already injured brain . I 50,159
Prior to advances being made in acute resuscitation and
transport of comatose head-injured patients, up to 30 per
cent of these patients were reported to have hypoxaemia
and 15 per cent had arterial hypotension on arrival at the
ARE department. 159- 164 In 1981, Miller's group 162 reported
a high frequency (53 per cent) of raised ICP (levels
> 20 mmHg) in head-injured patients during intensive
care, and data from the Traumatic Coma Data Bank (TCDB)
indicated t hat the incidence of raised ICP among severely
brain-injured adults requiring intensive care was even
higher. 165 In this study, 654 patients were examined and
72 per cent had ICP insult based upon end-hour ICP
recordings. 165 During their intensive care management,
head-injured patients may require to have repea t CT scans.
Andrews and co-workers l66 examined the number of sec­
ondary physiological insults occurring in TBI pa tients
being transported within the hospital, mos tly between the
ICU and CT scanning suite, and found a frequency of
308 I Head and neck injuries

(b)

Figure 14.14 (a) Haemorrhage in the spinal subdural space.

(b) Same case : axonal swellings in cervical nerve roots, beta amyloid
precursor protein staining. (Section courtesy of Professor JE Bell,
Edinburgh.)

Hae matoma

Cerebral oedema

Impairment of circulation

Rai se d ICP
Vasospasm

Infec ti on

Epil epsy

Hydrocephalus

Hypoxia

Hypercarbia

Hypotension

Fever

Ana emi a

Hyponatraemia

Hypoglycaemia

Figure 14.15 Primary and secondary brain injury. DAI, diffu se

axonal injury; ICP intracranial pressure; T81, traumatic brain injury.

50 per cent during tra nsit and > 65 per cent in the ICU dur­
ing th e subsequent 4 hours of continuous monitoring.
From the TCDB da ta , a significant increase in mortality
and morbidity w as associated with hy poxia or hypotension
from injUly throu g h resuscitation in the 717 adults who
were st udied after head injury.1 57 The relationship of out­
come and secondary physiological insults occurrin g during
the intensive care management after head injUly was deter­
mined in a prospective study involving 124 adults by Jones
and coll eag ues, 53 who used a computerized data collection
system to measure up to 14 clinica lly indicated physiolog­
ical v ariables (such as rcp, arterial blood pressure, CPP,
heart rate, oxygen saturation, and core and periph eral tem­
perature), minute by minute. 53 Abnormal valu es for each

~- -~~-
--~~~~-~
~

variable falling outside the pre-set threshold limits as
defined by the Edinburgh University Secondary Insult
Grading Scheme for ?5 minutes were analysed by insult
grade and duration. Ins ults were found in 91 per cent of
patients, regardless of the severity of head trauma, their
age or the admission injury severity score (ISS).53 The
cumulative durations were much greater than previously
recorded and the most significant predictors of m0l1ality in
this study were durations of hypotensive, pyrexic and
hypoxaem ic insults. 53 Hypotensive insults and pupillary
response on admiss ion were significant predictors of good
versus poor outcome. 53
Only a small amount of research has been done with
regard to investigating the clinical aspects and patho­
physiology in children with head injury apart from a few
specialized studies 167, 168 that primarily examined CBF,
cerebral hyperaemia, cerebral metabolic rate of oxygen
(CMR0 2) and arteriovenous difference of oxygen (AvDOJ,
This study found that cerebral hyperaemia was uncommon
(7 per cent) a nd that the CMR0 2 and cerebral oxygen
extraction was w ithin the normal range in 81 per cent of
children with severe head injuries. Cerebral metabolic rate
of oxygen and cerebral oxygen extraction fell between the
first and third day after the injury, and although the CBF
rose throughout the ac u te monitoring time this was not
significant. These findings led the investigators to conclude
that children were most vulnerable to secondary brain
damage short.ly after the injury. In addition, they found
that cerebrovascular autoregulation was preserved (signifi­
cant correlation between CPP and central venous pressure
[CVP] , P = 0.0003), the cerebrovascular resistance (CVR)
was normal (in 58 per cent), increased (in 32 per cent) and
could be impaired in the most severely injured. The ICP
was inversely propol1ional to the CBF (P = 0.009) and all
parameters (CMR0 2, CPP, 0 2/ partial pressure of arterial car­
bon dioxide [Paco2], cerebrovenous pH) were independent
determinants of CVR. 169
Another complicating factor for neurotrauma research
involving children is that the normal ranges for different
physiological param eters, such as ICp l70 and blood pres­
sure,171-174 are age related and change from birth through
childhood and adolescence to finally reach adult levels. Pre­
vious studies by the TCDB 23 grouped all children aged 0-15
years together and used the same end-hour threshold as that
used in their adult studies. Jackson's groupl75 attempted to
quantify the CPP secondary insults in children using only
three different adult CPP thresholds for their patients who
were aged 16 years or less. Similarly, Sharples and co-workers
used a single threshold value of > 20 mmHg to define raised
ICP in their study involving child ren aged 2- 16 years 01d. 166
In a more recent study, the cause and incidence of second­
ary insults in severely head -injured ad ults and children
were examined by Chambers and colleagues. 176 Again,
unified threshold values for CPP, ICP and arterial blood
pressure were used on their patients, regardless of their age
and developmental maturity.176 Using a standard definition
Secondary mechanisms of brain injury I 309
of abnormal physiological values, defining raised ICP,
hypotension and hypertension at all ages is too simplistic.
Jones and co-workers l77 from Edinburgh used age­
specific physiologica l norms to quantify secondary phys io­
logical derangements prospectively in 54 children aged
< 16 years with TBI. They found that secondary physiolog­
ical insults occurred frequently in children after head
injury, w ith rcp and CPP derangement demonstrated in
95.45 per cent and 77.27 per cent of patients respectively.
Furthermore, the duration of CPP derangement was found
to predict outcome (dead versus alive, P = 0.003).177 The
point at which physiolo gica I derangements become impor­
tant, i.e. by inducing ischaemi a in terms of magnitude and
duration, remains unclear, as does their effect on outcome.
A recent two-centre study 178 involving 86 head-injured
children, aged from 2 to 15 years, admitted within 24 hours
of injury, had prospective time series physiological data
downloaded from ICU monitors every minute for 6 physio­
logical variables (ICP, MAP, CPP, oxygen saturation, tem­
perature and heart rate) . Critical thresholds of CPP were
identified and, ideally, CPP should be kept above 48 mmHg
for children aged 2-6 years, above 54mmHg for children
aged 7-10 years, and above 58 mmHg for children aged
11-16 years. Duration of CPP was found to be the best pre­
dictor of survival (P < 0.000l), and a highly significant
discriminator between good (Glasgow Outcome Score
[GOSs] 4 and 5) and poor outcome (GOSs 1, 2 and 3)
(P = 0.005). A new cumulative pressure-time index (CPT)
was developed to include severity and duration and the
CPT was useful in predicting independent/poor outcome. 179
CEREBRAL OEDEMA
Cerebral oedema is defined as an increase in the volume of
the w hole (generalized) or part (focal oedema) of the brain
due to an increase in the water content. 180 Cerebral oedema
is a frequent accompaniment of shaking injuries and has
been estimated in as many as 42 per cent of such infants,
based on imaging. 181 The frequency is doub led in post­
mortem studies and Geddes ll6 ,117 confirmed that cerebra l
oedema was a major cause of death in 82 per cent with evi­
dence of hyp ox ic ischaemia in 77 per cent. It takes approx­
imately 6 hours to appear after the injury.!52
Cerebral oedema has several mecha nisms of production,
as see n in Table 14.7.182 The cause of the oedema in the
shaking injuries is probably multifactorial, with vascular
damage (as in the eye) causing vasogenic oedema, white
matter damage. disrupted venous damage from shearing,
damage to the blood-brain barrier, high central venous
pressure from chest compression, brain necrosis from
shock and imp aired perfusion, and secondalY hydro­
cephalus causing hydrostatic oedema from obstruction of
the arachnoid granulations by blood.
Cerebral oedema results in a deteriorating level of con­
sciousness. with signs of brainstem dysfunction in pulse,
respiration, blood pressure and pupils. which may progress
310 I Head and neck injuries

Table 14.7 Mechanisms of production of cerebral oedema

Intracellular (cytotoxic) Grey matter (especially astrocytes and oligodendroglia)

Energy failure of the cell (from hypoxia) causes membrane pumps to fail, i.e. inability to excrete
water or exclude sodium, plus intracellular proteolysis causes more idiogenic osmolysi s, finally
resulting in hydropic cells
Extracellular (vasogenic) White matter due to damage of blood-brain barrier with albumin leakage
Vasogenic Damage to the blood-brain barrier (from asphyxia and head inju ry) causes egress of glycine,
glutamate, noradrenaline and albumin into the brain; album in degrades to peptides with a strong
osmotic effect (cerebrospinal fluid [CSF] and protein similar to blood)
Hyd rostatic At the 'break-point' of cerebrovascular autoregulation an increased perfusion pressure is transmitted
to the microcirculation (from anoxic ischaem ic injury/hypercarbia or nitroprusside)
Hydrocephalic Re versa l of the no rma l transepend ymal flow of CSF back into brain parenchymal extracellular fluid
Osmotic Compartmentalized osmotic gradients cause shifts of fluid between: blood/ECF, blood/CSF,
CSF/extracellular fluid (ECF), and ECF/in trace llular fluid
Iatrogenic From inappropriate antidiuretic hormone, hypotonic fluids, disequilibrium syndrome from reducing
hypernatraemic states
Necrotic Ischaemia from any cause results in: (1) cellular necrosis and lysosomal rupture with release of
vasoactive peptides that are osmotically active; (2) release of thromboplastin, causing thrombosis
of the microvenous circulation and increasing vo lume of the infarct; and (3) cell membranes and
blood vesse ls be ing necrosed with no possib le osmotic gradient
Intramyelinitic (myelinac/astic) Toxic drug or metabolic effect

to signs of 'coning'. Coma itself maybe due to raised ICP or
to primary midbrain injury. Extensor hy pertonus follo wing
asphyxia may be due to oedema, but is more often a dys­
tonia as a result of basal ganglia involvement - i.e. post­
as phyxial rigidity, which is not influenced by treatment to
reduce ICP and which is often normal or secondary to
necrotic oedema. I S3
Imaging shows a reduction in lateral ventricular size
followed by third ventricle size and, eventually, the ambi­
ens cistern. The imaging loss of the 'sm iling face' means it
is imperative that a lumbar puncture is not performed as
there is some existing tentorial herniation. Scans show
hydrocephalic oedema as a hypod ense 'bat wing' around
the lateral ventricles. It is now possible to measure brain
water using MRI technology.
HYPOXIC/ISCHAEMIC INJURY
Ninety per cent of children dying from accidental head
injuries have evidence of severe hypoxic ischaemic brain
damage, 155 whereas 77 per cent of children dying from NAHl
are found to have hypoxic ischaemic damage at autopsy. 116, 11 7
The hypox ic/ ischaemic damage may be part of the pri­
mary injury (apnoea or medullary injUly) or result from
secondary insults such as hypotensive shock, raised ICP,
decreased perfusion pressure and seizures. The 'big black
brain' indicates particularly severe hypoxic ischaemic dam­
age, which is followed by death or the rapid development
of cerebral atrophy. Additionally, suffocation may have
occurred to quell the child's crying and hypoxic ischaemia
is almost always an agonal event in most infant deaths.
Cerebral Infarction from Impaired Perfusion
Brain infarction occurs as a result of hypotension,
raised ICP, brain shifts and vascular occlusions, spasm or
obstruction.
Glucose is the main energy source for the brain: under
aerobic conditions 38 molecules of ATP are produced but
only two are produced under anaerobic conditions. How­
ever, under anaerobic conditions the supply of glucose and
removal of lactic acid is essential. The ATP produ ced is
required for neurotransmitter formation, transpOli mecha­
nisms, maintenance of membrane pumps and polarization
of membran es.
The oxygen supply depends on the POz and the haemo­
globin concentration. Ischaemi a is far more dangerous
than hypoxia because there is now no glucose or ketones,
or any way of removing lactic acid, so neuronal necrosis
rapidly occurs.
Ischaemia initially results in swelling of the mitochondria
within the neurones and astrocytes as a result of water reten­
tion. The neurones release potassium, which is collected by
the astrocytes; glycogen also accumulates within the astro­
cytes, ca using further swelling. ls4 Second, there is leakage of
protein (and blood - haemorrhagic infarct) through the
endothelial cells into the extracellular space. Subsequent
breakdown of the red cells causes iron and bilirubin forma­
tion in the tissues, which is used to date the infarct. Finally,
lysosom al rupture, cell death and biochemical breakdown of
tissue proteins results in swelling of the infarct (further com­
pressing capillaries), breakdown of the blood-brain barrier,
loss of autoregulation, and sludging (from thromboplastin
release) and liquefaction of the tissues.
 Injury to the cervical spinal cord I 311

The necrotic tissue may be absorbed by macrophages vehicular accidents are most common causes in children,
leaving a cyst - as in periventricular leucomalacia - or may followed by sports-related injuries (e.g. rugby and horse
cause an astrocytic reaction, replacing the dead tissue with riding). Although the incidence is lower in children than
the formation of a gJial scar and capillary proliferation. in adults, it carries a worse prognosis and involves a high
Six to twelve hours after acute hypoxic ischaemia there incidence of severe, permanent neurological disability due
may be a therapeutic window, possibly due to a biochemical to spinal cord involvement. 190
cascade triggering apoptosis or a slowing down of metabo­ The anatomical and biomechanical characteristics of the
lism caused by adenosine and adrenocorticotrophic hormone. child's spine are very different from those of the adult and
the cervical spine is particularly at risk because of the high
weight and volume head-body ratios, ligamentous laxity
Patterns of Infarction and incomplete head control. Younger children have more
cervical lesions - the vertebrae are not yet ossified, the unci­
Hypotension does not result in uniform infarction of the
nate processes are flat and facet joints are horizontal, result­
whole brain. 18S The four feeding arteries to the circle of
ing in a different pattem of cervical spine injury in children.
Willis means there are areas where the pressures are equal,
Fractures involve the growth plates. 191 Zabramski et al 192
resulting in 'no flow', and ischaemia may occur in any
reported that 71 per cent of children with spinal injuries who
major vessel from Jow perfusion pressure.
were < 10 years old had cervical lesions, and that one-half of
When cerebral blood flow is severely impaired, water­
these lesions were between the occiput and C2.
shed zones between adjacent vascular territories are suscep­
There are six physical mechanisms that give rise to
tible. In the newborn this is between centrifugal and
spinal and spinal cord injUly:
centripetal arteries, causing periventricular leucomalacia. In
I. flexion dislocation;
the older patient it is between the middle and posterior cere­
2. flexion compression;
bral arteries (causing infarcts in the pericentral white mat­
3. compression burst fractures in which bone fragments
ter of the optic radiations and posterior temporal lobe) or
are exuded into the spinal canal;
between the anterior and middle cerebral arteries (causing a
4. spinal extension versus penetrating injuries;
wedge shaped infarct in the 'leg area'). This susceptibility is
5. distraction-rotation injury. 193
because cerebral arterioles are 'end arteries' and do not
anastomose.
Upper cervical spine injuries take the form of atlanto­
Patterns of infarction also result from brain herniations ­
occipital dislocation, from birth or other trauma, and odon­
pericallosal or subfalcine shifts, posterior cerebral (from
toid fractures that, in small children, are really epiphyseal
tentorial herniation) and posterior inferior cerebellar artery
detachments. Cervical dislocations and compression frac­
(PICA) from foramen magnum coning.
tures are most common in older children. Typical Jefferson
There is also selective vulnerability of cell types: neu­
fractures, Hangman fractures and cervical strains are
rones are the most sensitive, whereas microglia and blood
exceedingly rare.
vessel cells are the least sensitive. Ischaemia occurs in the
From these mechanisms of injury, primary spinal cord
brain-neocortex, basal ganglia, hippocampus and cerebel­
injlllY results from contusion and compression of the spinal
lar Purkinje cells. 186 In head injuries, hypoxic ischaemic
cord, which, within a few hours of injury, results in either
damage occurs to these selective areas as well as watershed
cord oedema with ischaemia or occlusion of the
zone infarction. 12
intramedullary vessels, resulting in haemorrhagic necrosis.
With resolution the necrotic areas leave cavitations or
syrinxes in the cord at the level of the injury. 194
INJURY TO THE CERVICAL SPINAL CORD
Pang and Wilberger llO described spinal cord injury with­
out radiographic abnormalities (SCrwORA), which is a
Injuries to the spinal cord in children are rare. They occur
traumatic myelopathy without evident fractures (due to
during birth, and after accidental and non-accidental
hyperflexion or hyperextension, or distraction or repetitive
injuries. In a study by Augutis and Levi 187 the incidence was
flexion-extension injuries), with primary damage to the
found to be 4.6 per million children per year (95 per cent,
cord, or cord injury and ischaemia secondary to hypoten­
confidence intervals 3.6-5.5); however, when pre-hospital
sion, or interference with the vertebral artery blood flow
fatalities were excluded the incidence was 2.4 (95 per cent
(bruising of the adventitia with intramural clots or occlu­
confidence intervals 1.8-3.1). Children's spinal injuries rep­
sion).195 Bollini 196 estimated that almost one-half of such
resent about 5-10 per cent of all spinal injuries.188.189 In
lesions were in the cervical region.
most series, boys predominate.
Spinal shock with a total flaccid paralysis below the
level of the region will last for a variable time after the ini­
ACCIDENTAL SPINAL INJURY
tial spinal injury until cord swelling resolves. Painful stim­
The aetiology of accidental spinal injury in children is dif­ uli wiU not elicit flexion of the limbs. The neurological
ferent from that in adults; falls and pedestrian-related findings will point to the classical syndromes of cord
312 I Head and neck injuries
involvement, i.e. complete, posterior cord or anterior spinal
artery territory, central cord, root involvement or Brown­
Sequard's syndrome cord lesion, whether the spine is stable
or unstable. 197
Treatment for the above spinal injuries may take the
form of halo distraction or other traction, occipitocervical
fixations with bone grafts or progressive traction with
transcranial tongs. Spine boards are not advisable because
they tend to force the large head into flexion, which can
compromise both airway and spine. Spinal cord injuries
in children are prone to result in spinal deformity because
of imbalance between the paralysed muscles involved in
postural control, or assymmetric lesions of growth plates.
Syringomyelia is common at the site of the injury but may
extend over several segments when the classical symptoms
of head and neck pain, made worse on straining, alert one
to the di agnos is. Survivors have a high frequency of bulbar
and associated brainste m deficits.
TRAUMATIC SPINAL INJURIES FROM BIRTH
Newborns suffer spinal lesions as a result of a traum atic
delivery due to breech and other malpresentation. Overall,
70 per cent of neonatal neck injuries result from breech and
30 per cent from cephalic deliveries. 198 Towbin 199 estimated
that spinal cord injury was respon sible for 10-33 per cent
of neonata l deaths 40 years ago. Additional ligamentous
and spi nal cord lesions are often found without fracture at
autopsy. These usual spinal injuries in the nevvb orn take the
form of atlanto-occipital and atlantoaxial dislocation, frac­
tures of the odon tOid, and complete transections of the
spinal cord. These injuries result in sudden death, tetra pare­
sis, bilateral Erb 's palsy, diaphragmatic and respiratory
paralysis, root pocket haemorrhage, spinal root avulsion,
spinal epidural haemorrhage and intraparenchymatous cord
haemorrhages. Newborns may be shocked, hypotonic and
apnoeic.
NON-ACCIDENTAL SPINAL INJURY
Caffey.11J in his origi nal description , labelled a NAHI as a
whip lash shaking injury. The pathology of spinal subdural
haemorrhages, in farcti ons and root avulsions from non­
accidental trauma are very similar to the spina l patholo gy
of newborns after bilth injury shown by Yates. 19S Both
trauma and whiplash are hyperextension injuries. Ged­
desI14.11S showed epidural spinal bleeding, localized axonal
damage to the craniocervical junction and damage to the
spinal nerve roots and brainstem in one- third of her fatal
cases.
The infant head pivots at approx imately the C2 level
and lower in the older patient, hence the cervicomedullary
damage. The posterior ligam ents and joint capsule are
more elastic in the infant and the cervica l spine is exten­
sile; traction will allow elongation, whereas the spinal cord
itself cannot stretch and aV1.llses. The cervical vertebral
column can move in the anteroposterior direction, allow­
ing a degree of slippage (spondylolisthesis). Whiplash
injury can therefore cause cord concussion fro m repe ated
partial subluxations, as well as traction with bruising and
vertebral artery and anterior sp inal artery lesions.
The presentation of infants with cervicomedullary
injury is with apno ea and seco ndary hypoxic damage.
Death from hypoventilation and apnoea may therefore be
from a primary injury to the respiratory centre in the
medulla or from associated vaso pa ra lytic shock w ith loss
of vasomotor tone, also from medullary injury. The result
is severe hypoxic ischaemic damage with oedema. These
cases highlight the need for the MRjCT scans to include the
cervical spine in all suspected cases and for the cord to be
examined at necropsy. Cullen 200 and Swischuk 201 described
compression fractures, subluxations, and fracture disloca­
tions in thoracolumbar levels (and one thoracic level
lesion) in 11 in fants suffering NAl.
GENETIC INFLUENCE ON RECOVERY FROM
TRAUIVIATIC BRAIN INJURY
Despite vigi lant neurointensive care, outcome remains
diverse in head-injured patients. Accumulating evidence
over the recent years indicate that genetic factors, particu­
larly apolipoprotein E (APOE) genotypes, may influence the
recovery after brain injury. 202.20J
The human APOE gene locus is located on chromosome
19 and has t hree allelic forms - epsion 2 (t:2), t:3 and t:4 ­
with a frequencies of 0.08, 0.77 and 0.15, respective ly, in
the general population. 204 lts gene product, apoE, is a 37­
kDa glycosylated protein, which forms a major component
of the lipid transport system in the brain,20s being synthe­
sized astrocytes in particular. 206
Teasdale and co-workers 207 examined the relationship of
APOE genotypes and functional outcome in 93 head-injured
patients and found that patients with APOE e4 allele were
more than twice as likely to have an unfavourable outcome
(dead, vegetative state or severe disability) 6 months follow­
ing the initial in su lt as those without it. This association has
been confirmed.202.20J
The influence of APOE genotypes on the conversion of
beta amyloid precursor protein (~APP) to beta amyloid
(~ A4) is another postulated mechanism that has gained
considerable interest in the last decade. In certain circum­
stances in predisposed individuals, ~APP may give lise to
deposits of :3A4, which is known to be toxic to neurones. 208
Beta amy loid depOSi t ions have been found in up to one­
third of patients who died from brain trauma. Nicoll's
group 209 found a hig her frequency of the APOE e4 allele
among thos e who had ~A4 deposition after head injury.
The majority of the scientific interest in APOE genotype
and brain trauma has focus ed thus far on ad ults and its
influence on the recovery of head-injured children remains
unknown.
 References I 313

REFERENCES
Teasdale G, Jennett B. Assessment of coma and impaired
consciousness. A practical scale. Lancet 1974; 2 :81-4.
2 Simpson D, Reilly P. Pediatric coma scal e. Lancet 1982; 2:450.
3 Raimondi AJ, Hirschauer J. Head injury in the infant and
todd ler. Coma scoring and outcome scale. Childs Brain 1984·
11 :J 2-35.
4 Jennett B. Teasdale G, Galbraith S et al. Severe h ead injuries
in three countri es.] Neu ral Neurasurg Psychiatry 1977 ;
40 :29 1-8.
25 Arnarson EO, Halld o rsson JG. Head trauma among children in
Reykjavik. Acta Paediatr 1995 ; 84:96-9.
26 Jennett B. Epidemiolo gy of head injury. ] Neurol Neurosurg
Psychiatry 1996; 60: 362 - 9.
27 Gabella B, Hoffman RE, Marine WW, Stallones L. Urban and
rural traumatic brain injuries in Co lorado . Ann Epidemiol
1997; 7:207-12.
28 Emanuelson I, Wendt L. Epidemiology of trauma tic brain
'
injury in children and adolescents in so uth- western Sweden.
Acta Paediatr 1997; 86:730-5.
29 Suominen P, Kivioja A, Ohman J e [ al. Severe a nd fatal

childhood trauma. Inju/y 1998 ; 29 :425-30.

5
Hutten loc her PRo Reye's sy ndrome: relati o n of o utco me to

30 Durkin MS, Laraq ue D, Lubman L, Barlow B. Epidemiology

therapy.] Pediafr 1972 ; 30:845-50.

and prevention of traffi c injuries to urb a n ch ildren and

6
Ses hia SS, Seshia MM, Sachdeva RK. Coma in childhood. Dev
adolescents. Pediatrics 1999 ; 10 3 :e7 4.
Med Ch ild NeuroI 1977; 19:614-28.
3 1 Morrison A, Stone DH, Redpath A et al. Trend ana lysis of
7 Morray JP, Tyler DC, J ones TK et al. Coma scale for us e in
socioeconomic differentials in deaths from injury in
brain-injured children. Crit Care Med 1984; 12:1018- 20.
childhood in Scotland, 1981-95. BM] 1999 ; 318:567-8.
8 Gordon NS, Fois A, Ja cob i G et al. The management of the
32 Thurman D, Guerrero J. Trends in hosp ita lization associated
comatose chi ld. Neuropediatrics 1983; 14:3-5.
with traumatic brain injury. ]AMA 1999; 282:954-7.
9 Yager lY, J ohns ton B, Seshia SS. Coma scales in pediatri c
33 Benoit R, Watts DD, Dwyer K et a1. Windows 99: a source of
pra cti ce. Am] Dis Child 1990; 144:1088-91.
suburban pediatric trauma.] Trauma 2000; 49:477 - 81;
10 Yates DW. RAD Lecture Notes on Emergency Medicin e.
discussion 481-2.
Oxford: Blackwell Scientific.
34 Thurman D. Traumatic brain inju ry in children and youth as a
II Tatman A, Warren A, Williams A, Powell JE, Whiteh ouse W.
public health pro b lem: an overv iew. In La ng lois JA (ed.)
Development of a modified paediatric coma scale in intensive
Traumatic Brain Inju ry in the United States: Assessing
care clinical practice. Arch Dis Child, 1997; 77:519-21.
Outcomes in Children. A tlanta, GA: Department of Hea lth and
12 J enn ett B, MacMillan R. Epidemiology of head injury. BM]
Human Services, Centers for Disease Control and Prevention
(c/in Res) 1981 ; 232:101 - 4.
2000, pp. 31-5. .
13 Kraus JF, Bla ck MA, Hessol N et al. The incidence of acute
35 Reid SR , Roesler JS, Gaichas AM , Tsa i AK. The ep idemiology
brain injury a nd se ri ous impairment in a defined population.
of pediatric traumatic brain injury in Minneso ta. Arch Pediatr
Am] Epidemiol1984 ; 119 :186- 201.
Adolesc Med 2001; 155:784-9.
14 Ann ege rs JF, Grabrow JD, Kurland LT, La ws ER Jr. Th e
36 Masson F, Thicoipe M, Aye P et al. Epid emiology of severe
in c idence, causes, an d secular trends of head trauma in
brain injuries: a prospective population-based study.
Olmsted County, Minnesota, 1935-1974. Neurology 1980;
] Trauma 2001; 51 :481-9.
30:912-19.
37 Adekoya N, Thurman DJ, White DD, Webb KW. Surveillance
15 Rantakallio P, von Wendt L. Trauma to the nervous system
for traumatic brain injury deaths: United States, 1989-1998.
and its sequelae in a one -y ea r birth cohort followed up to the
MMWR Surveill Summ 2002; 51:1-14.
age of 14 years.] Epidemiol Community Health 1985;
38 Williamson LM, Morrison A, Ston e DH. Trends in head injury
39:353-6.
mortality among 0-14 year olds in Scotland (1986-95).
16 Kraus JF, Fife D, Cox P et al. Incidence, severity, and external
] Epidemiol Community Health 2002; 56:285-8.
causes of pediatric brain injuly. Am ] Dis Child 1986;
39 Masson F, Thicoipe M, Mokni T et al. Epidemiology of
140:687-93.
traumatic comas : a prospective population-based study.
17 Chan BS, Walker PJ, Cass DT. Urban trauma: an analysis of
Brain Injury 2003; 17 :279-93.
1,116 pa ediatric cases.] Trauma 1989; 29:1540-7.
40 Andersson EH, Bjorglund R, Emmanuelson I, Stalhammer D.
18 Brookes M, MacMillan R, Cully S et a1. Head injuries in
Epidemiology of traumatic brain injury: a pOjlulation-based study
accident and emergency departments. How different are
in western Sweden. Acta Neu ral SCOlui 2003; 107:256-9.
children from adults? ] Epidemiol Co mmunity Health
41 Roberts I, Power C. Does the decline in child injury mortal ity
1990 ; 44:147-51. .
vary by social class? A co mpariso n of class-specific mortality
19 Sharples PM , Storey A, Ayn sley-Gree n A, Eyre JA. Causes of
In 198 1 and 1991. BM] 1996; 313:784-6.
fatal childhood accid ents invo lving hea d injury in northern
42 Reed lM, Browning JG, Wilkinson AG, Beattie T. Can we aboli sh
region, 1979-36. BM] 1990; 301:1193-7.
skull X-rays for head injury? Arch Dis Child 2005; 90:859-64.
20 Tepas JJ JIl, DiScala C, Ram enofsky ML, Barlow B. Morta lity
43 Minns RA, Millar C, Minns FC et al. Non-accidental head
and head injury: the pediatric perspective. ] Pediatr Surg
injury in Scotland : Scottish Database. In Minns RA, Brown JK
1990; 25:92-5; discussio n 96.
(eds) Shaken Baby Sy ndrome and Other Non-accidental Head
21 Hansen TB, Pless S, Gravers M. Head injuries among children
Injuri es. London: Mac Keith Press, 2005, pp. 154-84.
in the county of Ringkobin g. I. An a nalysis of hospital data.
44 Barlow KM, Minns RA. Annual incidence of shaken impact
Ugeskr Laeger 1991; 15 3 :2944-6.
syndrom e in young children. Lancet 2000; 356:1571-2.
22 Levin HS, Aldrich EF, Saydjari C et a1. Severe head injury in
45 Jayawant S, Rawlinson A, Gibbon F et al. Subdural
children: experience of the Traumatic Coma Data Bank.
haemorrh ages in infants: population-based study. BM] 1998;
NeurosurgelJ 1992; 31 :43 5-43; discussion 443-4.
317:1558-61.
23 AJdrich EF, Eisenb erg HM, Saydjari C et a1. Diffuse brain
46 Brown JK, Purvis RJ, Forfar JO, Cockburn F. Neurological
swelling in severely head-injured child ren. A report from the
aspects of perin ata l asphyx ia. Dev Med Child Neural 1974·
NIH Traumatic Coma Data Bank. ] Neurosurg 1992; 76:450-4.
16:567-80. '
24 Berney J , Favier J , Froid eva ux AC. Paediatric hea d trauma:
47 Whitby EH, Pal ey MN , Smith MF et a1. Low field strength
influence of age and sex. 1. Epid emiology. Childs Nerv Syst
magnetic resonance im ag in g of the neonatal brain. Arch Dis
J994; 10:514-16.
Child Fetal Neo /wta l Ed 2003 ; 88:F203-8.
314 I Head and neck injuries
48 Hold en KR, Titus MO, Van Tassel P. Cran ial magneti c
resonance imagi ng exam ination of normal term neonates :
a pilot study.) Child Neurol 1999; 14 :708 - 10.
49 Yonas H, Wo lfson SK Jr, Dujovny Met al. Selective
lenti culostriate occlusion in the primate. A hig hly focal
cerebral ischemia model. Stroke 1981 ; 12 :567 -72.
50 Shaffer L, Rich PM , Pohl KR, Gan esan V. Can mild hea d injUly
cause ischaemic stroke? Arch Dis Child 2003; 88: 26 7- 9.
51 Ko rs EE, Terwindt GM, Vermeu len FL et al. Delayed cereb ral
edema a nd fatal coma after minor head trauma: role of the
CACNA 1A calcium channe l subunit gene and relation ship with
familial hemiplegic migrai ne. Ann Neurol 2001; 49:7 53 - 60.
52 Umile EM, Sand el ME, AJvaria A et al. Dynamic imaging in
mild traumatic brain injury: support for the theolY of medial
tempo ral v ulne rability. Arch Phys Med Rehavil 2002;
83:1506-13.
53 Jones PA, Andrews P J , Midgley S et al. Measuring the burden
of seco ndary insults in head-inju red patients during intensive
care.) Neurosurg Anesthesiol 1994; 6:4-14.
54 Jiang JY, Gao GY, Li P, Yu MK, Zhu C. Early indicators of
prognosis in 846 cases of severe tra umatic brain injury.
) Nellrotrnuma 2002 ; 19:869-74.
55 Chiaretti A, Pia stra M, Pulitano Set al. Prognosti c factors and
outco me of children with severe head injury: an 8-year
expe rience. Childs Nerv Syst 2002; 18: 129-36.
56 Chiaretti A, De Bened ictis R, Della Corte F et al. Th e impact of
initial management on the outcome of children wit h severe
head inju ry. Childs Nero Syst 2002; 18:54 -60.
57 Coats TJ, Ki rk CJ, Dawson M. Outco me after severe head
injury treated by an integrated trauma system. ) Acc Emerg
Med 1999; 16 :1 82-5.
58 Zauner A, Bullock R, Di X, Young HF. Brain oxygen, CO 2 , pH,
and temperature monitoring: eva lu ation in the fe lin e brain.
Neurosurgery 1995; 37:1168-76 ; discussion 1176 -7 .
59 Jeevaratnam DR, Menon OK. Survey of intensive ca re of
severely head-injured patien ts in the United Kingdom. BM)
1996; 312:944- 7.
60 Matta B, Meno n D. Severe head injury in the United Kingdom
a nd Ireland: a survey of practice and impl ica tions for
management. Crit Care {"fed 1996 ; 24: 1743-8 .
61 Segal S, Gall ag her AC, Shefl er AG et a\. Survey of the use of
intracrani al press ure monitoring in children in the United
Kingdom. intensive Care Med 2001; 27:236 - 9.
62 Gambardella G, Zaccone C, Cardia E, Tomasello F. Intracranial
pressure monitoring in ch ildren: comparison of ex ternal
ventricular device wi th the fiberopt ic system. Childs Nem
Syst 1993; 9:470-3 .
63 Piek J, Bock WJ. Continuous monitoring of cereb ral tissue
pressure in neurosurgical practice: experienc es with 100
patients. ill tellsil!e Care IVIed 1990; 16: 104-8.
64 Yablon JS, Latner HJ, McCormack TM et a l. Clin ica l
experience with a fiberopti c intracrania l pres sure monitor.
) Ciill Moni t 1993; 9:171-5.
65 Shapiro S, Bow man R, Callahan J, Wolfla C. The fiberoptic
intra parenc hymal cerebral pressure monitor in 244 patients.
Surg Neurol 1996; 45:278-82.
66 Friedman WA, Vries JK. Percutaneous runnel ventri culostomy.
Summary of 100 procedures.) Nellrosurg 1980; 53:662-5.
67 Narayan RK, Kishore PR, Becker DP et al. Intracranial
pressure: to monitor or not to monitor? A rev iew of our
experience with severe head injury. ) Neuro surg 1982;
56:650 - 9.
68 North B, Reilly P. Compariso n amo ng three m ethods of
intracranial pressure recording. Neurosurgery 1986; 18:730-2.
69 Param ore CG, Turner DA Re lat ive risks of ventriculostomy
infecti on and morbidity. Acta Neurochir (Wi en) 1994;
127 :79-8 4.
70 Artru F, Terrier A, Gilbert let al. Monitoring of intracranial
pressure with intraparenchymal fiberoptic transducer.
Technical aspects and clinical reliability. Anll Fr Anesth
Real1im 1992; 11 :424 -9.
7J Pople IK, Mahlbauer MS, Sanford RA, Kirk E. Results and
complications of intracranial pressure monitoring in
303 children. Pediatr Neuro su rg 1995; 23: 64-7.
72 Jensen RL, Hahn YS, Ciro E. Risk factors for intrac ranial
pressure mon itoring in children with fiberoptic devices:
a critical rev iew. Surg Neuro/ 1997; 47:16- 22.
73 Meyer JS , Kondo A, Nomura F et al. Cerebral hemodynamics
and metabo lism following experimental head injUlY.
) Neuro surg 1970; 32:304-1 9.
74 Kaiser AM , VVhitelaw AG. Hypertensive response to raised
intracrani al pressure in infancy. Arch Dis Child 1988; 63:1461 -5.
75 Rossi S, Zanier ER , Mauri I et a l. Brain temperature, body core
temperature, a nd intracranial pressure in acute cerebral
damage. ) Neurol Neuro surg Psychiatry 2001; 71 :448-54.
76 Albrecht RF II, Wass CT, Lanier WL. Occurrence of potenti ally
detrim enta l tem perature alterations in hospita li zed patients at
risk for brain injury. Mayo Clin Proc 1998; 73:629-35.
77 Hutchinson PJ , al-Rawli PG , O'Connell MT et al. Head
inj ury monitoring using cerebral microdialysis and
Paratrend multipara meter sensors. leY/trolvl Neurochir 2000;
61 :88-94.
78 Gupta AK, a l-Rawli PG, Hu tchinson P J, Kirkpatrick P J. Effect
of hypothermia on bra in tissue oxygenation in patients with
severe head inju ry. Br) Allaesth 2002; 88: 188-92.
79 Smith LW, Fay T. Obse rva tions on human beings with cancer,
maintained at reduced temperature of 75- 90 degree
Fahrenhe it. Am) Ciin PatllOl 1940; 10:1-11.
80 Fay 1. Ea rly experiences with local and generalised
refrigeration of the human bra in.) Neurosurg 1959;
16:239 - 60.
81 Clasen RA, Pandolfi S, Russell J et a!. Hy pothermia and
hypotension in experimental cerebral edema. Arch Neurol
1968; 19:472-86.
82 Laskows ki EJ, Klatzo I, Baldwin M. Experimental study of the
effects of hy pothermia o n local bra in injury. Neurology 1960;
10:499- 505.
83 Rosom off HL, Gilbert R. Brain vo lum e and cerebrospinal fluid
pressure during hypothermia. Am) Physiol 1955; 18 3:19 - 22 .
84 Rosomoff HL, Shulm an K, Raynor R, Grainger W.
Experimental brain injury and delayed hypothermia.
Surg GYl1ecol Obstet 1960; 110: 27-32.
85 Clifton GL, AJlen S, Barrowdale P et a!. A phase II study of
mod era te hypothermi a in severe brain injury. ) Neurotrauma
1993; 10:263-71; discussion 273.
86 Marion DW, Oozeer R, de Vries L et a l. The use of mod erate
therapeutic hypoth ermia for patients with seve re head
injuries: a preliminary re port.) Neurosurg 1993; 79:354-6 2.
87 Cli fton GL, Miller ER , Choo i SC et al. Lack of effect of
induction of hypothermia after acute brain injUly. N Engl
) Med 2001; 344:556- 63.
80 Connell J, Oozeer R, de Vries L et al. Continuous EEG
monitoring of neon ata l seizures: diagnostic and prognostic
co nsideratio ns. Arch Dis Cilild 1989; 64:452 - 8.
89 Dusse r A, Na valet y, Dev ictor 0, Landrieu P. Short- and long­
term prognostic value of the electroencephalogram in children
with seve re head injury. Electroellcep iIalogr Clin Neurophysiol
1989; 73 :85-93.
90 Altafu ll ah I, Asa ik ar S, Torres F. Status epilepticLls: clinical
expe ri ence w ith two spec ial devices for continuous cereb ra l
mon itoring. Acta Neurol Scand 199 1; 84:374 - 8 1.
91 Aoki Y, Lombroso C. Progno stic valu e of
electroencephalography in Reye' s syndrome. Neurology 1973;
23:333-43.
 References I 315

92 Janati A. Erba G. El ectroencephal ograp hic correlates of

112 Mendelsohn D. Levin HS. Bruce D et al. Late MRI after head
near-drowning encephalopathy in children.
injury in children: rel ations hip to clinic a l features and
Electroencepi1alogr Clin Nellrophysiol 1982; 53: 182-9l.
outcome. Childs Nel'/! Syst 1992; 8:445-5 2.
93 Pampiglione G. Harden A. Resuscitation after
113 Caffey J. On the theo ry and practice of shaki ng infants. Its
cardiocirculatory arrest. Prognostic evaluation of ea rly
potentia l residual effects of permanent brain damage and
electroencephal ographic findings. Lancet 1968; I: I 26 1-5.
mental retardation. Am) Dis Child 1972 ; 124: 161-9.
94 Tasker RC. Boyd S. Harden A. Matthew DJ. Monitoring in
114 Guthkelch AN. Infantile subdural haematoma and its
non-traumatic coma. Part II. Electroencephalography.
relationship to whiplash injuries. BM) 1971; 2 :430-1.
Arch Dis Ch Ud 1988; 63 :895-9.
115 Minns RA. Brown JK. Neurological features ofNAHI and
95 Tasker RC. Boyd SC. Harden A. Matthew DJ. The cerebral whiplash shaken baby syndrome. In Minns RA. Brown JK
function analysing monitor in paediatric medical intensi ve (eds) Shaken Baby Syndrome alld Other NOll -accidental Head
care: applications and limitations. Intensive Care iVIed Injuries. London: Mac Keith Press. 2005. pp. 1-105.
1990; 16:60-8. Jl6 Geddes ]f. Vowles CH. Hackshaw AK et a1. Neuropathology
96 Lassen NA. Christiensen MS. Physiology of cerebral blood
of inflicted head injuIY in children. II . Microscopic brain
flow. Br) Anaesth 1976; 40:719-34.
inju ry in infants. Brain 2001; 124:1299-306.
97 Aaslid R. Markwalder TM. Nomes H. Noninva sive /17 Geddes JF. Vowles CH. Hackshaw AK et al. Neuropathology
transcranial Doppler ultra sound recording of flow velocity in of inflicted head injury in children. I. Patterns of brain
basal cerebral arteries.) Neurosurg 1982; 57:769-74. damage. Brain 2001; 124: 1290-8.
98 Weber M. Grolimund p. Seil er RW. Evaluation of
118 Barlow KJVI. Minns RA The relation between intracranial
posttraumatic cerebral blood flow velocities by
pressure and outcome in non-accidental head injulY. De1J
transcranial Doppler ultrasonography. Neurosurgery 1990;
Med Child Neuro11 999; 41:220-5.
27:106-12.
119 Ludwig S. Warman M. Shaken baby syndrome : a review of
99 Martin NA. Doberstein C. Zane C et al. Posttraumatic 20 cases. Ann Emerg Med 1984 ; 13: 104-7.
cerebral arterial spasm: transcranial Doppler ultrasound. 120 Schloff S. Mulhaney PB. Armstrong DC et al. Retinal
cerebral blood flow. and angiographic findings. ) Neurosllrg findings in children with intracranial hemorrhage.
1992; 77: 575-83. Ophthalmology 2002; 109: 1472-6.
100 Chan KH. Miller JD. Dearden NM. Intracranial blood flow 121 Reece RM. Nicholson CE. Inflicted Childhood
velocity after head injury: relationship to seve ri ty of injury. Neu rotrauma. Bethesda. MD: American Academy of
time. neurological status and outcome. ) Neurol Neurosurg Pediatrics. 2002.
Psychiatry 1992; 55:787-9l. 122 Goldstein B. Kelly MM. Bruton D. Cox C. Inflicted versus
101 Ungerstedt U. Microdialysis - principles and app lications for accidental head injulY in critically injured children. Crit Care
studies in animals and man.) Intern Med 1991; 230:365-73. Med 1993; 21:1328-3 2.
102 Hillered L. Persson L. Neuroc hemical monitoring of the 123 Kivlin JD. Simons KB. Laronitz S. Ruttum MS. Shaken baby
acutely injured human brain . Scand) Ciin Lab /Il1Jest 1999 ; syndrome. Ophthalmology 2000; 107:1246-54.
229(SuppL) :9-18. 124 Rao p. Carty H. Pierce A The acute reversal sign:
103 Hillered L. Persson L. Nilsson P et at Neurom etabolic comparison of medical and non- accide ntal injulY patients.
monitorin g of the ischa emic human brain using Ciill Radiol 1999; 54:495-501.
microdialysis. Acta Neurochir (Wim) 1990; 102:91-7. 125 Zimmerman RA. Bilaniuk LT. Bruce D et al. Interhemispheric
104 Hillered L. Valtysson J. Enblad P. Persson L. Interstitial acute subdural hematoma: a comp uted tomographic
glycerol as a marker for membrane phospholipid degradation manifestation of child abuse by shaking. Neuromdiology
in the acutely injured human brain. ) Neural Neurosllrg 1978; 16:39-40.
Psychiatry 1998; 64:486-9l. 126 Ewing-Cobbs L. Kramer L. Prasad Met a1. Neuroimaging.
105 Goodman JC. Valadka AB. Gopinath SP et a1. Extracellular physical. and developmental findings after inflicted and
lactate a nd glucose alterations in the brain after head injllJY noninfJicted traumatic brain injury in young childre n.
measured by microdialysis. Cnt Care Med 1999 ; Pediatrics 1998 ; 102:300-7.
27:1965-73. 127 Duhaime AC. Alario AJ. Lewander WJ et al. Hea d injulY in
106 Hovda. DA, Lee SM. Smith ML et al. The neurochemical and very young childre n: mechanisms. injury types. and
metabolic cascade following brain injury: moving from ophthalmologic findings in 100 hospitalized patients
animal models to man.) Neurofrauma 1995; 12:903-6. younger than 2 years of age. Pediatrics 1992 ; 90:179-85.
107 Nilsson OG. Brandt L. Ungestedt U. Saveland H. Bedside 128 Joffe M. Ludwig S. Stairway injuries in children. Pediatrics
detection of brain ischemia using intracerebral microdi a lysis: 1988 ; 82:457 - 6l.
subarachnoid hemorrhage and delayed ischemic 129 Boyles I. Minns RA. How Often Do Toddl ers Fall ancllnjure
deterioration. Neurosurgery 1999; 45: 1176-84; discu ssion Themselves? A prospect study. Special Study Module.
1184-5. University of Edinburgh. 2004.
108 Bullock R. Zaune A. Woodward JJ et al. Factors affecting 130 Hall JR . Reyes HM. Horvat M et a1. The mOl1ality of
exc itatory amino acid release follo wing severe human head childhood falls.) Trauma 1909; 29:1273-5.
injury.) Neurosurg 1998; 89:507-18. 13 1 Claydon SM. Fatal extradural hemorrhage following a fall
109 Bell MJ. Robertson CS. Kochanek PM et a1. Interstitial brain from a baby bouncer. Pediatr E111erg Care 1996 ; 12:432-4.
adenos ine and xanthine increase during jugular venous 132 Stoffman JM. Bass MJ. Fox AM. Head injuries related to the
oxygen desaturations in humans after traumatic brain injury. use of baby walkers. Can Med Assoc) 1984; 131:573-5.
Crit Care Med 2001 ; 29:399-404. 133 Plunkett J. Fatal pediatric head injuries caused by short­
110 Pang D. Wilberger JE Jr. Spinal cord injury without distance fall s. Am ) Forensic Med Patlwl 2001; 22: 1-1 2.
radiographic abnormalities in children. ) Neurosurg 134 Reiber GD. Fatal falls in childhood. How far must children
1982; 57:114-29. fall to sustain fatal head injury? Report of cases and review
III Marsha ll LF. Marshall SB. Klauber MR et al. A new of the literature. Am) Forensic Med Patho/1993; 14:201-7.
classification of head injury based on computerized 135 Milhorat TH. Pediatric neurosurgery. Contemp Neurol Ser
tomog ra phy.) Neurosurg 1991 ; 75(Suppl.):SI4-20. 1978 ; 16:1-389.
316 I Head and neck inj uri e s
136 Helfer RE, Siovis TL, Black M. Injuries resulting wh en sma ll
children fall out of bed . Pedia trics 1977 ; 50:533-5.
137 Margulies SS, Thibault KL Infant skull and suture properties:
measurements and implications for mechanisms of pediatric
brain injury. J Biol11ech Eng 2000; 122 :364-71.
138 Scarfo GB, Mariottini A, Tomaccini D, Palma L. Growing
sku ll fractures: progressive evolution of brain dam age and
effecti ve ness of surgica l treatme nt. Ch ilds Nen) Syst 1989;
5:163-7.
139 Carty H. Training in paedi atri c radiology. Clin Radial 1989;
40:227-8.
140 Royal College of Radiologists. Guidelinesfor Doctors, 5th
ed n. London: Royal College of Rad iologists, 200 3.
141 Zi mmerman RA, Bilianiuk LT, Bruce D et al. Computed
tomography of craniocerebral injuI)' in the abused chi ld.
Radiology 1979; 130:687-9 0.
14 2 Johnson H, Fae rber EN (ed.). Trauma. In CNS Magnetic
Resonance Imaging in Infants and Children. Lond on : Mac
Keith Press, 1995, pp. 98- I I 5.
143 Barlow KM, Gibson RJ, McPhillips M, Minns RA. Magnetic
reso nance imaging in acute no n-accid e ntal head inju ry. Acta
Paediatr 1999; 88:734-40.
144 Biousse V, Suh DY, Newman NJ et a l. Diffusion-weighted
magnetic reso nance imagin g in shaken baby syndrome.
Am J Oph tha11110 12002; 13 3:249- 55 .
145 Hovi nd KH. Traumatic brain injuries. In Raimondi AJ, Choux
M, DoiRocco C (eds) Head Injuries ill the Newborn and
Infant. New York: Springer-Verlag, 1986, p. 87.
146 Duhaime AC, Genarelli TG, Thibau lt LE et aJ. The shaken
baby synd rome. A clinical, pathological, and biomechanical
study. J Neurosurg 1987; 66 :409- I 5.
147 Morison CN, Minns RA. Biomechanics of shaking. In Minns
RA, Brown JK (eds) Shaking and Other Non-accidental Head
Inju ries ill Children. London: Mac Keith Press, 200 5, pp.
105-46.
148 Choux M, Lena G, Genitori L. Intracranial hematomas. In
Raimondi AJ, Choux M, DoiRocco C (eds) Head Iiljuries
in the Newborn and Infan t. New York: Springer-Verlag,
1985, p. 203.
149 Gedd es JF, Tasker RC, Hackshaw AK et aJ. Dural
haemorrhage in non-traunJatic infant deaths: does it explain
the bl eeding in 'shaken baby syndrome'? Neuropathol Appl
Neurobiol 20 03; 29: 14-2 2.
150 Byard RW, Blumberg P, Rutty G et al. Lack of evidence for a
causal relationship between hypox ic/ischaemic encep halopathy
and subdural haemorrhage in fetal life, infancy and ea rly
childhood. Pediatr Dev Pathol 2007; 10:348-50.
151 Ommaya AK, Gennarelli TA. Cerebral concussion and
traumatic unconsciousness. Correlation of expe rimental and
clinica l observ ations of blunt heael injuri es. Brain 1974;
97 :633-54.
152 Jaspa n T, Narbrou gh G, Punt J A, Lo we J. Cerebral
contusional tea rs as a marker o f child abuse: detection by
crania l sonography. Pediatr Radial 199 2 ; 22:237 -45.
153 Strich SJ. Shearing of nerve fibers as a course of brain
damage due to head inju I)'. Lan cet 1961; 2:443-8.
154 Crompton R. Closed head injuI)' in children. In Cromp to n R
(ed.) Closed Head InjUly in Children. Lo ndo n : Edward
Arno ld , 1986.
155 Adams JH, Mitchell DE, Graha m Dl, Doyl e D. Diffuse brain
damage of immediate imp act type. Its relatio nship to 'primaI)'
brain -stem damage' in head injuI)'. Brain 1977 ; 100:489-502.
155 Vowl es H, Scholtz CL, Cameron JM. Diffuse axona l injuI)' in
ea rly infa ncy. J Clin Pathol 1987 ; 40: 185-9.
157 Ches nut RM, Marshall LF, Klauser MR et al. The role of
secondaI)' brain injuI)' in determining outcome from severe
head injuI)'. J Trauma 1993; 34:216- 22 .
158 Graham Dl, Ford I, Adams JH. Ischaemic brain damage is
still common in fatal non-missile head injuI)'. J Neural
Neurosurg Psyclliatry 1989; 52:346-50.
159 Miller JD , Bec ker DP. SecondaI)' insults to the inj ured brain.
J R Call Surg Edill 1982; 27:292-8.
160 Gentleman D, Jennett B. Haz ard s of inter-hospital transfer of
comato se head -injured pati ents. Lan cet 1981; 2:853-4.
161 Kohi YM, Mendelo w AD, Teasdale GM, Allard ice GM .
Extra c ra ni al insults and outcome in patients w ith ac ute head
inju I)': relat io nship to th e Gla sgow Coma Sca le. Inju ry 1984 ;
15 :25-9.
162 Miller JD, Bu tte rworth JF, Gud ema SF et al. Furthe r
experien ce in the managem ent of severe head injuly.
J Ne urosurg 1981; 54:289-99.
163 Miller JD , Sweet RC. Na raya n R, Becke r DP. Early insults to
the injured brain. JAMA 1978 ; 240:439-42.
164 Price DJ, MUlTay A. The influence of hypoxia and hypotension
Oll recov ery fro m head inju I)'. InjUly 1972; 3:218-24.
16 5 Marmarou A, Anderson RL, Ward JD et al. NIND S Traumatic
Coma Data Bank : Intracranial press ure moni toring
methodo logy. J Neurosurg 199 1; 75 (Suppl.): 21 -7 .
166 Andrews PJ, Piper IR , Dea rd en NM, Miller JD . Seco nda I)'
insults during intrahospital transport o f head -injured
patients. Lancet 1990; 335:327-30.
167 Sha'lJles PM, Matthews DS, Ey re JA. Cerebra l bloo d flow
and metabolism in children wit h severe head iIJjuI)'. Part I:
Relation to age, Glasgow Coma Score, outcome, intracranial
pressure, and time after inju I)'. J Neural Nellrosu rg
Psychiatry 1995; 58:145-5 2.
168 Muizelaar JP, Ward JD, iVlarmaron A et al. Cerebral blood
flow a nd metabolism in severely heael-injured children. Part
1: Relationship with GCS score, outcome, ICP, and PVJ.
J Neurosurg 1989 ; 71:63-71.
169 Sharples PM, Matthews DS, Eyre JA. Cerebral blood flo w
and metabolism in children w ith severe head injuries. Part 2:
Cerebrovascu lar resistance and its determ inants. J Ne urol
Neurosurg Psychiatry 1995; 58:153-9.
170 Minns RA. Infectious and para infectious encep halopathies.
In Minns RA (ed.) Problems of Intracrallial Pressure in
Childhood. London: Mac Keith Press, 1991 , pp. 170- 282 .
171 Report of th e Second Task Fo rce on Blood Pressure Control
in Children: 1987. Task Force on Blood Pressu re Control in
Children . National Heart, Lung, and Blo od Institute,
Bethesda, i\IlaI)'land. Pediatrics 1987; 70:1- 25.
172 Donmall Me. Anthropom et ric and blood press ure studies in
children fro m Shetland: reliability of measurement and the
use of heig ht- we ig ht indi ces. Acta Paediatr SC(ll1d 198 5;
3 18 (Suppl.) :23-35.
173 Ro cce ll a EJ. Update on the 19~7 Task Force Report on Hi gh
Blood Pressure in Children and Adolescents: a working
group report from the National High Blood Pressure
Educatio n Program. National Hig h Blood Pressure Education
Program Wo rkin g Group o n Hypertensio n Control in
Children a nd Adolescents. Pediatrics 1995; 98:649 -58.
174 Rosner B, PrineasRJ , Logg JM, Daniels SR. Blood pressure
nomog rams for children and adolescents, by height, sex, and
age, in the United States. J Pediatr 1993; 12 3 :871-85.
175 J ackso n S, Piper JR , Wagsta ff A, Souter M. Quantification of
cerebra l perfu sion pressure (CPP) 'secondaI)' ins ults' in
child ren: a study of the effec ts o f using diffe ren t CPP
thres holds. British Journal of Neu roSlirgery, 1999; 13: 104.
175 Chambers IR, Treadw ell L, Mendelow AD. The cause and
incidence of secon daI)' insults in severely head-injured
adu lts and children. Br J Ne urosll rg 2000; 14:424-31.
177 Jo nes PA, A ndrews PJ, Eastman VJ, Minns RA. Traumatic
brain injuI)' in chi ldhood: Intensi ve care time-series data
and o utcome. Br J Neurosurg 2003 ; 17: 29 -39.

178 Jones PA, Easton VJ, Anderson PJD et al. Traum at ic brain
injury in chi ldren: intensive care time series data and
outcome. Br J Neurosurg 1982; 17:29-39.
179 Ch ambers IR, Jones PA , Lo TYM et al. Critical thresholds of
ICP and CPP related to age in paediatric head inju ry. J
Ne urol Nellrosllrg Psych 2007 ; 77:234-40.
180 Pappius HM, Fe ind el W. Dyna mics of Brai l! Edema. Berlin:
Spri nge r, 1976.
181 King WJ, MacKay M, Sirnick A. Shaken baby sy ndrome in
Canada: clinical cha racteristics and outcomes of hospital
cases. Can Med Assoc J 2003 ; 168:155-9.
182 Brown JK, Imam H. Interrelat ions hips of liver and brain with
special reference to Reye syn drome. J Inherit Me tab Dis
1991; 14:436- 58 .
183 Seshia SS, PN Chow, Sanka ran K. Coma foJlowin g
cardiorespiratory arrest in chi ldhood. Dev Med Child Neurol
1979; 21:143-53.
184 Fujimoto T, Walker JTJ , Spatz M, Klatzo I. Pathophys iol ogic
aspects of ischaemic edema. In Pa ppius HM , Feindel W (eds)
Dynamics of Brain Edema. Heidelberg: Springer, 1976,
pp. 171-80.
185 Bruce DA, Langfil l m , Mi ll er JD et al. Regional cerebral
blood flow, intracrani al pressure, and brain metabolism in
comatose patients. J Neurosurg 1973; 38:131-44.
186 Dearden NM. ls chaem ic brain. Lancet 1985; 2:255-9.
187 Augutis M, Levi R. Pediatric spinal cord injury in Sweden :
incidence, etiolo gy and ou tcome. Spinal Cord 2003;
41 :328-36.
188 Hamilton MG, My les ST. Pediatric sp in al injury: review of 61
deaths. J Neurasurg 1992 ; 77 :705-8.
189 Ham ilton MG, Myles ST. Pediatric spinal injury: review of
174 hosp ital ad miss ions. J Neu rosurg 1992; 77 :700-4.
190 Splendiani A, De Amicis R, Pupill o V et a l. Spi nal trauma in
childho od: a four-y ea r experience at a non-specialis t
hospital. RiIJista di Neuroradi%gia 2003; 16:455-7.
191 Aufderma ur M. Spinal injuries in juven il es. Necropsy
find ings in twelve cases. J Bon e Joint Surg Br 1974;
56B:513-19.
192 Zabramski JM, Hadl ey MN, Browner CM. Pediatric spinal
co rd and vertebral column injuries. Ban-ow Neural [nst Q
1986; 2:11 -17.
193 Lena G, Bollini G. Spinal inju ries in children. In Chou x M
et al (eds) Pediatric Ne urosurgery. Lond on: ChurchiJi
Livings tone, 1999, pp. 313 1-91.
References I 31 7
194 Tator CH, Koyanagi T. Vascular mechani sms in the
pathophysiology of human spinal cord injury. J Neurosurg
1997; 86:483 - 92.
195 Yates PO. Birth traum a to the vertebral arteries. Arch Dis
Child 1959; 34:436-41.
196 Bollini G. Thoracic and lumbar spine injuries in children. In
Fl oman Y, Fa rcy JPC, Arge nson C (eds) TllOra coiuml1ar Spine
Fractures. New York: Rave n Press, 1993, pp. 307 - 25.
197 Holdsw oli h F. Fractures, dislocations and
fracture-di slocations of the sp ine. J BOl1e Joint Surg (Am)
1971; 52A:1 534-S I.
198 Jones L. Birth trauma and the cervical spine. Arch Dis Child
1970; 45:1 47 .
199 Towbin A. Late nt sp in al co rd and brain stem injury in
newborn in fants . Del) }\lIed Child Neurol 1969: 11: 54- 68.
200 Cullen JC. Spina l lesions in bartered babies. J Bone Joint
SurgelY (Br) 1975; 57B :364-6.
20 1 Sw ischuk LE. Spine and spin al cord trauma in the battered
child synd ro me. Radiology 1969; 92 :733 - 8.
202 Friedman G, Froolll p, Sazbon L et aJ. Apol ipoprotein
E-epsilon4 genotype pred icts a poor outcome in surv ivors of
traumatic brain injury. Neurology 1999; 52:244- 8.
203 Lichtman SW, Seliger G, Tycke B, Marder K. Apolip oprotein
E and functi onal recovery from brain injury followin g
posta cute rehabilitation. Neu rology 2000; 55: 1536-9.
204 Utermann G, Langenbeck U, Beisiegel U, Weber W. Genetics
of the apolipoprotein E system in man. Am J Hum Genet
1980; 32:339-47.
205 Mah ley RW. ApoJipoprotein E: chol esterol transpOli protein
with expandin g role in cell biology. Science 19813;
240:622-30.
206 Boyles JK , Pitas RE , Wil so n E et aJ. Apo lip oprotein E
associated with astrocy1ic glia of the central nervous system
and with nonmyellnating glia of the peripheral nervous
system. J Clin [n/lest 1985; 76:1501-13.
207 Teasdale GM, Ni cho ll JA , Murray G, Fiddes M. Associati on
of apolipoprotein E polymorphism with outcome after head
injury. Lancet 1997; 350:1069 -7 1.
208 Mattson MP. Cellul ar actions of beta-amyloid precursor
protein and its solubJe and fibrillogenic deri va tives. Physiol
Rev 1997; 77: 1081-132.
209 Nicoll JA, Robelis GW, Graham DJ. Apo lipoprotein E epsi lon
4 allele is associated with deposition of amyloid beta-protein
fo llowi ng head inj ury. Natu re Med 1995; 1:135-7.
I CHAPTER 15 I
HEAT-INDUCED INJURY OR DEATH
Anthony Busuttil
Introduction 318
House fire deaths 318
INTRODUCTION
diminution in the incidence of domestic fire deaths since
Deaths in fires are the third leading cause of death in domes­
tic fatalities among children.! In the USA someone dies in a
fire about every 162 minutes and someone is injured in a fire
every 32 minutes, and many of these victims are chi ldren ;2
t he USA ranges only sixth in the incidence of deaths from
fires among the 25 developed coun tries for which statistics
are available.) In retrospect, many of these deaths are found
to have been preventable 4 and this raises a substantial pub­
lic health problem. 5 - 7 Financial and insurance losses are also
vely substantial.
Thermal injuries are also a common reason for referral
of children to the emergency services, with scalding being
the commonest reason for such referrals. Heat-induced
inju ries also form an important part of the gam ut of the
injuries seen in children who have been physically abused
by their carers.
HOUSE FIRE DEATHS
In the UK, house fires rank second after road traffic fatalities
as a cause of accidental deaths in children below the age of
15 years. When a house fire develops, children are more vul­
nerable, partly because they depend on the ad ults around
them to assist them out of the fire as they may be too small
and inexperienced to be capable of helping themselves and
extricating themselves out of the fire. The installation of
smoke alarms in houses has been a major reason for the
The pathologist's role 319
References 327
1987. Poor housing particularly associated with overcrowd­
ing and poorly built houses present a high lisk of fire deaths.
Single-parent families and families from a lower socioeco­
nomic background, particularly in associatio n with multi­
deprivation, show a much higher incidence of fire fatalities.
The misuse of a lcohol, and contro ll ed substances to a
lesser extent, by the adults caring for the child increases
the incidence of deaths in domestic fires, given that the
adults in such conditions succumb earlier to the fire and,
indeed, being a lready disorientated due to the intoxication
they are much less able to look after t hemselves and after
the child ren whom they are caring for. This asso ciatio n has
been shown to o'ccur both in the UK and in the USA.
Smokers' materials such as cigarettes, matches and
lighters are a very frequent cause of house fires. This has
led some countries to ensure that cigarettes lighters are
manufactured in a child-proof manner.
Deaths in fires usually are the result of smoke inhalation
rather than burns. s Although consumption of ambient
oxygen by the fire with environmental deoxygenation is
important, t he inhalation of carbon monoxide (and carbon
dioxide) is more frequently what causes death: thus smoke
detectors are valued as preventative measures. Smoke
detectors are widely used in several countri es and foun d
to be effective. Their use is becoming more widespread in
Europe and in the USA,9,lO and they have been sho wn to
have an important role in the prevention of conflagrations
both in public and private housing. II At present, regul a­
tions throughout the European Community tend to impose

the use of smoke alarms in public buildings rather than
homes where children live but there may be good reason
for extending their scope.
The banning of upholstery, certain items of furniture
and other fire-retardant items whose combustion produces
cyanide and other toxic compounds has also gone some
way in reducing deaths from such sources.
Fires and flames are a significant cause of death in
childhood in many countries in Europe. 12 Although there
has been a striking fall in the number of children who have
died from fire in recent years, the death rate from fires
remains unacceptably high. A major component of the
reduction has been the fall in the number of deaths from
the ignition of clothing following flame-proofing regula­
tions and the reduction in use of open fires. There have
been few specific studies of deaths from fire and flames at
the European level.
Each year in the USA, an estimated 700 children aged
5 years and under die in house fires. Representing 20 per cent
of the fire deaths each year, this age group has a fire risk
that is double the national average. Children playing with
matches, cigarette I ighters and other fire sources are thought
to have started about 91810 fires per year from 1993 through
1997, which resulted in an estimated 338 deaths and 2624
injuries each year. Pre-school children are also the most
frequent victims of fires started by children playing with
smokers' materials (over 30 per cent of the fires) and they
constitute 20 per cent of the total number of fire death vic­
tims. The United States Fire Administration (USFA) encour­
ages parents to teach children at an early age about the
dangers of playing with fire in an effort to prevent child
injuries, fire deaths and future fire setting behaviour.
Every year, more than 100000 household fires are
reported in the UK and more than 600 people die in these
fires. One in ten of these fire victims are children. The num­
ber of house fires in Scotland is 'markedly different' from
England, Wales and Northern Ireland according to the latest
government figures. In Scotland 46 per cent of all fires are
house fires, compared with 33 per cent in Northern Ireland
and 32 per cent in England and Wales. Broken down by fire
brigade areas, the figures show that all Scottish brigades
recorded that at least 38 per cent of all fires were house fires.
Outside Scotland, the largest percentage of house fires was in
Greater London. The Home Office research, made available
to chief fire officers across the UK, shows that around 75 per
cent of all fire casualties are in house fires: 466 people died
in fires in the home and another 14600 people were injured.
The number of firefighters injured - 655 men and women ­
was the lowest recorded for over 20 years, with over 50 per
cent being physical injuries rather than burns. There are
more than 100000 fires deliberately started in the UK each
year; some of these are targeted at domestic premises for a
variety of reasons.
In 2003, 111 people died in Scotland because of fires and
another 2300 people suffered non-fatal injuries. Misuse of
alcohol was a major contributory factor in 41 out of the
The pathologist's role I 319
76 fatal fires; carelessly throwing away cigarettes, matches
and lighters caused 39 out of the 76 fatal fires. Smoke
alarms were fitted in 44 out of the 76 homes where people
died, but only six alarms were known to have operated
properly; in the other 38 cases, 15 did not have a battery
fitted and three had flat batteries. Fire investigators said
that 39 out of the 85 house fire deaths were potentially
avoidable if smoke alarms had been fitted and in working
order. Of these fatal fires 34 started in the living-room, 21
in the bedroom and 13 in the kitchen. Polyurethane foam­
filled furniture contributed to 14 fatal fires.
Over 42000 children under the age of 15 years were
injured in the UK from burns and scalds in 2002, with
28000 of them being under 5 years of age and 95 per cent
of the thermal injuries occurring in the home. Over 50 per
cent of all severe burns and scalds happened in the kitchen.
Hot liquids were the cause of 70 per cent of the heat­
induced injuries.
THE PATHOLOGIST'S ROLE
Key questions to be addressed by the pathologist in fire
fatalities are: Can the victim be positively identified? Was
the victim actually alive or already dead at the time the fire
started? Was the death directly related to the fire and
caused by it? What exactly is the cause of death? Why was
the victim unable to escape the fire?
To respond to these questions the following information
must be correlated: (1) the circumstances preceding the
death; (2) the anatomy of the fire, i.e. evidence of the origin,
development and nature of the fire; and (3) the autopsy
findings.
When examining a fire it has to be kept in mind that fire
artefacts occur to varying extents in any body exposed to
a fire, irrespective of whether death was a result of the fire
or whether the person was already dead when the fire
started. These may be further aggravated at the time of
the rescue and recovery of the body from the site of the fire
occurs. These 'are:
• The 'pugilistic attitude' of the body. Characteristically,
the flexor muscles in both the upper and lower limbs,
which are on the whole stronger than the extensors,
will be damaged by the fire and contract, owing to the
heat -induced effects on the component proteins of the
muscles. This results in the body assuming a pugilistic
(boxer) or 'sit up and beg' posture, with the four limbs
solidly flexed.
• Loss of fluid from the tissues results in their
desiccation and brittleness, not least the skin and of
any other exposed underlying tissues that thus tend to
crumble on pressure and touch.
• Body weight may decrease with a loss of up to 60 per
cent, also mainly due to loss of water.
• Bone fractures are due to desiccation and direct heat­
induced effects where they are exposed. The heat of the
320 I Heat-induced injury or death
fire may result in fractures or render them so brittle
that they are fractured during recovery of the body_
Artefact ual fractures are particu lar ly common in the
sku ll , w here sometimes they can be difficult to
disting uish from ante-mortem injury,
• Body s hortening because the limbs, particularl y the
extremities, may be partly burn ed away an d muscles
are fl exed,
• Distortion of the facial features by bl ackening, skin
co ntraction and tightening, heat-induced damage to
the facial features and sin gein g of the hair.
• Accumulation of soot and perhaps th e emergence of
some blood-tinged fluid in the facia l openings.
• Irregular deep splits involving the sk in of the arms,
thighs and abdomen du e to ex posure to heat which
lead s to heat contraction of the skin. (These may be
misinterpreted as tea rs, lacerat ions and cuts inflicted
during life.)
• Accumulation of blood beneath the skull (giving the
appearances of an extradural haematoma); this may
also be mistaken for an an te-mortem injury.
Distinguishing between bu rns inflicted during life a nd
burns inflicted on an already dead body can be difficult, if
not impossible, at autopsy, Furthermore, genuine ante­
morte m nature of burn s may be obscured by the continued
ex posure to heat of the body after death.
Fire deaths result from:
• smoke inhalation;
• burns;
• heat shock;
• any combination of these,
Smoke Inhalation
Over 50 per cent of a ll fire fatalities occur as the res ult of
the inhalation of soot and gases generated in the course of
the fire. The hot air a nd steam generated may burn the fa ce
and enter into the upper a ir passages where they cause fur­
ther damage. This heat-induced effect may cause reflex
cardiac death (vagal inhibition).
Fires generate a wide variety of noxious gases that,
when inhaled, may ca use confusion and disorientation ,
incapacitation or dea th , often in combination with other
factors. With the exception of carbon monox ide, these
gases rarely reach lethal levels by themse lves.
Inhalation of soot particles dam ages the airways
because they are superh ea ted and contain toxic age nts. In
building fires, evid ence of soot inhalation at autopsy is
very common (90 per cent or more of cases). Autopsy evi­
dence of soot in the airw ays below the level of the vocal
cords is proof that the victim was alive at the time of the
fire, Soot in the oesophagus and admi xed with stomach
contents impli es the swallo wing of soot and has a similar
significance.
Blood taken at autopsy is analysed for carbon monoxi de
content and the result expressed as a percentage saturation
of the haemoglobin. Levels of up to 10 per cent carboxy­
haemoglobin saturation may be fo und in the normal popu­
lation in towns and cities as a res ult of atmospheric
pollut ion. A carboxy haemoglobin saturation level of over
50 per cent is generally accepted as sufficient in itself to
acco unt for death. Perso ns with pre-existing natural disease
of the heart or lungs may succumb to levels as lo w as 30 per
cent. In building fires approximately one-half of all victims
have a carboxyhaemoglob in level that is sufficiently high to
account for death. Carboxyhaemoglobin levels of 20- 30 per
cent produce dizziness, headache, nausea and fatigu e; levels
of 30-40 per cent produce imp aired judgement, confusion
and later unconsciousness, In a fire situation, levels above
30 per cent may explain a failure to escape.
The presence of a n elevated carboxyhaemoglobin level
and/or soot in the airways beyo nd the vocal cords provid es
proof that the victim was a live, but not necessarily conscious,
at the time of the fire, Inh aled carbon monoxide and other
noxious gases may have an additive effect and accou nt for
deaths with carboxyhaemoglobin levels below 50 per cent.
With the exception of cyanide estimation, laboratory tests
for the presence of these noxious gases is rarely, if ever, per­
formed on the bodies of fire victims. Cya nide is commonly
produced in fires and results from the burning of materials
such as wool, silk, horse hair, poly urethane a nd polyaclylo­
nitrile. Cya nide is a powerful, rapidl y acti ve and generally
cytotoxic poison, Normal blood cyanide levels are less than
8 ~mol/L; non-fa tal tox ic effects begin at about 50 i_,mol/L
and the fatal threshold is above 100 ~mol/L.
Bodies recovered From fires present severa l problems
of investigation a nd a close integration of info rmation
obtained from the exam inatio n of the scene, the exam ina­
tion of the body and the histOiY of the deceased is paliicu­
lady important. Several investigators with different areas
of ex pertise a re typically involved . In fire-related deaths a
fire investigator (usually an offlcer in the loca l fire brigade)
is in attendance and can provide valuable information on
the origin, development and nature of t he fire.
A full investigation of the circumstances preceding the
death requ ires the identification of the victim's past med­
ical a nd social history.
If the identity of a charred body is suspect, then identity
wi ll have to be proved scientifically. Wh en specific identi­
fication is not possible, circumstantia l evidence of identity
is usua ll y ava il a ble_
The majority of fire-rel ated deaths are accidental and
there is typically abundant coll ate ra l evidence from police
and fire brigade investigations to exclude suicide or homi­
cid e. The yo ung and the elderly are the common victims of
accid en ta l fires. The deaths are usua lly the result of care­
lessness: allowing clothing to bru sh aga inst fires; playing
with matches or other ligh ted obj ects su ch as cigarettes;
mainta ining faulty electrical and hea tin g appliances, as
well as being unable to effecti vely co mba t or escape a fire.
320 I Heat-Induced injury or death
ftre may result in fractures or render them so brittle
that they are fractured during recovely of the body.
Artefactual fractures a re palticularly common in the
skull, where sometimes they can be difficult to
disti nguish from ante-mortem injury.
• Body shortening because the limbs, palticularly the
extremities, may be paltly burned away and muscles
are flexed.
• Distortion of the facial features by blackening, skin
contraction and tightening, heat-induced damage to
the facial features and singeing of the hair.
• Accumu lation of soot and perhaps the emergence of
some blood-tinged fluid in the facial openings.
• Irregular deep splits involving the skin of the arms,
thighs and abdomen due to exposure to heat which
leads to heat contraction of the skin. (These may be
misinterpreted as tea rs, lacerations and cuts inflicted
during life.)
• Accumulation of blood beneath the sku ll (giving the
appearances of an extradural ha emato ma) ; this may
also be mistaken for an ante-mortem injury.
Distinguishing between burn s inflicted during life and
burns inflicted on an already dead body can be difficult, if
not impossible, at autopsy. Furthermore, genuine ante ­
mortem nature of burns may be obscured by the continued
exposure to heat of the body after death .
Fire deaths result from:
• smoke inhalation;
• burns;
• heat shock;
• any combination of these.
Smoke Inhalation
Over 50 per cent of all fire fatalities occur as the result of
the inhal ation of soot and gases gene rated in the course of
the ftre. The hot air and steam generated may burn the face
and enter into the upper air passages where they cause fur­
ther damage. This heat-induced effect may cause reflex
cardiac death (vagal inhibition).
Fires generate a wide variety of noxious gases that,
when inhaled, may cause confusion and disorientation,
incapaci tation or death, often in combination with other
factors. With the exception of carbon monoxide, these
gases rarely reach lethal leve ls by themselves .
Inhal ation of soot particles damages the airways
because they are superheated and contain toxic agents. In
building fires, evidence of soot inhalation at autopsy is
very common (90 per cent or more of cases). Autopsy evi­
dence of soot in the airways below the leve l of the vocal
cords is proof that the vict im was alive at the time of the
fire . Soot in the oesophagus an d admixed with stomach
contents implies the swallowing of soot and has a similar
s ig n ifi cance.
- ---
Blood taken at autopsy is analysed for carbon monoxide
content and the result expressed as a percentage saturation
of the haemoglobin. Levels of up to 10 per cent carboxy­
haemoglobin saturation may be found in the normal popu­
lation in towns and cities as a result of atmospheric
pollution. A carboxyhaemoglobin saturation level of over
50 per cent is generally accepted as sufficient in itself to
account for death. Persons with pre-existing natural disease
of the heart or lungs may succumb to levels as low as 30 per
cent. In building fires approximately one-half of all victims
have a carboxyhaemoglobin level that is sufficiently high to
account for death. Carboxyhaemoglobin levels of 20-30 per
cent produce dizziness, headache, nausea and fatigue; levels
of 30-40 per cent produce impaired judgement, confusion
and later unconsciousness. In a fire situation, levels above
30 per cent may exp lain a failure to escape.
The presence of an elevated carboxyhaemoglobin level
and/or soot in the airways beyond the vocal cords provides
proof that the victim was alive, but not necessalily conscious,
at the time of the fire. Inhaled ca rbon monoxide and other
noxious gases may have an add itive effect and account for
deaths with carboxyhaemoglobin levels below 50 per cent.
With the exception of cyanide estimation, laboratory tests
for the presence of these noxious gases is rarely, if ever, per­
formed on the bodies of fire victims. Cyanide is commonly
produced in fires and results from the burning of mate rials
such as wool, silk, horse hair, polyurethane and polyacryIo­
nitlile. Cyanide is a powerful, rapidly active and genera lly
cytotoxic poison. Normal blood cyan ide levels are less than
Bjtmol/L; non-fatal toxic effects begin at about 50f-Lmol/L
and the fatal threshold is above 100 f-Lmol/L.
Bodies recovered from fires present several problems
of invest igation and a close integration of informatio n
obtained from the examination of the scene, the examina ­
tion of the body and the history of the deceased is particu­
larly important. Several investigators with different areas
of expertise are typicaJly involved. In fire-related deaths a
ftre investigator (usually an officer in the local fire brigade)
is in attend ance and can provide valuable information on
the origin, development and nature of the fire.
A fu ll investigation of the circumstances preceding the
death requires the identiftcation of the victim 's past med­
ical and social history.
If the identity of a charred body is suspect, then identity
will have to be proved scie ntiftcally. When specific identi­
fication is not possible, circumstantial evidence of identity
is usually available.
The majority of fire-related deaths are acc idental and
there is typically abundant collateral evidence from police
and fire brigade investigations to exclude suicide or homi­
cide. The young and the elderly are the commo n victims of
accid ental fires. The deaths are usually the result of care­
lessness: allowing clothing to brush against fires; playing
with matches or other lighted objects such as cigarettes;
maintaining faulty electrical and heating appliances, as
we ll as being unable to effectively combat or escape a fire.
~
322 I Heat-Induced injury or death

Evidence of soot inhalation is very common (approxi­
mately 90 per cent of cases) in building fires but it is
uncommon to find any soot deposition beyond the primary
bronchioles. Deposition of soot on the tongue, in the nares,
the oropharynx or nasopharynx cannot be taken to imply
life during the fire. Deposition of soot below the level of the
larynx indicates that the victim was al ive at the start of the
fire. This soot that coats the mucosa of the tracheo­
bronchial tree can be identified through a window cut in
the trachea prior to removal of the neck and thoracic struc­
tures and is most easily seen if the mucus is spread on a
white sheet of paper. The presence of soot in the oesopha­
gus and stomach implies that it has been swallowed and
also indicates life at the start of the fire. On microscopy, the
soot particles within the tracheobronchial tree lie loosely
on the mucosal surface or embedded in the mucus. The par­
ticles are not incorporated into the tissues and are readily
distinguished from anthracitic pigment.
Smoke poisoning describes the effects of the various
noxious gases other than carbon monoxide that are pro­
duced by the thermal degradation of both natural and
man-made materials (Table 15.1). There are hundreds of
such products of combustion. Whereas modern synthetic
materials, for example PVC, may have increased this prob­
lem, natural materials, for example wood, wool, and silk,
produce similar gases and the pathological effects of these
noxious gases are difficult to separate from direct particu­
late injury. With the exception of carbon monoxide and
hydrogen cyanide, post-mortem analyses for toxic gases
are rarely, if ever, performed. Filter masks do not protect
against the inhalation of fumes; for this purpose the mask
must have a self-contained air supply.
These noxious gases rarely reach lethal levels by them­
selves, but they may cause incapacitation or death in com­
bination with other factors. Some gases have a synergistic
interaction, for exam pie carbon monoxide and hydrogen
cyanide, so that non-fatal levels of each in combination
may cause death. The concentration of the noxious gas is
of importance. Hydrogen cyanide produced from wool,
wood or polyurethane is rapidly fatal at 3000 ppm. Nitro­
gen dioxide from acrylonitrile is rapidly fatal at 2000 ppm.
Hydrogen chloride from PVC is rapidly fatal at 2000 ppm.
Hydrogen sulphide from rubber or wool has toxicity that
is similar to hydrogen cyanide and is rapidly fatal at
1000 ppm. Acrolein is an aldehyde product of combustion
of wood and paper and produces pulmonary oedema after
a few seconds exposure at 10 ppm. 14
Carbon monoxide is a colourless, odourless gas that is
formed by the incomplete combustion of carbon com­
pounds and is normally found in the atmosphere in levels
well below 1 ppm. Motor vehicle engines are responsible
for about 70 per cent of carbon monoxide liberated into the
atmosphere.
Carbon monoxide poisoning is a distinct cause of inhala­
tion injury, producing its effects by tissue hypoxia. The
affinity of haemoglobin for carbon monoxide is 200-300
times greater than for oxygen so that carboxyhaemoglobin
concentration is great even when the carbon monoxide con­
centration is less than 5 per cent in the inhaled gas. As well
as diminishing the oxygen-carrying capacity of the blood,
carboxyhaemoglobin (COHb) also alters the dissociation
characteristics of the remaining oxyhaemoglobin, making
less oxygen available to the tissues. The toxicity of carbon
monoxide depends upon: (1) the rate of inhalation of the gas
(i.e. concentration of gas in the inspired air and the duration
of exposure); (2) physical activity, which influences oxygen
requirements; and (3) individual variations in susceptibility.
Carboxyhaemoglobin is estimated by spectroscopic
methods making use of the fact that oxyhaemoglobin and
carboxyhaemoglobin have different absorption spectra.
The result is expressed as the percentage saturation and
is the ratio of carboxyhaemoglobin to total haemoglobin
X 100. The carbon monoxide oximeter is an automated
instrument that uses a spectrophotometric method. An
alternative gas chromatographic method is both specific
and very sensitive. Carbon monoxide liberated from a
known volume of blood is estimated on a gas partitioner
and the haemoglobin content estimated spectroscopically.

Table 15.1 Sources of toxic chemicals produced during conflagrations

Gas Source

Carbon monoxide, carbon dioxide All combustibles containing carbon

Nitrogen dioxide Cellulose, polyurethanes, acrylonitrile
Hydrogen chloride Chlorinated polymers, e.g. polyvinylchloride
Hydrogen cyanide Wool, silk, nylons, polyurethanes, N-containing plastics
Aldehydes Wool, cotton, paper, plasters, phenol-formaldehyde,
wood, nylon, polyester resin
Benzene Petroleum, plastics, polystyrene
Ammonia Melamine, nylon, urea-formaldehyde
Sulphur dioxide Rubber, thiokols
Phenol Phenol-forma Idehyde
Acrolein Wood, paper

Table 15.2 The effects on humons of different concentrotions of
carboxyhoemog/obin saturation
Percentage Effect
0-10 No immediate ill effects; fo und in cigarette
smokers or from industrial /motor ve hicle pollution
10-20 May produce dizziness and shortness of breath
on exertion
20-30 Dizziness, headache, nausea and fatigue
30-40 Impaired judgement, unconsciousness may occur
40-60 Unconsciousness and death likely
The effects of carboxyhaemoglobin at different percent­
age saturation levels are set out in Table 15.2.
INTERPRETATION OF RESULTS
Carbon monoxide at environmental levels of 1000 ppm can
kill in about half an hour, and at 5000 ppm is rapidly fatal.
About 85 per cent of building fire victims show evidence
of carbon monoxide inhalation and approximately 50 per
cent of victims have a carboxyhaemoglobin .level suffi­
ciently high to account for death. The normally accepted
fatal carboxyhaemoglobin level is 50 per cent.
Ethanol, a central nervous system depressant, might be
expected to have an additive or synergistic effect with carbon
monoxide but in practice this has not been demonstrated. An
elevated carboxyhaemoglobin level is usually but not univer­
sally (approximately 95 per cent) associated with evidence
of soot in the respiratory tract. The carbon monoxide and
cyanide produced in fires have an additive effect.
A carboxyhaemoglobin level below 10 per cent is not
proof that a v ictim was dead prior to the commencement of
the fire. Possible explanations may be: (I) little or no
carbon monoxide production in the fire due to abundant
oxygen allowing complete combustion, for example a
forest fire; (2) rapid death in a flash fire; and (3) death
following partially successful resuscitation.
Carbon monoxide may be produced after death. Blood­
stained fluid from the tboracic cavity in decomposed bod­
ies may have COHb levels as high as 80 per cent resulting
from the post mortem production of carbon monoxide. 15
Ethylene chloride, found in paint strippers and hairsprays,
is metabolized in vivo to carbon monoxide and may pro­
duce carboxyhaemoglobin levels of up to 40 per cent.
Hydrogen cyanide is a product of the thermal degrada­
ti on of many materials which contain nitrogen, for example
wool, si lk, horse hair, polyurethane and polyacrylonitrile.
Hydrogen cyanide is a potent toxin with a very rapid action.
Ii has a half-life in blood in the order of less than I hour,
bei ng metabolized to thiocyanate. Hydrogen cyanide in
bl ood is contained principally in the red blood cells and dis­
appears rapidly from plasma with a half-life of only 15 min­
:.Ites. Artefactual formation and metabolism of cyanide can
The pathologist's role I 323
occur in post-mortem samples of blood and tissue. Artefac­
tual formation of cyanide in post-mortem blood is more
likely in refrigerated samples than at room temperature. 16
Burns
The extent of damage caused by external heat depends on a
number of factors. Among these is the applied temperature ­
mammalian tissues only survive within the narrow tempera­
hire range of 20-44'C; the duration of the interval over
which the heat has been applied to the skin or other tissues
(e.g. mucosal; and the ability of the sUlface to conduct heat
away. Thus a burn may appear after 5 hours of exposure at
44'C but it only take 3 seconds to cause heat-induced dam­
age at a temperature of 60'C; a child's skin is about 10 times
softer than in an adult and it burns 40 times faster. The mor­
bidity and mortality associated with burns is related to the
depth of bum injury and the extent of injury, i.e. the size of
the burn relative to that of total body surface (TBS) area. The
obvious effect of dry heat is direct physical damage in the
shape of a burn and tbe mechanism of injury includes both
conduction and radiation. Further prolonged exposure of the
area to the heat can result in charring and carbonization of
the surface, singeing of the surface hairs and eventually
compete destruction of the tissues (cremation).
Jackson in 1947 17 described three zones in a burn: the
pOint of maximum damage w ith coagulation of proteins
locally is the zone of coagulation. This is surrounded by an
area of decreased tissue perfusion where the tissue is still
potentially sa lvageable, referred to as the zone of stasis.
In the outermost pari of t his area, tissue perfusion is
increased; there is a n increase in vascularity in this area
and it is the zone of hyperaem ia .
The older Wilson's tra ditional classification of depth of
burn injury is in degrees,18 i.e. first degree, second degree
and t hird degree. A first-degree bum involves on ly the epi­
dermis; it is characterized by erythema and mild pain, the
la tter resolving in 48-72 hours. Healing is usually unevent­
ful and completed in 5-10 days with no residual scarring.
Sunburn is the most common first-degree burn. Second­
degree burns involve the epidermis and a valiable portion of
the dermis. Superficial second -degree bums implicate only
the upper third of the dermis and are charactelized by blister
formation; they are extremely painful but heal in 7-14 days
with minimal scaning. A deep second-degree burn extends
beyond the upper third of the dermis, but not beyond the
dermis itself. These deeper burns are less painful than super­
ficial second degree burns; heal ing is extremely slow, some­
times requiring months and usually leading to dense
scarring (if the wound is allowed to heal primarily, rather
than skin grafted). The fluid loss and metabolic effects of
deep second-degree burns are essentially the same as those
seen with third degree burns. A/third-degree burn, or full
thickness bum, implicates the entire epidermis and dermis.
Primary re-epithelialization will not occur and the wound
324 I Heat-induced injury or death
will require skin grafting. Heat coagu lation of dermal blood
vessels leaves the tissue avascular with a characteristic waxy
white colour. Prolon ged contact of subcutaneous fat with a
flame source produces a leathelY brown, or black, chan-ed
appearance. There is charac te ristic lack of pain, due to heat
destmction of all nerve endings.
Burns are nowadays more frequently classified as either
partial thick ness or full thic kness . Partial-thickness wounds
contain viable epithelial elements capable of sponta ­
neously re-epithelializing the wo und (first-degree, superfi­
cia l and deep second-degree burns). Full-thickness burns
have no viable epithelial elements and always require cuta­
neous autografting. Partial thickness burns are usually
caused by a brief exposure to heat or contac t with hot liq­
uids and appear pink to mottled red, wet, covered with
ves icl es and bullae, and are painful. Full-thickness burns
result from contact with flame, electricity or chemicals.
They are often dry and charred, may be translucent, a nd
may have thrombosed superficial veins. They are insen sate.
Burns wi th chemicals can arise both from direct contact
and from fumes (e.g. chlorine), either acting directly on the
skin or mucosal surfa ces, or by inhalation. The degree of
tissue damage caused by them is dependent on the type of
chemical (alkalis are more caustic than acids in general
terms), its concentration, its quantity, the duration of contact
a nd the extent of penetration; if chemicals are lipid soluble
or if proteolysis occurs then this results in a greater depth of
penetration and at a quicker rate due to the liquefactive
necrosis that they induce, th us exposing deeper tissues to the
effects of the chemical. Chemicals that induce burns can be
categorized as strong acids, strong alka lis and hydrocarbo ns,
wi th hydrocarbons often having penetrating capacity.
Heat damage from hot fluids is usually referred to as a
scald. Water is the compon ent fluid involved, usually in the
kitchen or bathroom, but other fluid s such as oil, molten
metal , molten rub ber, other hot liquids and semisolid or
liquid foods (e.g. porridge, molten cheese) may be
involved; steam as a by-product of the heating process is
also very damaging due to its penetrative cap abilities.
Scald s do not usually result in charring, carbonization or
singeing of body hairs. Their effects are more like that of
first-degree burns from dry heat, with superficial redden­
ing, desquamation and loss of the epidermal layer, swelling
and blistering, usually with well-demarcated areas of dam ­
age, with unifo rmi ty all over the site of a rea of contact
with the fluid. Thus the severity of skin damage is more
uniform in sca lding and from the distribution of the scalds
it may be quite possib le to identify th e position of the
scalded victim in rel ation to the hot fluid; there may be
trickle or splash scalds, and horizontal fluid levels from
immersion in hot water.
The scald 's severity depends on the duration of exposure
of the skin to the hot fluid and its tem perature. Clothing
may prevent direct content of the fluid underlying the skin
but once it is soaked it may increase the duration of con­
tact of the hot fluid with the skin's surface. In many
in sta nces co ntact with the hot liquid is therefore momen­
tary and the large surface area of the skin on to which the
hot fluid has made contact a llows for rapid cooling, and
thus unless t he temp erature of the water is quite high, there
would not be a sufficiently lengthy period of exposure for
injury to result from such co ntact. These pheno mena are
quite impoltant in attempting to reconstmct how a child
has come by his scalds, palticularly in relation to accepting
or otherwise the his tory given by the carers when non­
accidenta l injury (NAI) is suspected.
In this respect the temperature of the hot water system
in households has to be regulated with care to ensure th at it
does not exceed celtain levels l9 and legislation and guide­
line in terms of plumbing of buildings are in existence.2o
Other instances of hea t-induced injury involve fire­
works, palticularly around the time of bonfire night in
November in the UK a nd New Year's Day festivi ties in other
countries 2 1,22 Some deaths from overheating have a lso
been described in children, particularly infan ts, from the
use of electric blankets.
Pathological Changes
Beneath the dead burned tiss ue, there is usually a zon e of
ischaemia of marginally viable tissue th at is readily con­
verted to non-viab le tissue (eschar) by any furth er insult,
such as hypoxia, decreased blood flow or infection. By this
process, a deep seco nd-degree burn frequ ently converts
over time to a third-degree burn. Prevention of wound
conversion of this type is of major importance in the res us­
c itation period and thus attention to oxygenation and
decreasing the risk of infection are essential elements of
the early management of burns. 23 ,24
A determina tio n of the burn surface area in the living is
obtained using the 'mle of nines ', each arm constituting
9 per cent, each leg 18 per cent, the anterior trunk 18 per
cent, the posterior trun k 18 per cent and the head 9 per
cent of TBS area. In assessing irregular areas of burn, it is
useful to remember that the surface area of th e person's
palm is roughly equivalent to 1 per cent of the TBS. In chil­
dren under 15 years of age the rel ative body surface area of
the head, upper leg and lower leg differs from an adult; this
ma y lead to inaccuracy in the estimation of burn injury;
use of the Lund Et Browder Burn Diag ram provides age­
adjusted data accurate for both adults and children.
It should be remembered that the extent of burn is often
over-estimated, and the depth of burn under-estimated
wh en the pati ent is assessed clinically.
In addltion to burn depth and extent of injury, other
factors determining morbidity a nd morta lity a re the loca ­
tion of the injury, the age of the victim and the presence of
injuries or natural disease. Age is a major factor in survival
for children under 2 years and adults over 60 years.
The dermis acts as a barri er prev,enting loss of body
fluid by evaporation an d the loss of excess body heat.

Consequently, loss of water through burn ed skin and loss
of heat play a major role in the pathophysiological
changes. The skin is also the primary protective barrier
against invasive infection so that wound infection is a
major cause of mortality and morbidity in the late post­
burn period. Within the superficial dermis are the nerve
endings that mediate pain. Consequently, partial-thickness
injuries tha t expose these nerves will be extremely painful,
whereas full-thickness burns that destroy the nerves are
usually anaesthetic.
The loca l and system ic responses to burns fo llow a time
course continuum but, for convenience, can be divided in to
three periods: (1) up to 48 hours; (2) 2-6 days; and (3) 7 days
to wound closure.
In the first 48 hours after severe burns, hypovolaemic
shock an d shock-induced organ failure (primarily renal
failur e) are the major threats to life. Hypovolaemia can also
lead to wound convers ion. Al though the exact pathophys­
iolo gy of the post-burn vasc ular changes and volume shifts
remains to be determined, two processes are involved: an
increase in microvascular fluid flux into the interstitium ­
both local to the burn and genera li zed (remote from the
burn wound) - and a generalized impairment in cell mem­
brane function, resu lting in cell swelling.
With modern therapy, adequate initial volume restora­
tion is achieved in more than 95 per cent of burns cases.
Correction of hypovolaem ia by the intravenous infusion of
fluid s is comp licated by ge neralized burn oedema forma­
tion, wh ich may resu lt in (1) further ischaemic insu lt to
already damaged cells; (2) chest wall oedema with resultant
increase in the work of breathing; and (3) upper airway
oedema. The last two complications may lead to rapidly
fatal respiratory failure.
Burned skin loses its elasticity. It is less compliant and
unable to stretch to accommodate an increase in interstitial
oedema. If burns are circumferen tia l, particularly a round
the distal extremities, a to urniquet effect is produced by
t hem. Initially this impedes venous return , resulti ng in an
increase in capillary pressure and further oedema. Impair­
ment of arterial blood flow may follow, with the potential
for distal tissue ischaemic necrosis. Tissue pressure can be
decreased by making an incision through the burned tiss ue,
i.e. an escharotomy. A chest wall escharotomy may be
required to ease respiratory difficulties; incisions are made
in the bilateral a nterior axillary lin es from the clavicles to
the costal margins; a transverse escharotomy at the level of
the costal margins connect ing the two vertical escharo­
tomies is helpful if the abdomi nal wall is burned to a sig­
nificant degree. Escharotomy can be a bedside procedure
and does not require anaesthesia because full-thickness
burns are insensa te. If th e deeper viable tissues are not
incised then bleeding should be minim al.
Third-degree or full-thickness burns often require exci­
sion prior to cutaneo us autografting. The two commonly
employed techniques of excisi on are tangential excisio n
(sequential shaving of no n-viable tissue until a uniformly
The pathologist's role I 325
viable wound bed is obtained) and excision to fascia . The
disadvantage of tangential excision is primarily that of
mass ive blood loss.
Burns also induce a system ic response to them due to
the presence of cytokines and other inflammatory media­
tors if the area of the burns exceeds the 30 per cent surface
area. These substances cause systemic effects:
1. increased capillary perm eability that leads to loss of
fluid and from proteins within the blood vessels into
the interstitial tissues.
2. peripheral and sp lanchni c vasoco nstriction.
3. decreased muscular contractility (probably due to the
effect of tumour necrosis factor alpha).
4. systemic h ypotension due to primary and seco ndary
fluid losses.
5. fall in systemic blood pressure - hypotension.
6. hypoperfusion of vital organs.
7. bronchoconstriction.
8. adult respiratory distress syndrome.
9. the basal metabolic rate trebl es.
10. No n -specific downregulation of the immun e system
affect ing both antibody- and lymphocyte-mediated
pathways.
NON-ACCIDENTAL BURNS AND SCALDS
Although a significant number of heat-i nduced injuries are
truly accidental - although often preventable - they result
from temporary lapses in the due care and attention that chil­
dren require in the home an d elsewhere. Some, however, are
due to neglect on the part of the carers, who show inadequate
and negligent parenting with failure to protect the child.
Others heat-induced injuries are deliberately inflicted. 2s - 27
Sometimes it is eventually suggested by the perpetrato r that
these are punitive in nature, allegedly punishing the child for
'bad' behaviour and failure to comply with parental instruc­
tions, for example in relation to 'potty training' or playing
with matches. They may also be sadistic, although less fre­
quently, in their production, with the perpetrator obtaining
pleasure by the infliction of pain and fear on the child. In
about 10 per cent of physically abused children bums are a
component of the injuries inflicted; about 5 per cent of sexu­
ally abused chi ldren also show bums. 28 ­ 31
Injury induced by liquids is usually referred to as sca lds.
These often blister, with peeling of the epidermis when the
blisters burst. The damaged area often appears soggy and
blanched. Their distribution may follow the contour of the
clothes given that the soaking of clothing with hot liquid will
increase the duration of the contact between the liquid and
the skin and may thus follow the contours of clothing. The
liquid will produce splash patterns - pouring and drip pat­
terns. Domestic hot water, boiling kettles and saucepans, pip­
in g hot cups of tea and coffee, and baths are frequent causes.
Scalds may also be found internally on the lips, gums,
cheeks , tongue an d pharynx as a ctlllsequence of swallow­
ing hot liquids. 32 Large areas of mucosa will peel off, and if
326 I Heat-induced injury or death
oedema supervenes as a reactive change there may be an
obstruction to upper air passages and to air exchange.
Contact with metallic, hot, dry surfaces will produce dry
or contact burns, such as with contact with central heating
radiators, an oven, electric heate r, clothes' irons, curling
tongs, fire surrounds, light bulbs, heating grills; serious
burns can result even though the contact with such objects
is usuall y of a very short duration. These burns may
demonstrate a branding type of appearance, with the burn
appearing dry and well demarcated, reproducing the pattern
of the object that has caused t he burn.
Lit cigarettes may leave very characteristic patterned
circular or oval marks on the skin and although these may
be accidentally induced , particularly in a crowded house­
hold, they are also often deliberately inflicted on the child.
In the latter instances areas of the body that are less easily
visible to a medical examiner or observer will be used by
the perpetrator: thus such Cigarette burns may be found on
the lower back, the nape of t he neck, the buttocks and the
legs. They are often multiple, and indeed may also be of a
different duration, and if they present in varying healing
phases then they are indicative of more than one episode of
such trauma. A pattern that is seen occasionally is that the
outline of the main bruise is oval in its configuration with
a roughly triangular ta il of further, usu ally more superfi­
ci al, burning radially confluent with the main burn but
located at an angle to it. This is due to move ment by the
cigarette across the skin or by the child attempting to move
away from the point of impact with the lit Cigarette end.
This appearance can thus be found in both accidental and
delib erately inflicted cigarette burns.
Accidental burns are found at sites where the child has
handled the lit cigarette or fallen on/brushed against it;
thus the back of the hands, the feet and the back are not the
usu al sites of an accidental contact.
Radiant burns may occur after prolonged contact with
the sun or a rather hot tIre. These affect a large anatomical
area of the body and are often ch aracterized by marked
erythema. All of the surface that is exposed will show heat­
indu ced injury. Clothing may have a protective effect and
thus patterns may be observed in such burns.
Chemical agents, not least acids and alkalis, cleaning
agents used in the house and also electricity sources will
also induce burns. The amount of heat generated and thus
the a mount of tissue da mage is equal to 0.24 x (voltage)2 x
resistance. 33 Thus voltage is the main determinant of tissue
injury, with electric burns thus being divided into low volt­
age burns (as with t he domestic current) and high-voltage
burns. Low-voltage burns are often deep-contact burns at
the entry and exit points of the current from the body.
When the latter involve voltages of over 1000 volts they
are subdivided further into true 'h igh-tension' injuries
caused by a high voltage current going through the body
and flash injuries caused by tangential exposure to high­
voltage current with no current actually flowing t hrough
the body. These may show extensive tissue losses and often
even involve the loss of limb. There may also be muscle
da mage with rhabdomyolysis, which, in turn, may lead to
acute renal failure. Contact with voltages of over 70000
volts is invariably fatal. In flash electric burns there is
exposure to radiant heat of the surface of the body, often
the face or on the arms, but clothing may also be set on fire
with further burns ensuing.
The depth of a burn is related to:
• the temperature of the heat source.
• the duration of contact between the skin or mucosa
and the warm object.
The usual temperature of hot water in the home is about
60°C and contact with water at this temperature results in
a full-thickness burn in child at about 10 seconds of con­
tact. Higher temperatures will cause burns much quicker
and at 70°C only 1 second of contact is sufficient for a full­
thickness burn.
About 3-10 per cent of burns in British children are due
to NAI.34 In American studies, 9-25 per cent of children
who are admitted to burns uni ts have been assaulted.
Younger children in the age range 5 months to 6 years are
often affected, with a peak noted in older toddlers around
the age of] years; about one-third of these children will be
repeatedly abused. This is in contrast with other forms of
abuse when younger children usually feature more promi­
nently.
In assessing whether a heat-induced injury is accidental
or otherwise the presentation of the child and the history
given by the carer at medical presentation is impoJta nt, as
is the interaction of the carers with medical and nursing
staff. If the injury had been inflicted non-accidentally then
the appearance of the injury and the history given by the
carer may not tally: the carer may even suggest that the
injury was not induced by heat or the explanation given
for the contact with the warm solid or liquid is incongru­
ous, for example claiming t hat a two-year-old has opened
a tight, large, hot water tap in a bath. Siblings may be
blamed and the episode is claimed to be unwitnessed; when
the child is asked, there is a different version of events
given. Associated concomitant injuries and previous heat­
induced damage may fe ature in the past medical history in
such instances.
If t he child has been deliberately immersed into hot
water then a clearly delineated tide mark may be seen if the
lo we r limbs are involved . Similarly, if the buttocks are
immersed then a 'doughnu t' demarcation will be observed
on the buttocks. The deliberate infliction of scalds on a
child may be the result of chastisement for failure by the
child to comply with potty training and in bed wetting. In
many instances, there will also be splash marks elsewhere.
Children thus non-accidentally assaulted should be
exa mined for other signs of abuse and neglect. The differ­
entiation between accidental and non- acc idental burns is
not easy and, unless suspected in all >circumstances, non­
accidental injUly will be missed?5
 References I 327

BURNS CAUSED BY ELECTRICITY Electricity can cause disruption of the body's normal
electrical activities, with neurological 39 dysfunction being
An electric current is a flow of electrons along the path of
present in some form, even if only temporalY, in virtually
least resistance toward a natural ground with any sub­
all patients. 4o Transient nerve injuries resulting in tempo­
stance or object along this path decreasing the rate of flow
rary numbness and tingling are most common. Mass depo­
(resistors) or increasing the rate of flow (conductors). The
larization of the brain may lead to a loss of consciousness
amnesia and coma. Spinal cord involvement may result i~
skin is a natural resistor to electrical flow: dry skin has a
resistance of 40 000-100 000 ohms, wet skin approximately
transverse myelitis. Transverse myelitis may have delayed
1000 ohms and calloused skin 2000000 ohms. The thin
onset and is associated with poor prognosis for recovery.
skin and high water content of the skin of children has a
Electrical injuries also may affect the heart 41 and about
resistance much lower than that of an adult. J6 The internal
25 per cent of patients with electrical injuries have cardiac
resistance of the body is estimated to be between 500 and
dysrhythmia. Many of these are benign and transient (e.g.
1000 ohms, with bones, tendons, and fat providing most
sinus tachycardia , premature atrial ventricular contractions
resistance. Nerves, blood vessels, mucous membranes and
and conduction disorders). Sudden death from an AC elec­
muscle are the best conductors.
trical injury often results in ventricular fibrillation, although
With burns resulting from an electric current, the cross­
asystole and other dysrhythmias are common. Ventricular
sectional area is inversely proportional to tissue damage.
fibrillation is three times more likely to occur if the flow of
The pathway of the current plays an important role in
current is arm-to-arm. Tnle myocardial infarction is rare
determining injury, with a vertical pathway being more
unless the patient has pre-existing cardiac disease.
dangerous than a horizontal (hand-to-hand) pathway.J7
Another mechanism of injury is related directly to the
Standard household current in the UK is 230-240 volts
amount of heat generated by the flow of electrical current
(alternating current lAC]) at a frequency of 50 hertz; in the
through body tissue. At higher voltages, higher tempera­
USA and Canada it is 110 volts, with a frequency of 60 hertz.
hIres are achieved thus resulting in greater direct thermal
Skeletal muscle is stimulated into spasm and tetany by cur­
injury. High-tension voltages cause devastating injuries
rents with frequencies of 40- 110 hertz. Most low- and high­
from huge amounts of internal thermal damage.
tension electrical current is AC. Alternating CUITent produces
Vascular injury occurs as a result of vascular spasm, and
tetany and the 'locked-on' phenomenon as the fle xo r muscle
the heat generated can also cause coagulation and vascular
groups are usually stronger and predominate. As a result, an
occlusion. Damage to the vascular wall may produce delayed
individual's grasp is uncontrollably locked on to an object in
thrombosis and bleeding. Compartment syndrome may
which an electric current is passing, which can increase the
develop as a result of acute ischaemic insult to the muscula­
length of time the current passes through the body and may
ture. Renal injuries may occur as a result of rhabdomyolysis.
result in greater injury. In contrast, direct current (DC) tends
Rhabdomyolysis causes myoglobinuria from massive release
to produce a single large muscular contraction t hat often
of myoglobin, which, on crystallization in the kidney tubules,
throws the patient away from the source. However, at high
may cause acute renal failure.
voltages, both AC and DC produce similar effects. Contact
with high-voltage CUITents results in injuries that are associ­
ated with arc burns and flash burns. Arc temperahlres may
REFERENCES
reach up to 5000· C and are usually responsible for the severe
thermal injuries. Runya n SA, Casteel C (eds) The State of Home Safety in
Lightning involves a single massive current impulse that Americr: Facts about Unintentional Injuries in the Home,
is roughly equivalent to a DC blast of 2000-2 billion volts 2nd edn. Washington DC: Home Safety Council, 2004.
of extremely short duration (0.1-1 ms).J8 The peak tempera­ 2 Centers for Disease Co ntrol and Prevention. Web-based Injury
Statlstlcs Query an d Reporting System (WISQARS) [online).
ture in the lightning strike channel is 30000·C. However,
2005. National Center for Injury Prevention and Control,
the short duration usually precludes serious direct thermal Centers for Disease Control and Prevention.
injllly. Four modes of lightning injury have been described: www.cdc.gov/n cip c/w isqars
direct strike, side flash (discharge from an object near the 3 Internation al Association for the Study of Insurance
victim)' stride potential (enters one foot and exits via the Economics. World Fire Statistics: Information Bulletin of the
other), and flashover (energy passing outside the body with World Fire Statistics. Geneva: The Geneva Association, 2003.
4 Ma lon ee S, Istee G, Rosenberg M et al. Surveillance and
vaporization of surface water and blast effect to clothing). prevention of reside ntial - fire injuries. N Engl J iVIed 1996;
Lightning may cause full cardiac arrest by inducing either 335:27-32.
asystole or central apnoea. Massive depolarization of the 5 Sorensen B. Prevention of burns a nd scalds in a developed
heart leads to asystole. However, the heart usually restarts counBy. J Trauma 1976; 16:249-58.
6 Linares AZ , Linares HA. Burn prevention programmes for
spontaneously in normal sinus rhythm. Massive depolariza­
children: are they effective? Burns 1979; 6:26-9.
tion of the brain is believed to stun the respiratory centre, 7 MacK ay A, Rothman K. The incidence and severity of burn
causing a much longer duration of central apnoea. If artifi­ injuries following Project Burn Prevention. Am J Publ Health
cial respiration is provided, many patients can survive. 1982; 72:248-5 2. >
328 I Heat-induced injury or death
8 Istre GR, McCoy MA, Osborn L et al. Deaths and injuri es from
hou se fires. N Engl } Nled 2001; 344: 1911-16.
9 McLoughlin E, Marchione M, Han ger L et al. Smoke detecto r
legislation: its effects on owner-occupier homes. Am } Publ
Health 1985; 75:858-62.
to Ahrens M. US Experience with Smoke Alarms and Other Fire
Alanns. Atlanta, GA: Quincy [MA) National fire Protection
Associatio n, 2004.
11 Mill er R, Reisinger K, Blatter M, Wu cher F. Pediatric
co unselling and subseq uent use of smoke desecraters.
Ani J Publ Health 1982; 72 :392-3.
12 Carlson A, Uden G, Karlsson ED. Burns injuries in sma ll
children, a population-based study ill Sweden . } Clin Nursing
2006; 15:129-34.
13 Anderson RA, Watso n AA, Harland WA. fire deaths in the
Glasgow a rea: general cons iderat ions and patho logy.
Med Sci Law 1981; 21: 175-83.
14 Napier DH. Haza rdou s materials and· the gases they produce.
Med Sci Law 1977; 17 :83-90.
15 Kojima T, Nishiyama Y, Yashiki M, Une I. Post mortem
formation of carbon monoxide. Forensic Sci Int 198 2;
19: 243-4.
16 Anderson RA, Ha rl a nd WA. Fire deat hs in the Glasgow area:
the role of hydrogen cyanide. Med Sci Law 1982 ; 22:35 -40.
17 Hettiaratchy S, Dz iew ulski P. ABC of burns: pathophysiology
a nd types of burns. BM} 200 4; 828:1427 -9.
18 Knight B. Burns and sca lds. In Kn ight B [ed.) Forensic
Pathology, 2nd ed n. London: Arno ld, 1996, pp. 305- J 7.
19 Weaver AlVI, Himel MHM, Edli ch RF. Immersion scald burns:
strate gies for preventio n. J Emerg lVled 1993; 11 :39 7-402.
20 Erdmann T, Felman K, Rivara F et a!. Tap water burn
prevention: the effect of legi slati on. Pediatrics 1991:
88:572-7 .
21 Sm ith GA , Knapp Jf, Barnett TM, Shield s BJ. The rockets· red
glare, the bomb s bursting in the air : fireworks- related injuries
to children. Pediatrics 1996; 98: 1-9.
22 D'Regenio P, Cafaro L, Santon3sras i F, Taggi F et al.
Capodanno Senza Danno: the e ffects of an intervention
program o n fireworks injuri es in Nap les. Am J Pub' Healtlr
1996; 86:84-6.
23 Walker AR. Emergency department management of house fire
burns and carbon monoxid e pOiso ning in chi ldren. Cur Opin
Pediatr 1996; 8:239 -42.
24 Wo lf SE, Rose JK, Desai MH et a!. Mortality determinants in
massive pediatric burns. An analysis of 103 chil.dren with
;'80% TBSA burns [;'700/0 full thickness). Ann Surg 1997;
225: 554-65, discussion 565-9.
25 Ayoub C, Pfeifer D. Burns as a manifestation of ch ild abuse
and neglec t. Alii J Dis Child 1979; 133:910- 14.
26 Sto ne NH , Rinldop L, Humphrey CR et a1. Chi ld abuse by
burnin g. Surg Clin North Am 1970; 50:141 9- 24.
27 Showers J, Ga rrison KM. Bu rn abuse; a four -yea r study.
J Traumn 1988; 28: 1581 - 3.
28 Hight DW. Bakalar HR , Lloyd J. Inflicted burns in children:
recognition and treatment. JAMA 1979; 242:517 - 20.
29 Keen JH, Lend rum J, Wolman B. Inflicted burns and scalds in
children. BM} 1975; 4:268-9.
30 Hobbs CJ. When are bum s not accid enta l? Arch Dis Child
1986; 61:357-61.
31 Lenoski IF, Hunter KA. Speciflc patte rn s of inflicted burns
injuri es . J Trauma 1977 ; 17:842.
32 Canady JW, Thomps on SA, Bardach J. Oral commissure burns
in chi ldren. Plost Recon struct Surg 1996; 9 7:738-44,
discu ss ion 745, 74 6-55.
33 Luce E. Electrical burns. Clin Plast Surg 2000; 27:133 - 43.
34 Andronicus M, Oates RK, Peat J et a1. No n-accid enta l burns in
children. Bums 1998; 24:552 - 8.
35 Benger JR, McCab e SE. Burns and scalds in pre-school
children atte ndin g accident and emergency: accid ent or
abuse? Emerg Med J 200 1; 18:172-4.
36 Ga rcia CT, Smit h GA, Cohen DM, Ferna ndez K. Electrica l
injuri es in a pediatric emergency depa ltment. A nn Emug IVIed
1995; 26:604-8.
37 Nguyen BH, Mac Kay M, Bailey B, Klassen TP. Epidem iology
of electrical- and lightning-related deaths and injuri es amo ng
Canad ian ch ildren and you th. Injurv Pre" 2004; 10:122 - 4.
38 Jain S, Bandi V. Electrica l a nd li g htning injuri es. Crit Care
Clin 1999; 15:319 - 31.
39 Lee RC. Injury by electrical forces: pathophysiology,
manifestations, and thera py. CUlT ?robl Slirg 1997 ;
34:677-764.
40 Rai J, Jeschke MG, Barro w RE , Herndon DN. Electrica l
injuries: a 30-year review. J Trauma 1999 ; 46:933 - 6.
41 Bailey B, Gaudreau! t p, Thivierge RL, Turgeon JP. Cardiac
mon itoring of children with hou seho ld electrical injuries.
Ann Emerg Med 1995; 25:612-17.

ASPHYXIAL DEATHS IN CHILDREN
Anthony Busuttil
Petechiae 329
Scene of death 330
Traumatic asphyxia in children 330
Entrapment asphyxia 330
Foreign body inhalation 330
Plastic bag asphyxia 331
Overlaying and wedging 331
Strangulation 331
The term 'asphyxia' comes from the Greek sphygmos
(meaning absence of a pulse) and is applied to pathological
unnatural states in which the body is deprived of oxygen
while there is a concomitant excess of carbon dioxide, i.e.
a hypoxaemia (hypoxia) and hypercarbia (hypercapnoea)
together. This may result in loss of consciousness with the
coma thus produced leading to death. Children die from
mechanical asphyxia in many situations and exhibit simi­
lar pathological changes that one would expect in an adult
dying in a similar situation.
In the international classification of diseases, accidents
caused by asphyxia are classified under E91O-E913:
• E910 = Accidental drowning and submersion.
• E911 = Inhalation and ingestion of food causing
obstruction of respiratory tract or suffocation.
• E912 = Inhalation and ingestion of other object
causing obstruction of respiratory tract or suffocation.
• E913 = Accidental mechanical suffocation.
PETECHIAE
This pathological feature has long been recognized in the
forensic context. As a finding of medicolegal import,
petechiae were first described by Ambrose Tardieu in
1855. 1 It was, at one stage (and perhaps by many still is),
I CHAPTER 16 I
Hanging by a ligature 332
Drowning and near drowning 332
Imposed airways obstruction 332
Abuse of inhalants (solvent abuse) 333
Reverse suspension 333
Chemical asphyxia 333
Prevention 333
References 334
considered to be pathognomonic and an almost sine qua
non of asphyxia.
These are haemorrhages arsing from tiny blood vessels,
probably mostly venules (capillary haemorrhages may be
too small to be visible naked eye being 1-2 mm lesions or
smaller), which rupture when the pressure within them is
increased and the endothelial lining is distended beyond its
innate elasticity. Contribution by hypoxaemia is likely in
that the vessel wall is also affected by a decrease in oxygen
tension, becoming leaky and pennitting the extravasation of
red blood cells. These lesions tend to occur at sites where the
connective suppot1ing tissue is loose, provided that the pres­
sure is locally raised in the specific area. 2 Thus an applica­
tion of a sphygmomanometer cuff at a level sufficient to
occlude, selectively, the venous supply for a period will
result in petechiae formation distal to the edge of the cuff.
Another cause of petechial haemorrhages is a blood­
clotting problem. A low platelet count from any cause, for
example idiopathic thrombocytopenic purpura, aplastic
anaemia, or leukaemia, may produce petechiae. Excessive
capillary fragility such as Henoch-Schonlein purpura may
have a similar effect (see Chapter 4).
Petechiae may also be found when there is toxic damage
to the endothelium, particularly in association with dissemi­
nated intravascular coagulopathy, as in coliform and
meningococcal septicaemia. In septicaemia, petechiae tend to
330 I Asphyxial deaths in children
be widespread. They are also found in situations when
microembolic phenomena occur, such as fat and air
embolism, amniotic fluid embolism and sub-acute bacterial
endocarditis. In asphyx ia, these haemorrhages tend to be dis­
tributed externally above the level of the obstruction and not
below the level, and are thus commonly found: on the palpe­
bral and bulbar conj unctivae ; on the face - mainly around
the eyes; behind the ears (pinnae); and on the mucosal sur­
faces of the lips and nose. In children it is not uncommon to
find these haemolThages in a single site only and not neces­
sarily the eye. 3,4 They may be present over the face a nd ante­
rior chest wall in still births and in early neonatal deaths of
babies who die as a result of acute hypoxic/ischaemic insult,
particularly retroplacental haemorrhage (see Fig. 10.10).
They also occur naturally in normal children, albeit in
sma ll numbers. In a total of 116 children under the age of
1 yea r, who were fully examined in child surveillance clin­
ics in Newcastle-upon-Tyne, Downes et al fo und that 27.6
per cent of children had one or more petechi ae, 8.6 per cent
two or more, and 2.6 per cent had more than two. s
Petechiae can also occur if the child has been experienc­
ing pronounced and recurrent Valsalva manoeuvres, as
with coughing (for example pertussis infection , bronchioli­
tis, vomiting [gastroenteritis], crying, straining, temper
tantrums). This may also explain why petechiae may be
discovered after active cardiopulmonary resuscitation. 6
They a lso occur interna ll y with any form of hypoxaemia
as the final common pathway for death. They are character­
istic of the sudden infant death syndrome, in which they are
found on the thymus, the epicardium and the pleurae. 7 - 9
SCENE OF DEATH
As with aU other deaths it is essential that the death scene
is thorou gh ly examined 10 in all cases when asphyxia
appears to have been the mode of death. In this respect, if
the baby has already been moved it maybe usefu l to seek a
reconstruction of the in cident of w here exactly the dead
child was found, by using dolls that can be manipulated by
the carers in line with their recollection of events. I! These
interactions with the family may be very emotionally
fraught and have to be carried out with great sensitivity by
police officers or others who have been specia lly trained . 12
TRAUMATIC ASPHYXIA IN CHILDREN
In this situation the thoracic cavity is transfixed and no
respiratolY movements are thus possible. There are classic
signs of congestio n above the obstruction, central cyanosis
and petechial haem orrha ge formation also above the
obstructio n, usually ending at about the level of the clavi­
cles. Children who are trapped under masonry and other
items, for example in explosions, in the course of crowd
stampedes, in the course of warfare and as a result of other
environmental catastrophes such as earthquakes and gas
explosions, will show features above the level of obstruc­
tion associated with t his condition.
Another accidental asphyxial cause of death in yo ung
children is being run over at low speed by reversing cars,
not infrequently driven by one of the other members of
their own family and sometimes in their own driveway. The
child's short stature precludes their visibility to the driver
with a consequent accidental knocking over of the child
who, having a low centre of gravity, will tend to fall below
the ve hicle rather than being lifted up and over the car.
Many children surv ive this type of insult, provided that
the duration of appl ication of the compressive force is not
prolonged, the velocity of imp act is low and the weight of
the object trapping them is not considerable. 13 - IS Very
often they present with features of cerebral anoxia and
convulsions. They frequently have visceral injuries and
extensive soft tissue crushing injury. Drag marks in the
shape of directional scuff abrasions are often found on the
body, indicating the direction in which the vehicle has
moved over the child's body. Many of these children are
yo unger than J years and boys predominate.
This matter is a serious public health issue. In Aus­
tralia, 16 this type of accident acco unts for 8 per ce nt of pae­
diatric pedestrian fatalities, in New Zealand 10.7 per cent
and in the USA 20 per cent. 17 ,18 Ultrasonic transceivers
located on the rear bumpers have been introduced as prox­
imity warning devices.
ENTRAPMENT ASPHYXIA
The inquisitive nature of children during play actIVIties
may result in their entrapment in objects and spaces from
which they are not able to extricate themselves and thus
they die as a consequence. Discarded chest freezers, fridges,
old safes, large tnmks and suitcases adopted by children as
play areas.can cause such problems.
Entrapment in the luggage compaliments of cars (car
trunks, boots) may have a similar effect. In the latter, there
may also be the added effect of heat if the vehicle has been
parked in a particularly hot day at sites where the sun is shin­
ing directly on to it; heat stroke (hypertherm ia) may develop
in such cases in addition to asphyxial changes. Such deaths
can occur if the outside temperature exceeds 29.5"C. Cars
parked in direct sunlight can reach interna l temperatures of
55°C to 78°C when the outside temperature is 27 -38°C. The
less well ventilated the space in which the child is trapped,
the more likely is it that excessive internal temperatures are
reached . This temperature rise occurs within about 15 min­
utes of co nsta nt exposure to tllis temperature. 19,20
FOREIGN BODY INHALATION
An aspirated foreign object, which is splid or semisolid, can
lodge in the lalynx, trachea or main bronchi of a child. If the
 Strangulation I 331

object is large enough to occlude the airway completely, it were at play. Knight 27 suggested an overstimulation of the
will lead to immediate asphyxia by preventing any gaseous sympathetic nervous system (resulting in arrhythmias, par­
exchange in the lungs and death results within minutes. Pas­ ticularly ventricular fibrillation) caused death.
sage of the object beyond the calina may still be quite dan­ In some of these cases in older children, a gas or a
gerous and can cause serious breathing problems; it may vapour, for example solvent vapour, may have been intro­
also result in death from bronchospasm in those who are duced into the polythene bag and this may also have influ­
susceptible. Common objects include seeds, nuts, bone frag­ enced directly the mode of death.
ments, nails and screws, small toys and pins. 2J.22
As the angles of bifurcation of the main stem bronchi are
OVERLAYING AND WEDGING
acute, for the first 15 years of life foreign bodies may find
their way into either side of the lower airways.23 Once aspi­
Overlaying is the accidental death of a child by smothering
rated, objects may subsequently change position or migrate
resulting from a larger individual sleeping on top of the baby
distally, particularly after spontaneous or external attempts
in the course of deep sleep or sleep induced by extraneous
at removal of the object, after thumping on the back of the
intoxication . This has been a well documented condition
patient or attempting to make him or her cough or retch.
since biblical times 28 and tended to be much commoner in
This may cause delayed obstruction. Inhaled vegetable mate­
Victorian times 29 when parents in a state of inebriation may
rial may swell over subsequent hours or days, and cough,
have taken to bed a large number of their children. 3o This
stridor, wheeziness, breathlessness and cyanosis may ensue.
hazard is greatest in infants below the age of 5 months but
Other objects, such as peanuts and other organic foreign
can occur in children of up to 2 years of age. 3J ,32
bodies, may, in addition, excite an acute inflammatory
The rolling over by the adult to co me to lie on top of the
response with tissue swelling, which makes the level of
child exerts pressure on the child 's face against bedding,
asphyxia gradually worse; children aged 1-3 years are more
the mattress or into the body of the sleeping adult or a
at risk and death results at a rate of 0.7 per 100000 per
co-sleeping older child. Owing to the pressure being exerted
annum in the USA. This is due to the tendency of young
on the child's chest, he or she is unable to cry out and
children to put everyth ing into their mouths and the way in
attract attention. Some of these babies show no pertinent
which they chew food. Because their molars are unerupted
clinical signs, often not even petechiae. 33 Some babies
they tend to use their incisors; objects a nd fragments of food
show the presence of contusions and abrasions, as well as
are then propelled posteriorly, thus exciting a reflex reaction
an unusual distribution of the lividity demonstrati ng the
and a tendency to inhalation. 24 ,25
points of compression. Indeed, it is stated that if a child
who appears to have died of overlaying shows other exter­
nal blunt force injuries, it is likely that the child has been
PLASTIC BAG ASPHYXIA
the subject of non-accidental injury.34,35
In wedging, the child may be wedged by the bed's
In the later 1950s, polythene bags started to be used for
co-occupants against a wall, between the mattress and the
packaging and as supermarket carrier bags. It was soon
wall, the bed frame or an adjacent piece of bedroom furniture,
reported that children had died directly as a consequence of
the cot sides, railing in the cot sides, the head- or footboard of
these plastic bags being placed over the head, although not
the bed, other co-sleepers including other children etc., with
necessarily secured around their neck. 26 In most countries,
similar consequences. 36,37 Another hazard has been associ­
it is now obligatory by law that all polythene bags used for
ated with co-sleeping on waterbeds. Kirchne.-J7 identified 515
wrapping have holes intrinsically cut into them close to the
such deaths in children below the age of 2 years during a 7­
base to allow air to enter if they are slipped over the head;
year period. These included 121 deaths due to overlaying by
however, this legislation does not cover such household
a parent, sibling or by other adult - with 77 per cent of the
items such as rubbish bin bags, carrier bags etc., and the
deaths involving children younger than 3 months; 394 deaths
role of child carers cannot be overestimated with respect to
due to wedging in beds - 296 on regular ad ult beds, 79 on
the prevention of these deaths.
adult waterbeds - with onJy two involving previous alcohol
The imm ediate cause of death in these deaths was at one
or drug abuse, 10 on adult day beds, nine in adult-sized beds
time thought to be due to a decreasing oxygen concentra­
fitted with bed rails. 38 The larger the number of persons occu­
tion within the bag and rebreathing of carbon dioxide, with
pying the beds and the higher the weight of the occupants,
consequent narcosis and loss of consciousness. Physical
the greater is the risk involved. Unsafe sleeping environments
obstruction of the nose and mouth may also have aided
for babies are further discussed in Chapter 11.
and abetted this mode of death. In addition, it was pro­
posed that the bag becam e electrostatically charged and
adhered to the face, aided by condensation of water vapour STRANGULATION
from the expired air. However, as there is a dearth of typi­
cal asphyxial features and the persons concerned looked Non-intentional, accidental self-sJrangulation of young
pale and placid, it was postulated that other phenomena children with loose wires, cords and other potential ligatures
332 I Asphyxial deaths in children
found commonly around the house and often in close
proximity to their beds are well documented. 39 - 4 ! Entangle­
ment in such cords was responsible for 14.3 per cent of Amer­
ican childhood deaths. Other causes were plastic bags, bedding
(non-plastic), cords of blinds or curtains, cords by which
dummy teats (pacifiers) were attached to them and other types
of cord. 32 An intravenous tubing set was also reported as hav­
ing caused such a fatality in a I-month-old boy and a near
fatality in an 8.S-month-old boy,42 and the wiring of an
apnoea monitor, which was non-fatal, in another baby.43
HANGING BY A LIGATURE
Deaths by hanging in children below the age of 14 years
are uncommon. Much of the published literature relates to
individual case reports, with no epidemiological studies
being undertaken. 44 - 46 Cases usually occur against a back­
ground of unhappiness or frank depression. A history of
poor achievement at school, parental and peer bullying and
similar adverse effects come to light after death when a
psychological review is carried out. There may have been
previous attempts at self-harm.
Autoerotic unintentional deaths from suspension have
been reported in children aged 9 years and above. 47 - 49
DROWNING AND NEAR DROWNING
Drowning is defined as death from asphyxia within 24
hours of submersion in water. Near-drowning refers to sur­
vival for 24 hours after a submersion episode. Drowning by
definition is fatal but near drowning may also be fatal. 50
Drowning in children is often classified according to the
temperature of the water into which immersion has taken
place: warm-water drowning at temperatures greater than
or equal to 20°C, cold-water drowning in temperatures less
than 20°C, and very cold-water drowning when the tem­
perature is less than or equal to 5°C.
When a person is submerged in water and attempts to
breathe, he or she may inhale or aspirate water into his or
her airways, and laryngospasm may develop, thus leading
to deceased oxygenation, hypoxaemia and brain death.
Vomiting may occur at this time and gastric contents may
also be aspirated into the airvvays, further complicating
matters. About 10 per cent of persons who drown do not
inhale water or gastric contents.
In near drowning the cardinal features are those of
hypoxaemia of the internal organs. This may affect the
brain, but also and perhaps more importantly in the sur­
vival situation, the lungs.
Most individuals will aspirate less than 4 mL/kg of body
weight; 11 mL/kg body weight is required for the aspirated
fluids to alter the blood volume and about twice this level
is needed to alter electrolyte levels in the blood. Thus fresh
water has earlier effect than salt water. In most patients
who survive, hypovolaemia is a frequent finding owing to
the low oxygen tension causing endothelial cellular dam­
age and hyperpermeabili ty with leakage of body fluids into
the extracellular space. 51 Hyponatraemia may also develop
if large quantities of fresh water are swallowed or if the
syndrome of inappropriate antidiuretic hormone secretion
occurs.
Inhalation of 13 mL/kg of fluid can result in signifi­
cantly impaired gas exchange. Fresh water, being hypo­
tonic, damages type 2 pneumocytes and causes a
disruption of alveolar surfactant. Sea water draws fluid
from the blood into the alveoli and thus dilutes surfactant,
resulting in lower residual pulmonary functional capacity
and pulmonary oedema, with development of an acute res­
piratory distress syndrome.
There may also be plugging of small airways by debris,
which increases airway resistance. Release of inflammatory
chemical mediators from the lungs leads to local vasocon­
striction. Biochemical changes may lead to cardiac dys­
rhythmia or asystole due to hypoxaemia. There may also be
metabolic acidosis and acute pulmonary coextension.
In the USA, for each death from drowning (about 1500
children per annum) an additional four hospital admissions
per drowning death occur after submersion incidents. 52 ,53
Drowning and near drowning are further discussed in
Chapter 18.
IMPOSED AIRWAYS OBSTRUCTION
One form of child abuse that has been shown to occur under
the age of I year is for the parent, usually the mother or
another carer, to obstruct the mouth with a soft object such
as a pillow, clothing or by pressure against her breast or in
some other way. The child does not seem to struggle too
much or too ovelily and develops hypoxia and cyanosis; he
or she may also demonstrate convulsions due to cerebral
hypoxia and, eventually, a respiratOlY and cardiac arrest.
This may occur when a mother with psychosocial problems
is attempting to attract attention to herself, formerly referred
to as Munchausen's syndrome by proxy.
Coveli video surveillance has been used to document
such abuse, with a strong debate as whether this is ethical
and legal. 54 Whether the end (namely convincing the Crim­
inal Court of the abuse by the compelling evidence of a
video) justifies the means is dubious in the mind of many
of the professionals involved in such cases. This form of
surveillance was first used in 1983 55 and since then several
reports of it have appeared. 56 - 62 The method used is to have
the child attached to a multifunction polygraph (recording
respiratory raster, ECG, EEG, pulse rate, blood pressure)
in a cubicle, thereby ensuring that the infant cannot be
moved away from the area of surveillance and thus having
physiological criteria that can be compared and timed with
images recorded on camera.
These children present with an aClJte life-threatening
event, 'breath holding attacks' or recurrent convulsions.
-- -- ----------------------------------
-- --
-- ----

It may be extremely difficult to differentiate between a nat­
ural phenomenon and abuse phenomenon.63
It can be extremely difficult to find any signs of the
external airways obstruction, although a very careful
search for bruising on the face, on the inside of the lips and
cheeks should be made; in the mucosa l sites bruising may
be accomp anied by minor abrasions.
Internally, the presence of haemosiderin in the lungs,
within both alveo lar cells and in the interalveolar septa, is
of great significance. In the acute phase there will have
been haemorrhage within alveoli and this may give rise to
frothy frank ly blood-stained fluid emerging from the nose
and mouth.
ABUSE OF INHALANTS (SOLVENT ABUSE)
Inhalants, usually hydrocarbons, are breathable chemical
vapours that produce mind-alteri ng effects similar to the
effect of alcohol consumption or the use of controlled dnlgs.
They are an ingredient in many househoJd products, such as
agents used in cleaning, decorating, painting (turpentine,
white spirit), sta ins, vamishes, g lu es, air fresheners, hair con­
ditioners, dry cleaning agents, shoe polish, colour markers,
na il polish removers, spot removers and degreasers and
deodorants, as well as in fuels such as petrol (gasoline), butane
and lighter fuel. Anaesthetic agents, such as ether, chloroform,
halothane and nitrous oxide, may be used in a similar manner.
The so lvents are taken in to the body by: squirting
directly into the mouth from cans and canisters, inha ling
from bags or an aerosol can, sn iffin g directly from the lid
of the container, breathing in directly (huffing) or after
application of the substance in a rag or handkerchi ef. Plas­
tic bags containing the vapour may be placed directl y into
the mouth. There are about 1400 compounds that can act
in this manner, including nitlites, alcohols and halo­
genated compounds, in addition to hydrocarbons.
Death may occur in a variety of ways:
• direct acute toxicity of the brain, leading to respiratory
a nd cardiac arrest;
• cardiac arrhythmias brought on by the substance
inhal ed, followed by cardiac arrest; this may be
brought about by sudden SPUltS of muscular activity
and release of a bolus of catecholamines, as if the
subject while under the influence is chased and made
to run away;
• suffocation by the item being used to present the
vapo ur to the mouth, for example plastic bag or
balloon;
• vomiting w hile unconscio us from the effects of the
inhalants;
• suffi xa tion by displaci ng air from the lungs with an
irrespirable vapour;
• acute parasympathetic stimulation with inhibition if
me substances cause irritation and/or cooling of the
upper gullet or n asophary nx;
Prevention I 333
• accidents while und er the influence of these
substa nces, due to loss of judgement and impairment
of other cognitive functions;
• burns produced by inadvertent ignition of the
inflammable accelerant used;
• long-term toxic effects of the solvent on the marrow,
liver, kidneys and brain;
• chronic respiratory problems due to chronic bronchitis
and recurrent inhal ation al or chemical pneumonitis.
Note that in some older children, these substances may
be used as sexual aid in the course of autoerotic activities.
In cases when such abuse is suspected, there may be lit­
tle to find at autopsy. There may be a rash around the
mouth and blisters where the substance had been app lied
and produced a solvent degreasing effect on th e skin. The
so lve nt may still be appreciable through its smell at
autopsy. One lung should be retai ned in a nylon fibre bag
to allow 'a ir' to be sampled from it directly from its airways
for gas chromatography assay. The blood , the liver and the
kidneys can also be assayed toxicologically. Histology may
confirm chronic changes in the internal organs particularly
the lungs. No petechial haemorrhages are found either
internally or externa lly, except in the few instances of
suffocation.
REVERSE SUSPENSION
This is perhaps a rarity, which may occur in children dur­
ing play activity when death results from exhaustion of
breathing du e to splinting of the diaphragm 64 ,65 by the
upwards displacement of abdominal contents. Death occurs
slowly from ex haustion of respiratory effort.
CHEMICAL ASPHYXIA
This term is applied to situations when irrespirable gases
find their way into the child's environment. For example,
carbon monoxide in fires and from exhaust fumes (in
closed garages) and from charcoal fires (barbecues) in an
enclosed space, chlorine escaping from swimming pools,
hydrogen sulphide and methane from refuse tips, old mines
and slurry pits are among the gases that can cause death
after inha latio n. These events usu ally involve older, active
children and, more frequently, boys, in common with many
environm ental mishaps.
PREVENTION
As incidents involving asphyxia are common, public
health offiCials, community doctors, midwives and health
visitors have been involved in several programmes aimed
at alerting parents to the dangers of asp hyxia. 66 In severely
334 I Asphyxial deaths in children
disab.led children accidental asphyxial deaths are more
common. 57 In the neonate and in young children below the
age of 12 months sleeping practices also have to be care­
fully scrutinized to enquire that they are not potentially
dan gerous to the child. 6B - 72 Co-sleeping in particular
should be viewed with some discretion, particularly if the
parents indulge in smoking and drinking of alcohol. 73 ­ 75
REFERENCES
Tardieu A. Etude medico -lega le sur les sev ices et ma uva is
tra itmen ts exe rces sur des enfants. Ann Hyg Publ tvled Legale
1860; 13 :36 1-98.
2 Luke JL. Conjunctival petechiae. IV Eng/} iVIed 1971 ;
28 4:1101.
3 Rao VJ , Wetli CW The foren sic sig nifica nce of co njun ctiva l
petechiae. Am } Forensic ivfed Pathol 1986; 9:32 ­ 4.
4 Jaffe FA. Petechial haemorrha ges. Am } Forensic iVIed Patl10l
1994 ; 15:203 -7.
5 Dow nes AJ, Crossland OS , Mell o n AF. Preva lence and
distribution of petechiae in well babies. Arch Dis Child 2002;
86:291-2.
6 Hood I, Ryan 0 , Spitz WU. Resuscitation and petec hiae. Am }
Forensic iVIed Pat/101 1988; 9:35-7.
7 Beckwith JB. Intrathoracic petechial hemorrhages; a clue to
the mechanism of death in SlOS. Ann N Y Acad Sci 1988;
533:37-47.
8 Krous HF, Jordan J. A necro psy study of distribution of
petechiae in non-SlOS. Arch Pathol Lab Med 1984; 108:75- 6.
9 Kleeman WJ, Wiechern V, Schuck M, Troger HD . Intrathoracic
and subconjunctival petec hi ae in SlOS . Forel1sic Sci lilt 1995;
72:49-52.
10 Bass M, Kravath RE, Gla ss L. Death-scene inves ti ga ti on in
sudden infant death. N Engl} Nled 1986; 315 : JOO-5.
II Iyasu I, Hanzlich R, Rowl ey 0, Willin ger M. Proceedings of
'Workshop on Guid elin es for Scene Investigat ion of SlO S'.
} Forensic Sci 1994; 39 :11 26 - 36.
12 CDC. Guid elines for dea th scene investigation of sudde n,
unexplained infant deaths: recommenda ti ons of th e
interagency pane l on SlOS. MMWR 1996; 45:RR-IO.
13 Haller JA, Donahoo JS. Traumatic asphyxia in ch ildren:
pathophysiology and management.} Trauma 1971; 11 :453-7.
14 Campb ell-Hewson G, Egleston CV, Cope AR. Traumatic
asphyxia in children. Accidel1t El11 erg iVIed 1997; 14(1):47-9.
15 Sarihan H, Abes M, Akyazici R et a!. Traumatic asphyx ia in
children. Cardiovasc Surg 199 7; 38:93 -5.
16 Holland AJA, Liang RWY, Sing h SJ et al. Driveway motor
ve hi cle injuri es in children. iVI}A 2000; 173: 192 - 5.
17 Brison RJ, Wicklund K, Mueller BA. Fata l pedestrian injuries
to yo un g children: a differe nt patte rn of injury. Am } Public
Health 1988; 78 :79 3- 5.
18 Roberts I, Norto n R, J ackson R. Driveway-related c hild
pedestrian injuries; a case control study. Pediatrics 1995;
95:405-8.
19 Sumra ll RE, Petty CS, Holman W. Temperature in closed
auto mob iles in hot weather. Forensic Sci Gazette 1976; 7:7-8.
20 Surpure J. Heat-rela ted illness a nd the a uto mob il e. Ann
Em erg iVIed 1982; It :263-5.
21 McGuirt WF, Ho lmes HD, Feehs R et a l. Tracheo-bron c hial
fo reign bodies. Lm)'ngoscope 1988; 98:615- 18.
22 Weiisberg D, Sc hwartz I. Foreign bodies in the tracheo­
bronchial tree. Chest 190 7 ; 1: 730 - 3.
23 Cleveland RH. Symmetry of bronchia l angles in chi ldren.
Radiology 1979; 11 3:89 - 93.
24 Byard RW. Mec han isms of death in in fa nts a nd young
children fol.l.o wing fore ign body ingestion. } Forensic Sci
1996; 41 :438-41.
25 Baharloo F, Veyckemans F, Fra nci s C et al. Tracheo-bronchial
fore ign bodies: presentation a nd management in children and
adults. Chest 1999; 115 : 1357 - 62.
26 Plast ic bags (Ed itoria l). BM} 1959; 1 :1463-4.
27 Knight B, Sa ukko P. Forensic Pathology, 3rd edn. Oxford:
Oxford University Press, 2003 .
28 The Bible, 1 J<jngs Ch . 3 v.19.
29 Templeman C. 258 cases of suffocation of children. Edin Med
} 1891; 38:22- 6.
30 Nakam ura S, Wind M, Dane ll o MA. Review of hazards
associated w ith chi ldren placed in adult beds. Arch Pediatr
Adolesc iVIed 1998; 1 53: 10 19- 23.
3 I Drago DA, Dannenberg Ai. Infant mechan ical suffocation
dea th s in the United States 1980-1997. Pediatrics 1999;
15 3: 103 -59.
32 O'Hara M, Harruff R, Smialek J , Fow ler D. Sleep location and
infa nt s uffocat io n : how good is the evidence? Pediatrics
2000; 10 :9 15-20.
33 Moore L, Byard RW Path o log ical fInd ings in hang ing and
wedgin g deaths in infants and yo un g children. Am } Foren sic
ivIed Pa tl101 199 3; 14 :296- 302.
34 Collins K. Death by overlaying and wedg ing: a 15-yea r
retrospective study. Am} Foren sic iVIed Patho1200 1; 15 : 155-9.
35 Mitchell E, Kraus HF, Byard RW Pathol ogica l findin gs in
ove rl aying.} Clin Forensic iVIed 2002; 9:133 - 5.
36 Byard RW, Beal S, Bourne AJ. Potentially dangerous sleep ing
environment and accidental asphyxia in infancy and ea rly
ch ildhood. Arch Dis Child 1994; 71 :497-500.
37 Kirchner JT. Deaths associated with small children sleeping in
adult beds. Am Acad Fam Phy 2000; March I ,
www.aaf.org/afp.
38 Nakamura S, Wind M, Danello NA. The pediatric forum:
should infants sleep with their parents 7 Arch Pediatr Adolesc
ivIed 2000; 154:1171-3.
39 Nixon JW, Kemp AM, Levene S, Sibert JR. Suffocation,
choking and strangulation in children in Eng land and Wales;
epidemiology and prevention. Arch Dis Child 1995; 72:6 - 10.
40 Rauchschwalbe R, Mann NC. Pediatric wind ow cord
strangulations in the United States, 1981-95. }AMA 1997;
72: 1696-8.
41 Sabo RA, Hanigan We, Flessner K et a!. Strangulation injuri es
in children.} Trauma 1996; 40:68-72.
42 Garros D, J<jng WJ, Brady-Flyer Bet a!. Strangulation with
intravenous tubing; a previously und ec id ed adverse even t in
children. Pediatrics 2003; III :732-4.
43 Emery JL, Taylor EM, Carp ente r RG, Wa ite AJ. Apno ea
monitors and accidental strangulation. BM} 1992; 304 :11 7.
44 Petruk J, Shiels E, Cummings GE, Francescutti LH, Fatal
asphyxiations in ch ildren involving drawstrings in cloth in g.
Can Med Assoc} 1996 ; 155: 1417- 19.
45 Bya rd RW, Ma rcopo ul os D et al. Early adolescent suicide; a
co mparati ve study. } Clin Forensic Med 2000; 7:6-9.
46 Wyatt J, Wyatt PW, Sq uires T, Busuttil A. Hang ing deaths in
children . Am } Foren sic Med Pathol 1998 ; 19 :3 43-6.
47 Feldm a n KW, Simms RJ. Stra ng ulatio n in children:
epidemiology a nd clinica l co urse. Pediatrics 1980;
65: 1079-85.
48 Cooke CT, Cadden GA, Hilton JMN. Hanging dea ths in
children. Am } Forensic Med Pathol 1989; 10:98- 104.
49 Uva JL. Autoero ti c asphyxiation in the United States.
} Foren sic Sci 1995 ; 40:574 - 8 1.
50 Bya rd RW, Lipset J. Drowni ng deaths in tod d lers and pre­
ambulatory children in South Austra lia. Am} Forensic iVIed
Path ol 1999; 20:328-3 2.


51 Nodlkel JH. Drow ning. N Engl ] Med 1993; 328:253-6.
52 American Academy of Pediatrics Committee on Injury and
Poison Prevention. Drowning in infants, children and
adolescents. Pediatrics 1993; 92 :292-4.
53 Wintemu te GJ. Childhood drowning and near drowning in the
United States. Am ] Dis Child 1990; 144:663-9.
54 Byard RH. Covert video surveillance in Munchhausen
syndrom e by proxy - ethical comprom ise or esse ntial
technique? Med ] Aust 1994; 160 :352-6.
55 Rosen CL. Frost JD, Glaze DG et al. Ch ild abuse and recurrent
infant apnea.] Pediatr 1986 ; 109 : 1065- 7.
56 Southall DP, Stebbens VA, Rees SV. Apnoe ic episodes induced
by smothering. Two cases identified by covert video
surveillance. EM] 1987; 294:1637-41.
57 Samuels MP, McLaughlin W, Jacobson RR. Fo urteen cases
of imposed airway obstruction. Arch Dis Child 1992;
67:162-79.
58 Byard RW, Burnell RH. Covel1 video surveillance in
Munchausen syndrome by proxy: ethics compromise or
essential technique? Med ] Austr 1994; 160 :352- 4.
59 So uthall DP, Plunkett MCB , Banks MW et al. Covert video
recordings of life-threatening child abuse: lessons for child
protection. Ped iatrics 1997; 100:73 5-60.
60 Foreman OM, Farsides C. Ethic al use of covert videoing
techniques in detecting Muncilausen synd rome by proxy.
BM] 1993 ; 307:611-13.
61 Thomas T. Covert video surveill anc e. New Law] 1994;
114:9 6-7.
62 Shabde N, Kraft AW. Covert video surveillance: an important
in vestigative tool or a breach of trust? Arch Dis Ch ild 1999;
81 :291-4.
References I 335
63 Recurrent apparent life- threatening events and intentional
suffocation (Editorial). Arch Dis Child 1999 ; 81:189.
64 Friberg TR , Weimreb RN. Ocular manifestations of grav ity
in version. ]AMA 1985; 253:1755-7.
65 Busuttil A, Obafunwa J. Recreational abdominal suspe nsio n: a
fatal practice. A case repol1. Am ] Forensic iVIed Pathol 1993 ;
14: 141-4.
66 Towner E, Errington G. The Epidemiology oj Choking in
Childhood and [l17pli catiollsjo r PrelJentiol1. A report prepared
for the Department of Health in London, 2002.
67 Ama nuel B, Byard RW. Accidental asphyxia in severely
disabled children. ] Paediatr Child Health 2000 ; 36:66- 8.
61l Beal SM, Byard RW. SIDS in South Australia 1968-97. Part 3:
Is bed sharing safer in infants? ] Paediatr Child Health 2000 ;
36:552-4.
69 Byard RW, Beal SM. 'V'-shaped pillows and un safe infant
sleeping.] Paed ia rr Child Health 1997; 33:171 - 3.
70 Byard RW, Bourne AJ, Beal SM. Mesh side cots: yet another
potentially dangerous infant sleeping environment. Forensic
Sci [l1t 1996 ; 83: 105- 9.
71 Moore L, Bourne AJ, Beal S et at. Unexpected infant death in
association with suspended rocking cradles. ] Forensic Med
Patho11995; 16 : 177-80.
72 Byard RW, Beal SM. Simpson A et al. Accidental infa nt deaths
and stroller prams. Med] At/st 1996; 165:140-1.
73 Byard RW. Is co-sleeping in infancy a desi rab le or dangerous
practice? ] Paediatr Child Health 1994; 30: 198 - 9.
74 Byard RW. ls breast feeding in bed always a safe practice?
] Paediatr Child Health 1998; 38:498-9.
75 Scragg R, Mitchel l EA, Taylor A et al. Bed sharing, smok ing
and alcohol in the SIDS. BM] 1993; 307: 1312-18.
I CHAPTER 17 I
ACCIDENTAL INJURIES IN CHILDREN
Anthony Busuttil
Overview of paediatric trauma 336
Bicycle helmets 338
Falls 339
Playground injuries 340
Sports injuries on snow and ice 340
Riding injuries 341
OVERVIEW OF PAEDIATRIC TRAUMA
the age of 2 years are involved in many serious accidents,
Tra uma of children has always been a major public health
concern in that a significa nt degree of disability, loss of
schoo ling and mortality arise as a consequence. Three
major sUlveys that looked at these aspects are: (1) the
Canadian Hospital Injury Reporting and Prevention Pro­
gram (CHIRPp),l based on 16 out of the 750 hospitals in
Canada; (2) the Australian National Injury Surveillance and
Prevention Program;2 and (3) the programme commis­
sioned by UNICEF Innocenti Research Centre of Florence)
(Table 17.1). The annual hospital health-care spend in the
USA accrues to abo ut £5.1 million 4 for injuries incurred
by children. In 1971, deaths from injuries in countries that
belong to the Organisation for Economic Co-operation
and Development (OECD) in the 1- to 14-year age group
accounted for about 25 per cent of mortality in chi ldh ood
(Table 17.2) . In the 1990s, more than one in three deaths
in this age group were the result of injury. This has not
changed in any appreciable manner since then.
In such assessments of paediatric trauma, recorded sur­
veys vary in the cut-off point in relation to the ages con­
sidered. It is often the case in many of the published series
that persons up to 18 years of age are considered. This may
skew the figures, given the increased activity of adoles­
cents over the age of 14, a nd the high incidence of injury
after the age of 12 years. It is unlikely that children under
-- - -- ----
Agricultural injuries 341
Prevention 341
Older children and substance abuse 341
Accidental poisoning 342
Hypersensitivity 342
References 342
an d any injury below the age of 1 year is sufficiently
uncommon that it should always be treated with a mod­
icum of suspicion .5 A careful match should be attempted in
all instances between the findings on medical examination/
investigation and the narrative given by the carers to doc­
tors as to how the accident occurred. Furthermo re, the for­
mat and data conventions followed by different countries
vary,6 as does the accuracy of data coding.7
In general terms, boys are injured abo ut twice as often as
girls,S except in household injuties when girls tend to pre­
dominate. The difference between sexes increases after
infancy and continues through a ll age groups into adoles­
cence. 9 The largest gro up of injured chi ldren is between the
ages of 5 and 9 years, followed by children aged 1-4 years
and children aged 10-14 years. Infants under 1 year have the
lowest rate. JO
The vast majotity of injuries occurred between noon and
midnight when children are most likely to be out of the
structured schoo l environment. The two most common sites
of injuries were the road and home. II Other places included
recreation areas, public places, schools 12 - 14 and, less fre­
quently, farms and workpl aces.
Children from deprived backgrounds are more prone to
injury. This may be related to a lack of adequate supervision,
increased risk-taking and a greater proportion of the ch ild's
day spent on the street. 15-18
~
 Overview of paediatric trauma I 337

Table 17.1 Mortality (deaths per 700000) from injury: ch ildren aged 7- 74 year53

All injuries Transport Fire and flames Falls Poisoning Drowning Firearms Intentional Others

UK 6.06 2.91 0.66 0.26 0.12 0.39 0.0 1 0.80 0.91

Canada 9.68 4.33 1.01 0.20 0.12 1.26 0.12 1.45 1.19
USA 14.06 5.76 1.65 023 0.17 1.74 0.4 2.74 1.37
New Zealand 13.67 6.93 0.82 0.58 0.13 1.87 0.14 1.74 1.46
Australia 9.53 3.37 0.68 0.22 0.14 1.97 0.04 0.85 1.26
Mexico 19.75 6.05 0.62 1.09 0.55 3.30 0.35 2.90 4.89
Denmark 8.10 3.99 0.52 0.13 0.05 1.07 0.07 0.94 1.33
Spain 8.12 4.02 0.30 039 0.17 1.12 0.05 0.33 1.74
Italy 6.14 3.30 0.18 0.51 0.12 0.46 0.03 0.50 1.04
Czech Republi c 11.95 4.64 0.34 0.61 0.49 2.23 0.03 1.61 2.0
Portugal 17.76 8.65 0.6 0.75 0.46 1.34 0.06 4.23 l.67
Korea 25.57 12.59 0.91 1.18 0. 75 5.14 0 1.08 3.92
From the WHO mortali ty database, with perm ission.

Table 17.2 Mortali ty fram injuries in Organisation for Economic
Co-operation and Development (DECO) countries, by age and
gender 1991- 7995, per 700000 af the population
Age (years)
1-4 5-9 10-14
Boys 21 16 16
Girls 14 7 7
From the WHO mortality database, with permission.
The use of the term 'accident' to describ e such traumatic
episodes is often a misnomer, as few of th ese injuri es are
actually due to unpredictable irremedi a bl e ev ents and situ ­
ations; rather they are more often due to interaction with
the environment in which children are living and playing,
the products that they use and the level of supervisio n t hat
they are allocated. Many of these injuries a re preventable
and thus the main thrust of the investigation a nd ana lys is
of figures from childhood injuries should be in attem pts at
prevention. 19-21
There is no simple correlation between mOltality staris­
tics and non-fatal injury data. 22 FUithermore, the relarive
population m ortali ty data are hig her in Eastern Europea n
countries. 23 Why this should be so is difficult to determine
but, at least in palt, it may be re lated to the different
perception of the seriousness of injury a nd a populist
fatalistic attitude; there may be a tendency to decry the
seriousness of injury combined with Jess willingness to
seek medical help for injury until it is too late.
Accident Proneness in Children
Intrinsically the behaviour of children is based on curiosity
and constant activity. Many paediatricians would class as
abno lmal an inactive, sedentary young child . This combina­
tion of childhood inquisitiveness, mobility and activity te nds
to be impulsive and experimental. This is superimposed on
a n obvious backdrop of inex perience and a n inherent imma­
turi ty of social and intellectual skills to cope wit h adverse
situat ions as and w hen they present th emselves. By their
very nature, children take risks and are sensation seeking.
All of th is compounds their proneness to injury.24
Falls are therefore inevitable. The graspi ng and pulling
of objects towards the infant may result in objects being
pulled on to the child, and to scalds and burns.25 As the
child's territory expands and manual dexterity and p hysical
strength improve, play activity becomes more adventuro us
and the opportunities for injuries increase. As organized
sporting activities are taken up these further raise th e prob­
ability of injury, be this in organized and team sporting
activity - participation in team sports such as rugby, soccer
and hockey - or, more simply, in such recreational activiti es
as kicking a ball, riding a bicycle or using a sWing. 26 The
more v igoro us the sport (e.g. karate, judo, Tae Kwon Do,
boxing and skateboarding) the more likely is the occurrence
of inju ryY
The role of carers in ens uring that the children do not
take excessive risks, and indeed to be t here to rescue them
from the hazardous situ at ions in w hi ch they find them­
selves, is essential. When the carers are employed in this
capacity by nurseries and schools, there is an ex pectation
of a duty of care. Sometimes - indeed , increasingly fre­
quently - this is called to account in a cou rt of law when
this duty is perceived to have been breached. 28 Participa­
tion in sport also carries with it the possibility and, per­
haps, in more viol ent sporting activity, a probability, of
almost inevitable injury. This willing, wholehearted partic­
ipation by the child with the sanction of his pa rents, often
expressed legally in th e max im volenti non fit iniuria ('to one
who is willing, no ha rm is done'), m ay decrease the culpa­
bility and answerability of the carer when things go wrong
338 I Accidental injuries in children
but it also bears the expectation and responsibility from
them that all that can be done to supervise children's activ­
ities - to decrease the risk, to provide protective clotting
and equipment - has been fulfilled in full exercise of their
duty of care. 29
Causes and Mechanisms of Injuries
Falls are the leading cause of injuries. Among young chil­
dren these are often from fu rniture, such as cllanging tables,
sofas and couches, cribs, beds or fallin g down stairs, pal1ic­
ularly in walkers,JO during momentary periods of inattention
by their carers or because of the activity of older siblings.
As toddlers begin to explore, the oppOl1Unity for falls from
heights increases; this includes recorded falls down stairs.
Less common are falls from balconies, porches or out of
windows;Jl these inciden ts are more frequent in warmer
climates when children have more open access . Nearly
20 per cent of the children adm itted after falls have fallen
2.5 metres (8 feet) - the height of one storey or more.
The next major mechanism of injury is motor vehicle
crashes with children as occupa nts. Most crashes involve
collisi ons with other vehicles. Only one-quarter of those chil­
dren injured have been using some kind of protective
restraints, such as child safety seats or safety belts. Many
unrestrained children were hurled aro und inside the car dur­
ing a collision or thrown out of the car through windows or
doors. Breast-feeding in the car or holding a child in one's lap
is dangerous as the child will be unprotected during a crash.
Many other children are injured fatally when they walk,
cycle or play on the roadway and are struck by motor veh i­
cles.J2 Thousands of children are killed annua lly on the
roads but most of the deaths are incurred when the child is
on foot. There is correlation between the number of child­
hood fatalit ies on the roads and the size of the popUlation of
a country : the bigger the child population and the lower the
socioecon omic status, the higher is the likelihood that unsu­
pelvised children are on the road and thus find their ways
into the path of moving traffic. This may also be the reason
for the disproportionately higher incidence among children
of certain ethnic groups.JJ Many of the deaths occur in the
urban location at speeds of less than 30 miles per hour and
usually within a Sh0l1 distance away from their homes.
Most of these deaths occur in the holiday periods and a t
other times during the parts of the day w hen the children
are not actually at school. Thus more injuries occur in the
summer months. The institution of school crossing patrols
manned by retired individual s and the training of children
in road safety has produced major dividends in terms of
reducing co llision-related injuries and deaths amo ng chil­
dren; the aims have been to teach children to be vigilant
and to improve their skills in quickly integrating and
assessing movement, relative velocity, distance and depth
in relation to moving motor vehicles, along with the inter­
pretation of road signs. 34 The setting aside of exclusive
cycle tracks in urban areas has also assisted in the decrease
of such incidents. J5 ,36
Children are most frequently hit while walking along the
side of the street, during play in the street or while darting
into or crossing the street. Children are occasionally hit by
motor vehicles while bicycling, but they may also be injured
when fallin g off their bicycles after hitting objects such as
trees, kerbs or walls. Few of those children admitted to
trauma units after bicycling-related injuries, were wearing
helmets. 37 Other motor vehicle-related injuries involve
motorcycles, all-terrain vehicles and recreational vehicles.
Children with hearing impairment from any cause,
including chronic ot itis media a nd 'glue ear', are more
prone to accidents as they may not be aware of vehicles in
their vicinity.
Consequences
Many injuries occur while a child is engaged in some kind of
activity such as nll1ning, playing, climbing or bicycling. Even
when the child is in a more station ary position, such as rid­
ing in a car, the child is part of an activity involving motion
and speed. The combined effects of motion, speed and impact
can injure many regions of a child's body. Two or more body
regions were affected in about one-half of the children. Chil­
dren typically had multiple injUlies that resulted in several
diagnoses: head inj ury is the most frequent diagnosis among
children recorded, followe d by fractures to the bones of
extremities and torso. The third most frequen t diagnosis is
open wounds, mostly lacerations.
The severity of injuries is directly linked to how the chi l­
dren were injured. Children who are injured in motor vehi­
cle crashes and as pedestrians are more likely to have four
or more functional limitations. The force and speed of col­
lisi ons are factors determining the severity of injuries. Chil­
dren injured by falls are less seri ously injured and have
fewer functional limitations.
BICYCLE HELMETS
Experienced, careful bicycle riders crash every 4500 miles
on the average. Head injury causes 75 per cent of our 800­
plus an nual deaths from bicycle crashes. Medical research
shows that a bicycle helmet can prevent 85 per cent of
cyclists' head injuries. 38.39
A helmet reduces the peak energy of a sharp impact. This
requires a layer of stiff foam to cushion the blow when
crushed. Most bicycle helmets do this with the use of
expanded polystyrene (EPS), the white picnic cooler fo am.
Once crushed, EPS does not recover. Another foam,
expanded polypropylene (EPP), does recover but is much less
commonly utilized. A stronger EPS called GECET appeared
in 1992 and is widely used now and another synthetic foam,
expanded polyurethane (EPU), is used in Taiwan as it has a

unifonn cell structure and crushes without rebound; it is
heavier and requires a manufacturing process that is not
environmentally friendly. The spongy foam placed on the
inside a helmet is for comfort and fit, not for impact.
The helmet must stay on your head even when the head
is hit or bumps on to a hard unyielding surface, more than
once in a moving traffic incident: usually there is a bump
with a ca r first, and then with the road. It thus needs a
strong strap and an equally strong fastener. The helm et
should sit level on the head and cover as much as possible
of its surface contour. Above all, with the strap fastened,
one should not be able to get the helmet off your head by
any combination of pulling or twisting. If it comes off or
slips enough to leave large areas of the head unprotected,
the straps must be readjusted otherwise a better-fitting hel­
met will be requ ired . The strap must feel comfortably snug
when riding the cycle.
Most helmets made of EPS foam ha ve a thin plastic
outer s hell. The shell helps the helmet skid easily on rough
pavement to avoid jerking of the neck during such an
impact; it also holds the EPS together after the first impact.
Some excellent helmets are made by moulding EPS in the
shell rather than adding the shell later.
Excessive vents in the helmet mean less actual area con­
tact with the head, which could concentrate force on one
point. 'Aero' helmets are not noticeably faster and in a
crash the 'tail' could snag or knock the helmet aside.
Skinny straps are also less comfortable. Dark helmets are
hard for motorists to see. Rigid visors can snag or shatter in
a fall. Helmet standards do not address these problems a nd
should be reconsidered.
There can be little doubt that bicycle helmets have pre­
vent some of the more serious head injuries and undoubtedly
they have saved the lives of young cyclists. 4o - 42 However, it
has been observed that some children, particularly older
children, need to be objectively convinced that helmet wear­
ing is indeed not just useful in preventing injUl)' and that the
money spent on their purchase rather than other cycl in g
clothes and accessories is money well spent. Above all cycle
helmets had to be shown not to make their wearers less 'cool'
and trendy but vice versa, and also that their use makes the
wearers appear more professional. 43 A variety of strategies
have been devised worldwide to encourage the purchase and
wearing of cell helmets with varying success levels. 44- 47
FALLS
All children trip and fall, indeed from the time that initially
they stand up on their own two feet. Falls a re more com­
mon while the muscle tone is not fully developed but
become more frequent once aga in when gait allows chil­
dren to explore farther and when their innate inquisitive­
ness takes them further afield.
Contusions, haematomas, lacerations and abrasions
tend to occur at the site of such fa ll s. OccaSionally, if more
Falls I 339
severe force is in volved or in falls from a height, fractured
long bones a nd closed head injuries may occur 48
The incidence of injury depends to a major extent on the
height from which the fall occurs and also on the terrain
that the chi ld lands on - the harder and more compact the
latter, the higher is the incidence and severity of injury.49
The height of climbing frames, swings, etc. is another
factor involved in injury causation and the height of the
fa ll is directly related to the severity of the injuries sus­
tained by children in playgrounds. Injuries are more likely
to occur with equipment that is two 2 metres above the
ground ,5o so that reducing the height of swings, monkey
bars and other equipment improves safety.
Reports of children falling out of windows are all too
frequent in the summer. Height off the ground and landing
surface are two critical factors in the injuries to children
that result from falls. Height is directly related to the speed
at wh ich a child falls and the force of impact upon landing.
Falls are particularly common among children of p re­
school age who are developing their balance and learning
to walk, climb and run ; however, their physical skill s
develop sooner than their abilities to anticipate the conse­
quences and risks of their actions.
Amongst falls from heights, about one-quarter of chil­
dren fell from w indows. Examples of other falls from
heights include falls from roofs, out of trees, from ski lifts,
off porches or balconies and down flights of stairs. Even
escalator falls have been known to cause serious injuries. 51
The incidence of injuries among children h as a distinct
seasonal pattern, with an increase in the warmer months of
the year. This is even more so for falls from windows as
they are more likely to be open in warmer weather. Tod ­
dlers and pre-school children are particularly vulnerable as
they lack the judgement needed for risk assessment an d
safety precautions. Falls occurred in th e chi ld's hom e in 96
per cent of the cases. Falls from two- to three-storey win­
dows were most common (72 per cent), followed by fo ur
to five storeys (18 per cent). Only 8 per cent of children
admitted to trauma centres fell from firs t-sto rey windows;
however, 3 per cent fell from six storeys or more.
Head injuries, with or without dam age to other body
regions, occur among 65 per cent of the children . A small
child 's head is at special risk for injury from a fall because
it is disproportionately large and heavy in relation to the
rest of the child 's body. Height, rate of fall and the force of
impact have combined effects that result in injuries to
multiple body regions in about one-half of injured children.
The greater the height of the fall, the higher the death rate.
Among the children who fell from windows, six storeys or
more, 8 per cent died. Among those who fell four to five
storeys, 5 per cent died; among those who fell two to three
storeys, 2 per cent died. Non e died who fell from one storey.
Upon hospital discharge, 73 per cent of the children who
fell from windows had no identified functional limitations.
For the 21 per cent who had one to' t hree limitations, these
were most commonly in bathing, dressing, walking and
340 I Accidental injuries in children
se lF-Feed ing. By contrast, the children with four or more
functional limitations at discharge had more difficulties in
cogn ition , behaviour speech, vision and hearing, reflecting
the severity of head injuries in this group. Among t he chil­
dren who fell from six or more storeys, 42 per cent had one
to three limitatio ns; 8 per cent had four or more limitations.
More functional limitations were identified as the height of
t he fall increase d; children who fell from four to five storeys
had the largest proportion of functional limitations - four or
more. Ho wever, because more children died who fell from
fi ve storeys or more, there were fewer surv ivors in that group
with four or more limitations.
Almost one-half (47 per cent) of the falls occurred on
playgro unds in recreational areas, one-quarter (25 per cent)
occurred at home, and almost one-fifth (19 per cent)
occurred on sc hool playgro unds. About 90 per cent of these
injuries occurred in falls from playground eq uipment:
sw ings, slides, climbers (monkey bars, jun gle gyms or other
climbing apparatus) and trampolines. Children fell from dif­
ferent types of playground equipment depending where the
equipment was located. At home, children most often fell
from swings or trampolines, whereas at schoo l and in recre­
ation areas they most often fell from climbing equipment.
Children of different ages fell from different types of
playground equipment. The youngest children fell most
often from slides, whereas older children fell most often
from climbing equipment.
Seven of the children died in the hospital: six from head
injuries and one as a result of a severed sp in al cord. Three
of these children fell from slides, two fe ll from sw in gs, o ne
fell from monkey bars and one fell from a trampoli ne. Four
of the children who died fell from playground equ ip men t at
home; the other three fe ll fro m playground equ ip me n t in a
recreational area .
PLAYGROUND INJURIES
Playground s a re considered as oases from road traffic ­
places w here children can safely enjoy themselves and
indeed develop interpersonal, social and team skills. As
childhood obesity is bandied around as being of epidem ic
proportions, t he impo rtance of play activities cannot be
overemphasized. Ho wever, playgrounds - whether at
hom e, at schoo l or at rec reational areas - can also be dan­
gerous. To be safe, all public and private playgrounds must
be well designed , we ll mai nta ined a nd used properly. Chil­
dren need good su pervision , and sho uld be taught how to
play safely. Children enjoy experi men tatio n a nd explo­
ration and will incur risks, w hether calculated or hapha z­
ard, a nd thus although play areas have to be challenging
and inn ovative to keep the ch ildren 's interest, there should
also be a backdrop of safety built into playgrounds. Mod­
ern playgrounds wi ll often h ave eq uipment such as a
seesaw, merry- go -round, swings, slide, climbing frame,
walking bridge, jungle gym, c hin-up bars, sandbox, spring
rider, monkey bars, overhead ladder, trapeze/trapeze rings,
pl ay houses and a maze. Th e scope of many of these items
is to assist children to deve lop physical coordination,
strength, and flexibility, as well as providing recreation
and enjoyment. Common in modern playgrounds are 'play
structures' that li nk many different pieces of eq uipment.
A little more than one-half of the children injured are
boys, with the largest percentage of injured children in the
age bracket of between 5 and 9 years 0ld s2
Protective surfaces (suc h as rubber mats, wood mulch,
bark or soft sand) of sufficient depth should be installed
und er and aro und playground equipment; 53 there is no
place for concrete, asphalted s u rface or sto ne slabs. 54 .55
This, togeth er with t he removal of monkey bars from parks,
resu lts in a significant reduction of inju ry rate. 56
Equipment must be insta lled properly, with sufficient
space a round it. Pl aygro und eq uipment needs to be
inspected periodically for structural defects, peeling paint or
splinterin g wood a nd these should be repa ired. Protective
surfaces under playground equipment should also be
inspected periodically, as surfaces such as wood chips or
sand lose their protective quality if they get compacted by
regul ar usage or are worn awayS7
SPORTS INJURIES ON SNOW AND ICE
Areas with cold te mpera tures, snow and ice can make ideal
conditions for recreation, but sports on snow and ice create
specia l ha za rds for children. Over one-half of the children
with t his type of injury are injured while sledging; the next
largest groups have injuries caused by skiing, snowboard­
ing or playing hockey - fewest children were injured while
ice-skating.
Falling or hitting objects such as trees, rocks or fro ze n
ground injured most children. Head injury was the most
frequent diagnosis. One out of four children had broken or
fract ured bones or had multiple injuries or required sur­
gery. The average length of hospital stay was 6 days. More
than 14 days were spent in the hospital by on ly 10 pe r cent
of the children .
The combination of speed with a fall or coUision on ice
and snow increases the risk of serious injuries . Over one­
half of the children had head inj uries and few, if any, of
these children had been wearing helmets.
Data from a study of sk i rental fac iliti es in 1999 looking
at 353 rental shops in five European countries (France,
Austria, Germany, Sw itzerl and and Ita ly) had shown that
71 per cent failed to meet the sta ndard and abo ut 13 per
cent just about reached an acceptable level. In tota l, 70 per
cent of rental facilities still set the bindings incorrectly ­
66 per cent of bindings in children were found to have been
set dangerously hi gh. 58 - 60 At the biannual [SSS sy mposium
held in Japan in 200 5, injury statistics presented from
Scotland, Fra nce and Iran show a rate'expressed as injuries
per 1000 part icip ant days of 1-2 per 1000 fo r alpine skiing

and 3-4 per 1000 for snowboarding and ski-boarding.
More common injuries are knee injuries in alpine skiers,
wrist injuries in snowboarders and lower leg injuries in
ski-boarders. Children as a group are twice as likely to be
injured as adults, especially if using rental gear and on
their first day of skiing as were those who had received less
than five formal lessons and those skiing with friends.
Children were shown to be more frequently injured on lift
systems and as a result of bindings failing to release.
In relation to the use of protective equipment by skiers
the debate about helmet use continues. In Norway, the rates
of use amongst those injured were 31 per cent (alpine ski),
26 per cent (snow board) 26 per cent (telemark ski) and 28 per
cent (ski-board). In Sweden, where about 58 per cent of all
people on the slopes now wear a helmet, a 28 per cent rela­
tive increase in the risk of a head injury was observed in
non-helmet wearers. The main protective effect of headgear
is against more minor injmies such as abrasions and bruises.
The helmets in use at present simply cannot do much in
high-speed impact accidents. Wrist guards were devised to
reduce the incidence of wrist fractures amongst snowboard­
ers but there are currently no international standards for
them, so it is velY difficult for a boarder to decide which to
use and which is the best buy. Release bindings had led to a
reduction in the number of lower leg fractures and may
reduce the number of knee injuries that were still occurring.
International standards have still to be agreed for them.
RIDING INJURIES
Each year children are injured and some die as a result of
incidents while riding a horse. An adult horse can weigh
over 500 kg, gallops at speeds up to 65 km per hour and
kicks with the force of 1.8 times its weight. About 65-75
per cent 51 - 54 of children admitted after riding accidents
have been riding the horse, with most injuries occurring
to the head and in the limbs. Many of the fatalities and
injuries can be avoided or have a reduction their severity
through a combination of better adult supervision and the
use of appropriate protective helmets. 55 .55
AGRICULTURAL INJURIES
Children in farming communities often assist their families
during peak harvest times and indeed may seek employ­
ment as fruit and vegetable pickers during vacation peri­
ods. In the course of holidays on farms, children may also
find themselves in trouble with farming equipment. This
problem is well recognized in Britain as it is in Australia 57
and in the USA;68 between 1979 and 1988, 167 children
died in farming accidents in Britain and about 300 children
die annually in the USA.69
Tractor-related injuries are the most common in
Britain,l° even although there is legislation in existence
Older children and substance abuse I 341
forbidding children below the age of 13 years to drive a
tractor on private ground. Children are also prevented
by law from riding as passengers on farming machinery. This
applies not only tractors, but also to harvesters and trailers.
Quad bikes are quite popular on farms and are often an
irresistible attraction to children; crashes with these have
resu Ited in injury and deaths.
PREVENTION
A major public health issue 71 is the effective prevention of
injuries and fatalities from injuries in children and, in most
countries, there is a constant active monitoring of the
incidence of serious injury and fatalities resulting from
accidents; any patterns that emerge from such studies are
addressed and this occasionally involves new Jegisla­
tion. 55 ,55.72 Preventative measures that are adopted are reg­
ularly audited in terms of their effectiveness and fine tuned
as required.73
The availability of guidelines and legal sanction have
enabled the commencement of, and the frequent successful
completion of, litigation procedures in the civil and, more
rarely, the chimerical courts, in addition to public inquests
and fatal accident inquiries.
OLDER CHILDREN AND SUBSTANCE ABUSE
In the older age groups, alcohol use, solvent misuse 74 and
drugs may creep into the picture; in addition young adoles­
cents may carry weapons as part of gang culture or for
alleged personal protection from bullying and other preda­
tory behaviour. 75 In other countries firearms may be used
in much the same way and because of easy access to them,
injuries and deaths often result from them.
More than one person per week dies from volatile sol­
vent abuse (VSA). The average home has 30 kitchen and
bathroom products that can be abused, including: butane
gas cigarette lighter refills, liquefied domestic gas, solvent­
based adhesives, deodorant aerosols, pain relief sprays,
aerosol air fresheners, hairspray, other aerosols, correction
fl uids, petrol, certain paints, paint thinners and removers,
clry cleaning agents, petrol lighter fuel, nail varnish and
varnish remover, some shoe and metal polish, and plaster
remover. In 2001, VSA was responsible for seven times the
number of fatalities than those related to ecstasy (3,4 meth­
ylenedioxymethamphetamine [MDMA]). The 2003 annual
report from the European Union's drug agency warned that
VSA is an often overlooked problem, with a big impact on
public health of young people; the use of solvents and
inhalants is the next common substance misuse after
cannabis and alcohol in 15- to 16-year-olds. One in seven
15- and 16-year-olds in the UK sniffs solvents to get high.
More than one person dies every week in the UK from the
effects of solvent abuse - between 70 and 100 each year.
 342 I Accidental injuries in children
The dru gs charity SOLVE IT says so lven t abuse kills 60-75
youn g people each year. One- thi rd of young people that die
from VSA are first-time so lvent abusers; in 1997 in 37 per
cent of deaths there was no prior known history of abuse.
Most deaths from VSA occurred in the 14-18 age ran ge .
Solvent abusers can be male or female, al thou gh there are
higher numbers of solvent-related de aths in boys, probab ly
due to differences in sniffing behaviour.
Sniffing so lvents may cause intox ication similar to the
effects of alcoho l,76 so a so lvent sniffer may become
drow sy, confu sed, an d aggressive, may take more risks
than they would when sober, a nd so on. Accidents, there­
fore, are quite comm on and so metimes fatal. 77 Over one­
half of the deaths that have been linked to solvent sniffing
appear to resu lt from the direct toxic effects of t he chemi­
cals that w ere sniffed , but other de aths result from acci­
dents, choking on vomit or suffocati on. Deat hs are often
sudden, and often a mechanism of death invo lv ing cardiac
arrest app ears to be the cause . Gas fuels continue to be
associated with the majori ty of deaths, accountin g for
about two-thirds of VSA deaths. Sniffing the butan e gas
in li ghters causes a cardiac arrhythmia. 7s .79 Possession or
abuse of a vo latile substance is not a c riminal act in itself,
although shopkeepers can be fi ned under the Intoxicating
Substa nces (Supply) Act 1985 if they knowingly sell to an
abuser who is und er 18 year old. The investigations of such
deaths require carefu l sampling; to enabl e later sampling of
the air wi thi n the low er airways, it is essential to retain a
lun g, with its main bronchus tied up securel y.8o.sl
ACCIDENTAL POISONING
The innate inquisitiven ess of children and th eir tendency to
sample and taste wha t th ey encoun ter in t heir immediate
environm ent may result in accidental poisoning by: con­
sumption of mushrooms, benies, fruits and vegetation grow­
ing wild, which are intrinsically poiso no us; and the
sampling of liquids held in bottles, including alcoholic bev­
erages, or other containers of liquids found in the house or
in ga rages and outho uses. They are attrac ted by the colour,
odour and labels of bottles and contain ers. Poisons range
from med icines (such as a nalgesics, contraceptive pills, iron
tab lets, vitamins an d antidepressants) to com mon fluids or
items foun d in the ho use (cleaning fluids, a lco hol, cigarettes,
c ray ons and button batteries) to dangerous garden plants
(fox glove and lab urnum seeds). Samplin g by children of
medication not intended fo r th em is a recurring prob lem,
particularly if the tablets and capsules in questions are
coloured a nd resem ble sweets. Poiso ning is the fourth most
co mmon ca use of accidental deaths in children.
Children under 5 years of age, as well as ado lescents, are
pro ne to poisoning but accidental ingestions are most com­
mon in ch ildren under 5 years old . Abo ut 90 per cent of a ll
poisonings occur at ho me, the most common sites being
the ki tche n and the bathroom . Unintentional pOisonings
- --
-
occur most frequ ently whe n routines are disrup ted, for
exa mpl e moving and vacations. Chi ld safety caps have
help ed decrease the number of p oisonings but they are not
100 per cent effective and should not give a fals e sense of
security.82,83 It goes without sayin g that a ll potential poi­
sons should be prop erly lab ell ed, stored out of reach of
children and locked away.
In a household where drugs are misused it is po ss ible for
a child , even a tod dler, to gain access to liquids, su ch as
methadone, and co ntrolled substances of abuse, such as
heroin or cocain e powder, which they inges t with the con­
sequence of serious harm as a direct effect of such inges­
tion. Passive smoking of cannabis fr om parents smoking
close to the child in enclosed restricted environm ents has
been known to result in inhal at ion of enough smoke to
result in intoxication of the child .
HYPERSENSITIVITY
Allergies with the potential for a fatal hypersensitivi ty
respo nse have become more freq uent over the years S4 - at
least in the public perception. According to the Natio nal
Institute of Allergy, food allergy occurs in 6-8 per cent of
children und er the age of 4 years, as well as 4 per cent of
adults. About 30000 episodes of anaphylaxis and 100-200
deaths per year occur in the USA. At present, the only ways
to manage food allergies are to avoid the food s that cause
reactions and to treat the sympto ms caused by allergic reac­
tions. About 20 per cent of th e population worldwide can be
considered to have som e form of atopy; this is more common
in infants and children und er the age of 3 yea rs. B5 ,S6 Nut
allergy is parti cularly important;S7 the inadvertent consump­
tio n of nuts by children who happen to be sensitive to them
may cause deaths unless immediate action is taken. In the UK,
5 out of the 20 fatal a naphylactic re actions reco rded each
year are due to food .s8 In addition to peanuts, the foll owi ng
fo ods can result in life-threaten ing reac tions on re-exposure:
tree nuts, seafood, seeds and cows' milk. 89 - 93 A number of
these children always cany with th em an emergen cy pack
that contains t he equipment required for an adrenalin injec­
tion, along with instructions as to how it shou ld be used if
they were to develop an acute hypersensitivity respo nse;94
these are carried to school or other extra-domestic sites. 95
Children with bronchial asthma and wh o are kn ow n to
respo nd adv ersely to external allergens, su ch as cats, may
also develop potentially life-t hreateni ng episodes if exposed
acutely to t he antigen .
REFERENCES
Macke nzie SG, Pless lB. CH1RPP: Canada 's princ ipal injury
surveillance program .il1jul)' Pre/! 1999; 5: 208- 13.
2 Stone DHG , Morrison A. Smith GS. Emergen cy Depart ment
inju ry surveill ance system : the best use of resources? inju1Y
Prel) 199 9; 5: J66 -7.
-

3 ECOHOST. Child Injuries in Industrialised Countries. Report
commissioned by UNICEF. London: ECOHOST, 2000.
4 Malek M, Chang BH, GalJagher SS, Guyer B. The cost of
medical care for injuri es to children. Ann Emerg Med 1991;
20:997-1005.
5 Barber MA, Sibert JR. Diagnosing physical child abuse: the
way forward. Postgrad Med ) 2000; 76:743-9.
6 Moyer LA, Boyle CA, PoJJack DA, Validity of death certificates
for injury-related causes of death. Am) Epidel11iol 1989 ;
130: 1024-3 2.
30 Kendrick D, Marsh P. Babywalkers: prevalence of use and
7
Maclntryre CR, Ackland MJ, Chandra raj EJ. Accuracy of
injury coding in Victorian hospital morbidity data. Ausr NZ )
PubI Health 1997; 2:779-83.
8 Walton W, Jarvi s SN, Lowe P Jet al. Annual incidence of
unintentional injury among 54,000 children. Injury Prell
1996; 2:16- 20.
9 Towner EM, Jarvis SN, Walsh SS, Aynsley-Green A.
Child Pedrstrian Accidents: a Research Revirw. Project Report
Measuring exposure to injury risk in school children aged
116 Et 117. Crowthorn e, Berkshire: Transport Research
11-14. Bi\lIJ 1994; 308:449 - 52.
Laboratory, 1995.
10 Levene S. Coron ers' records of accidental deaths. Arch Dis
Child 1991; 66:1239-41.
II Murdock R, Eva J. Home accidents to children under 15 years:
a survey of91O cases. BM) 1974; 3:103-6.
12 Lenaway DL, Ambler AG, Beaudoin DE. The epidemiology of
school-related injuries: new perspectives. Am) Prep Med
1992 ; 8: 193-8.
JJ Latlamme L, Menckel E. School injuries in an occupational
health perspective: w hat do we learn from community based
epidemiological studies? Injul:V Preu 1997; 3:50-6.
14 Hammarstrom A, Janlert U. Epidemiology of school injuries in
the northern part of Sweden. scal1 ) Soc IvIed 1994;
22: 120-6.
15 Alwash R, McCarthy M. Accidents in the home among
children under 5: ethnic differences or social disadvantage?
BM) 1988; 296: 1450-3.
16 AvelY lG, Vaudin IN , Fletcher JL, Watson 1M. Geographical
and social variation s in mortality due to childhood accidents
in England and Wales 1975-84. Publ Health 1990 ;
104: 17I -82.
17 Campbell H, O'Dri sco ll S. Th e ep id e miology of leisure
accidents in Scotland. Health Bulletin 1995 ; 53:280-3.
18 Hapgood R, Kendrick D, Marsh P. How well do
soc iodemogTaphic characteristics explain variation in childhood
safety practices? ) Publ Health IvIed 2000; 11 3 :285-9.
19 David RM, Pless LB. BMJ bans 'accidents '. Accidents are not
unpredictable. BM)2001; 332:1321-2.
20 Guyer B, Gallagher S, Chang Bet al. Prevention of childhood
injuries: evaluation of the Statewide Childhood Injury
Prevention Program (SCIPP). Am) Publ Health 1989;
79: 1521-7.
21 Gielen A, Wilson M, McDonald E et al. Randornised trial of
enhanced anticipatory guidance for injury pTevention. Arch
Pediarr Adolesc Med 200 1; 155:42-9.
22 Botting B, Crawley R. Trends and patterns in childhood
mortality and morbidity. In Botting B (ed,) Tile Health oj Our
Children: a reuiew in the mid-/990s. Decennial supple ment
OPCS DS No. 11,62-81. London: Office of Population
Censuses and Surveys, 1995.
23 King A, Wold B, Tudor Smith C, Harel Y. The Health oj Youth;
a Cross-national sumey. WHO regional Publications,
European Series No. 69, WHO, 1966.
24 Bijur P, Golding J, Haslum M, Kurzon M. Behavioural
predictors of injury in school-age children. Am ) Dis Child
1988 ; 142: 1307-12.
25 Waller AI, Marshall sw. Childhood thermal injuries in New
Zealand resulting in death and hospita lization. Burns 1993 ;
19:371-6.
References I 343
26 Mott A, Evans R, Rolfe K et al. Patterns of injuries to children
on public playgrounds. Arch Dis Child 1994: 71 :3 28 -30.
27 Kujala UM, Taimela S, Antti-Poika I et al. Acute injurie s in
soccer, ice hockey, volleyball, basketball, judo and karate:
analysis of national registry data. BM) 1995; 311: 1465-8.
28 Swan P. British Schools, Safety Incidents and The Courts.
www.articlecity.com/artic les/lega l/ a rticl e_3 8 5.sh tml.
29 Lane K. General Defences Relevant to the Tort of Negligence.
www.city.londonmet.ac.uk/-shlane/Generaldefences.htm.
relationship with other safety practices. Inju/y Prev 1998;
4:295-8.
31 Spiegel C, Lindaman F. Children can 't tly: a program to
prevent childhood morbidity and mortality from window falls.
Am) Publ Hralrh 1977; 67:1143-7.
32 Christie N. Social, Economic and Environmental Factors in
33 Lawson SD, Edwards PJ. The involvement of ethnic minoriti es
in road accidents: data from three studies of young pedestrian
casualties. Traffic Eng COl1trol 1991 ; 32: 12-19.
34 Boxall JA. School crossing patrols: how effecti ve are they?
Traffic Eng Control J 938; 29:586. Report CR 192. Crowthorne,
Berkshire: Transport and London: Departme nt of TranspOI1.
35 Christie N. Social Economi c and Environmen tal Factors in
Child Pedestrian Accidents: A Research Review. Project Report
no. 116. Craw thorne, Berkshire: Road Research Laboratory,
1995.
36 Jorgensen E. Bicycle tracks in urban areas in Denmark.
Evaluation of the effect on safety. In Biecheler M, Lacombe C,
Muhlrad N (eels) Evaluatiol1 85: International Meeting on the
Evaluation of Local Traffic Safety Measures, Proceedings of
the Paris Conference, 20-23 May 198 5, pp. 755-61.
37 Bicycle Helmet Safety Institute, 199 5.
http: //www. bhsi.org/webdocs/hend erson. htm.
38 Cameron M, Heiman L, Neiger D. Evaluation of the Bicycle
Helmet Wearing Law in Victoria During its First 12 Months.
Report No. 32. Victoria, Australia: Accident Research Centre,
Monash Unive rsity, 1992.
39 Vulcan A, Cameron M, Watson W. Mandatory bicycle helmet
use: expe rience in Victoria, Australia. World) surg 1992;
16:389-97.
40 Finch C, Newstead S, Cameron M, Vulcan A. Head Injury
Reductions in Victoria Two Years After Introduction of
MandatolY Bicycle Helmet Use. Report No. 51. Victoria ,
Australia: Accident Research Centre, Monash University, 1993.
41 Cameron M, Vulcan A, Finch C, Newstead S. Mandatory
bicycle helmet use following a decade of helmet promotion in
Victoria, Australia - an evaluation. Accident Anal Prey 1994;
26:325- 37.
42 McDermott F. Bicyclist head injulY prevention by hel mets and
man datory wearing legislation in Victoria, Australia. Ann R C
surg Engl 1995; 77:38-44.
43 Puczynski M, Marshall D. Helmets! All the pros wear them.
Am) Dis Child 1992; 146:1465-7.
44 Lee A, Mann N, Takliti R. A hospital - led promotion campaign
aimed to increase bicycle helmet-wearing among children
aged 11-15 living in West Berkshire 1992-98. Injury Prev
2000; 6:151-3.
45 Britt J, Silver I, Rivara F. Bicycle helm et promotion among
low income preschool children. Injurv Prey 1998 ; 4:230-3.
46 Abu larrage J, De Luca A. Abularrage C. Effect of educat ion
and legislat ion on bicycle helmet use in a multiracial
population. Arch Pediatr Adolesc Med 1997 ; 151 :4 1- 4.
47 Dannenberg A, Gielen A, Beilenson Pet al. Bicycle helmet
la ws and educat ional campaigns: an evaluation of strategies
344 I Accidental injuries in children
to increase c hil dren's helmet use. Am} Publ Healt/r 1993;
83:667-74.
48 MacKay M. P layground injuries. Injury Prev 2003; 9: 194 -6.
49 Cha lm ers DJ , Ma rshall SW, Lan g ley JD et al. Heig ht a nd
surfacing as risk factors for injury in fa ll s from playground
equipment: a case co ntrol study. InjU/)' Prev 1996; 2:98 -1 04.
50 Laforest S, Robaitaille, Lesage D et al. Surfa ce charac teristi cs,
equipment height, and the occurrence and severity of
playground injuries. Injury Preli 2001; 7:35-40.
51 Platt SL, Fine JS, Foltin GL. Escalator-related injuries in
children. Pediatrics 19 97: 100 :E 2.
52 Lilli s KA, Jaffe DM. Playgrou nd injuries in children. Pediatr
Emerg Care 1997; 13: 149-53.
53 Laforest S, Robita ill e Y, Dorval D et al. Severity of fall injuri es
on sand or grass in playg rounds. } Epidemiol Community
Health 2000; 54:475- 7.
54 Ro sevea re CA, Brown JM, Barclay Mcintosh JM, Chalm ers DJ.
An intervent ion to red uce playground equipment hazard s.
injury Prev 1999; 5: 124-8 .
55 Sosin D, Su rface-specific fall injury rates on Utah school
playgrou nd s. Am} Publ Health 1993; 83:733-5.
56 Mott A. Rolfe K, Ja mes R et al. Safety of su rfa ces and
eq uipment for c hildre n in playgro und s. Lancet 199 7:
349 : 1874- 6.
57 Sub ve rt JR, Mott A, Rolfe K et a l. Preventing inj uries in
public playg round s t hro ug h partnership between health
services a nd loca l a uth ority: co mmuni ty interwoven study.
BM} 1999; 318:159 5-8.
58 Mayr J , Russel 0, Sp itzer P et al. Playground accidents. Acta
Paediatr 1995; 84:573-6.
59 Macar1hur C, Hu X, Wesson DE, Parkin Pc. Risk factors for
severe injuries associated with falls from playground
equipment. Accident Anal Prev 2000; 32:377-8 2.
60 Langran MA. Summa ry of the research presented at the 16th
In ternationa l Sy mposium of the ISSS held in at Mount Arai,
Japan, Ap ril 2005. www.ski-injury.com.
61 Barone GW, Rodgers BM. Pediatrics equestrian injuri es :
a 14-yea r rev iew.} Trauma 1989; 29:245-7.
62 Bond GR, Christoph RA, Rodgers BM. Pedi atric eq uestrian
injuri es; assessing the impact of helmet use. Pediatrics 1995;
95:487-9.
63 Holland AJA, Roy V, Goh V et al. Horse-related injuries in
chil dren. Med} Aust 2001; 1756 ;609-11.
64 Ghosh A, Di Scala C, Drew C et al. Horse-rel ated injuries in
pediatric patients.} Pediatr Surg 2000 ; 35: 1766 - 9 .
65 Aronso n H, Tough Sc. Horse-rel ated fatalities in the p rovince
of Alberta 1975-1 980. Am} ForC'l1sic fVlt:'d Pat/lOl 1993;
14:28-30.
66 Ch itn av is JP, Gibbins CLMH, Hirigoyen M et al. Accidents
with horses. What ha s cha nged in 20 yea rs? Inj!lI)' 1996;
27: 103 - 5.
67 Byard RW, Gilbert J, Lipset J , James R. Farm - and tractor­
related fatalities in South Australia. } PC'diarr Child HC'olrh
1998; 32:139-41.
68 Cogb ill TH, Busch HM, Stiers GR. Fa rm accidents in ch ildren.
Pediatrics 1986; 76:562-6.
69 Rivara FP. Fatal and non-fata l farm injuri es in the United
States. Pediatrics 1985; 76:5 67 - 73.
70 Cameron D, Bishop C, Sibert JR. Farm accidents in children.
EM} 1992; 305:23-5.
71 Millward LM , Morgan A, Kelly MP. NH S Hea lth Development
Agency. Health Developm ent Agency. Prevention and
Reduction of Accidental Injury in Children and Older People.
Evidence briefing, June 2003.
72 Grayling T, Hallam K, Graham D et al. Streets Ahead: SaJe
and Liveab/{:, Streets Jor' Children. London: In stitute for Publi c
Policy Research, 2002.
- --=- -
73 Town er E, Dowswell T, Mackereth C, Jarvis S. What Works in
Preventing Uninte ntional Injuri es in Chi ldren and Young
Ado lescents 7 An Updated Systemat ic Review. London:
Hea lth Develop ment Agency, 2001.
74 Sva nstro m L, Haglun d BJA. Evidence-based Safety Promotion
and Injury Prevention - a n Introducti o n. Department of
Publ ic Hea lth Sciences, Di vision of Soc ial Medici ne,
Karolin ska lnsti tutet, Stockho lm , Sweden, 2000.
7 5 Ano nymous. Solvent abuse: Little progress a fter 20 years
(Editori al). 8M} 1990: i:J00.
76 McKeganey N, Norrie J. Association between illegal drugs a nd
weapon carry in g in young people in Scotland; schools'
survey. EM} 2000: 320:982-4.
77 Joseph DE, Parke r S. inl7alants. Drugs oj Abuse. New York:
United States Drug Enforcement Administration , 2005.
78 First Nations and Inuit Health Committee, Canadian Paediatric
Soc iety (CPS). Inhalant abuse. Paediatr Child Hea lth 1998;
3:123- 6.
79 Shep herd RT. Mechanism of sudden death associated with
volatile substance abuse. Hum Toxicol 1989; 8:28 7-91.
80 Anderson HR, MacNair RS , Ramsey JD. Deaths from abuse of
substances : a national epidemiological study. EM} 19 85 ;
290:304-7.
81 www.tox la b.co. uk/vo la tile.htm and
www.toxlab.co. uk/dasguide.htm.
82 S ibe rt JR , Clarke AJ, Mitc hell MP. Improvements in child
res ista nt co ntai ne rs. A rch Dis Child 1985; 60:1155 - 7.
83 Rodge rs G. The sa fety effects of child-res istant pack aging fo r
ora l prescripti on drugs : two decades of experie nce. }AMA
1986 ; 275: 166 1-5.
84 Isolauri E, Huurre A, Salminen S, Impivaa ra O. The all ergy
epide mic ex tends beyond th e past fe w decades . Ciil1 Exp
A llagy 2004; 34:1007-10.
85 Host A, Halken S. A prospective study of cow milk a llergy in
Danish infants during the first 3 years of life. Clinical course
in relation to clinical and immun o log ical type of
hypersensitivity reaction. Allagy 1990; 45 :5 87 - 96.
86 Sampson HA, Scanlon SM. Natural hi story of food
hypersensitivity in children with atop ic dermatiti s. } Pediarr
1989; 115:23-7.
87 Bock SA, Atkins FM. The natural histolY of peanut alJergy.
} Allng.l' Ciill illll11ullol 1989; 83:900-4.
88 Pumphrey RS. Lessons for management of anaphy laxis from a
study of fata l rea ctions. Clin Exp AI/{:'rgy 2000; 8: 1144-5 0.
89 Pa llares D. Allergy to macadamia nut. A 1111 Allergy Asthma
1111111ullol 2000; 85:385-6.
90 YUll gin ge r .liN, Sweeney KG, Sturn er WQ e t a l. Fata l food ­
induced a naphy laxis. }AMA 1988; 260: 1450-2.
9 1 Sa mpso n HA, Mend e lso n L, Rose n JP. Fatal a nd near-fatal
anaphy lac tic reaction s to food in chil dren a nd ado lescents.
N ElIgl} MC'd 1992; 32 7:330- 4.
92 Bock SA, Mun oz - Fu rl ong A, Sampson HA. Fata liti es due to
anaphylactic react io ns to foods. J Allergy Clin Immunol 200 1 ;
1: 191-3.
93 Yunginger JW, Nelson DR , Squi llace DL et aJ. Laboratory
investigation of deaths due to anaphylaxis. ) Forensic Sci 1991;
36:857-65.
94 Simons FE, Chan ES, Gu X, Simons KJ. Epinephrine for the
out-of-ho sp ital (first aid] treatment of anaphylaxis in infants:
is the ampOUl e, syringe, need le met hod practical? ) Allergy Clin
Immunol 200 1 ; 108: 1040-4.
95 Simons FE, Gu X, Silv er NA, Simo ns KJ. EpiPen Jr versus
EpiPen in young children weighing 15 to 30 kg at risk for
anaphylaxis. ) Allergy Clin Immunol 2002; 109: 171-5.
­

DROWNING AND NEAR DROWNING
John Pearn
troduction 345
e causes of childhood drowning: a perspective 345
e drowning medium 346
TRODUCTION
child abuse or non- accidenta l injury, homiCide, euthanasia
J:o wning remains one of the most common causes of death
:!l early chil dhood . In developed and affluent countries, after
?: child has passed th e neonatal period, immersion incidents
~main one of the three principal causes of death.l,2 In many
:<a es and commun ities within Europe, the USA, Canada and
.-. ~ tra la sia, drowning is the leading cause of unintentiona l
C:~th in children und er the age of 5 yea rs.) This age group
:.So ex hibits the highest drowning mortality rate worldwide,
=:.xcept in Africa. 4 Immersi on fatalities rank as the seco nd
~d ing cause of acc idental death for all older children and
-cenage rs and this is the third leading preventable cause of
:'eath.5 [n Australia, a water-orientated society, drowning
- . replaced motor vehicle accidents as the leading cause of
=.:.: childhood deaths from injury.6 Autopsy di agnos is of the
- owned child is challenging for the pathologist and is co n­
-:-onti ng for all who work proactively to redu ce the incidence
uC: im mersion accidents.
Every case of drowning and near-d rowning has poten­
forensic and legal implications. These app ly not on ly
- :he wo rk of pathologists, coroners, police officers an d
G.\. 'ers in the case of immersion fatalities, but also are
apa bl e as part of the clinical skills required by rescuers,
.'-- la nd ers, a mbulance officers, paramed ics a nd the emer­
=e cy room team. [n addition to the elucidation of the
;mato micopathological features of drowning, and the esta b­
!i hme nt of precedents and causes of death, the fore nsic
i sues of immersion incidents may also include those of
I CHAPTER 18 I
The pathophysiology of drowning 349
Forensic immersion syndromes 351
References 359
a nd culpable neglect. The demonstration a nd interpretation
of secondalY trauma due to attempted cardiopulmonary
resuscitation after immersion also depends upon meticu­
lous autopsy technique.
Of all health professionals, pathologists develop the most
pragmatic perspective of the serio us threats to children's
health, welfare and safety,7 Both the paediatric pathologist
and the forensic pathologist have inescapa ble ethical roles in
advocacy for the reduction of immersion accidents, a signif­
icant proportion of which a re preventable.
THE CAUSES OF CHILDHOOD DROWNING:
A PERSPECTIVE
Childhood drowning ra n ks high as a cause of child death,
especially in the pre-school years (Table 18.1). The sites of
the drowning incid ent include the family bath tub, private
and public toilets, buckets and pails, private, family-owned
swimming pools, public swi mmin g pools, fish ponds and
ornamental pools, building site tre nches and drains, agri­
cultural dams and tanks, resi dential and neighbourhood
canals, lakes, creeks and riv e rs, boating marinas and the
open sears) (Table 18.2). Children who drown in each of
these sites have their own s ite-specific precedents a nd
forensic concomitants B ,9
The causes, sites, survi val rates and modus operandi of
immersion deaths all differ when child victims are compared
346 I Drowning and near drowning
Table 18.1 Relative rank order (per cent) of traumatic and injury
deaths in childhood (1-14 years)
Cause of death Percentage
Drowning (accidental) 32
Motor vehicle occupant death s 29
Pedestrian run downs 17
House/ca rava n fires 8 when the duration of immersion is measured in minutes only.
Homicide, child abuse 6 The second tool, indispensable to the pathologist, is ref­
Fa lis, playg round acciden ts, accidenta I 6 erence to the defined syndromic profiles of child drown­
strangulati on
Other, e.g. poisoning, electrocution 2 distinct child immersion syndromes. lJ Such form indispen­
Total 700
Data typical of tropica l and temperate developed nations. Comp il ed from
the Austra lian Insti tute of Health and We lfare 6 and other sources.
Table 18.2 Relative rank order (per cent) ofsites of drowning of
children aged 0-5 years
Site Percentage
Private swimm ing pools 64
Family bath tubs 16
Creeks 11
Dams, building trenches, sewers 5 subsequent proceedings.
Waterholes, fish ponds 4 Many childhood immersion fata lities are depressingly
Data typical of tropical and temperate developed nations. Compiled from
the Brisbane Drowning Stud y l 7.43 and other sources.
with adu lt subjects. The common patterns of adult immersion
incidents, involving alcohol, suicide lO and boating accidents,
are rarely encountered in the case of childhood immersion
victims. The common situation of difficulty with cadaver
id entification often encountered in long-immersed adult vic­
tims is very rare indeed; the identity of childhood victims is
usua lly known and not open to question at the time of
extraction from the water. Immersion times of childhood vic­
tims are measured in minutes rather than hours in more than
99 per cent of repo rted series. 11 The challenge of child immer­
sion deaths to the forens ic pathologist is that every case is
individual and specific. Yet, however rare might be the cir­
cumstances sunounding an individual traumatic child death,
diagnosis as to the cause of death and elucidation and recon­
struction of its precedents are of the greatest impoliance.
In approaching a case of immersion death or near-death
in childhood, the pathologist needs to adopt two appro aches.
The first is that of the usual detailed history-taking, followed
by meticulous autopsy examination of the body with, in
some instances, particular attention to post-mOliem radi­
ographic survey, blood analyses for drugs, chemical analyses
of lung tissue, DNA extraction and storage, post-mo rtem
photography and, rarely, bone marrow extraction for diatom
examination (Table 18.3). The diagnosis of drowning is one of
the most difficult in forensic pathology 12 In the case of child
victims this difficulty is compounded when there may be no
pathognomonic signs of drowning and where there may be
no objective post-mOliem signs of intentional injury, even
when such is undoubtedly the case. Alveolar oedema, alveo­
lar haemorrhages and emphysema aquosum are often the
only histological featu res of drowning; even these may not
be demonstrable in the case of an infant or child drowning
ings. Much recent research has built up patterns of quite
sable templates not only for the pathologist, but also for
the investigating police and presiding coroner (Table 18.4).
Childhood immersion fatalities do not form a spectrum of
immersion in cidents; rather they constitute subsets of quite
distinct forensic syndromes that are unrelated to each other
except by the end-point of drowning. Overall, 95 per cent
of such cases are accidental (unintentional). In homicide
cases, however, drowning is the method of killing in some
10-20 per cent (see Table 18.4). Reference to such immer­
sion syndromes in the pre-autopsy analysis of any individ­
ual case is of great help to the pathologist in the choice
of ancillary investigations (Table 18.3) and may be of the
greatest help to family doctors, coroners and lawyers in
repetitive as many of these deaths are preventable. Never­
theless, all of those persons who are involved in the pre­
mortuary chain - parents, siblings, neighbours, rescuers,
bystanders, paramedics and police officers - are inevitably
enmeshed in an intense personal and tragic incident. What
happens to such individual s often depends on the determi­
nation of diagnostic truth, and such in turn depends on the
patholo gist's skill in recreating the events surrounding
such immersion deaths. Because of the well-recognized, and
indeed historic, difficulty in confirming death by drown­
ing, there exists no greater challenge to the attending
pathologist.
THE DROWNING MEDIUM
In at least 90 per cent of drowning fatalities, death is due to
the inhalation of water into the alveoli, with hypoxaemia
and subsequent brain hypoxia. The water in which children
drown contains a case-specific mixture of inorganic sub­
stances, gases, dissolved organic matter and particulate mat­
ter. Particulate matter may include bacteria, fung i, diatoms,
zooplankton and phytoplankton. Unlike the situation that
is more frequently encountered in the forensic analysis of
adult immersion victims, such water composition does not
usually have clinical, forensic or legal overtones. Very rarely
children drown in vessels t1lled with paint or fertilizer, or in
agricultural, industrial or domestic chemicals.
 The drowning medium I 347

Table 18.3 Forensic investigotions apposite to immersion victims

Radiology Total body X-rays essential after immersion death

Skeletal survey usually normal in private swimming pool drownings. Exclude cervical spine fracture-dislocations
in deaths in public swimming pools, in sea or surf, in diving or suspected diving activity
In bath tub drownings, skeletal survey excludes old fractures
In unidentified/decomposed cadavers, dental pantograms are useful
Histology Histology: of lung, brain and heart is essential, sampling of other organs advisable
Lung: vascular congestion, oedema, alveolar macrophages, alveolar haemorrhage, emphysema aquosum;73 foreign
material, including stomach contents in larger airways or bronchioles

Neck muscles: intramuscular haemorrhages 74

Brain: vascular congestion, microhaemorrhages; immunohistochemical changes in hypog lossal nuclei 22

Electran microscopy: not required routinel y
Alcohol /drug assays Blood alcohol: measure in all immersion victims over the age of 10 years; 60 per cent of adult male drownings
have raised blood alcohol levels
Antiepileptic drug levels when prescribed
Screen for narcotics, psychotropic drugs, amphetamines and other drugs of abuse in all teenage immersion deaths
Urine screen: gas chromatography-mass spectrometry (GC-MS); detects prescribed medications, drugs of abuse,
'over-the-counter' drugs, dietary supplements 75
Serum assay: quantitative estimation of anticonvulsants, other prescribed medication and drugs of abuse using
high-performance liquid chromatography assay [HPLCF 5
DNA extraction Blood or tissue sampling for DNA identification not routinely required in most paediatric drownings; exceptions
are suspected neonaticide, bath tub drownings, suspec ted child homicide and when maceration makes visual
identification impossible
Electrolytes Measurement of electrolytes rarel y helpful
Serum K may be increased due to haemolysis
Serum Na may be elevated due to inhaled or swallowed sea wa ter, or pre- or post-mortem dehydration
Vitreous electrolytes are stable for some hours after death and may not reflect agonal serum changes
Other biochemical Glucose: not routine; hypoglycaemia may be related to alcoholic intoxication, exhaustion by physical
studies activity or prolonged pre-death immersion in cold water/ 6.n vitreous glucose is more stable
Glucose in serum and vitreous should be measured in known diabetics
Urinary myoglobin: not routine, normal level < 50 ng/mL; can be very high in drowning but non-specific;
elevated levels occur in fatal burns, heat stroke, asphyxia and drowning; may indicate exertional

mu sc le activity or post-mortem change 78

Diatoms Rarely needed; many different diatom genera and species have siliceous skeleton; femoral bone marrow sampled;
tissue acid-digested or macerated, interpretation difficult, diatoms absorbed from the gastrointestinal tract
pre-mortem; water from the putative drown ing site essential for interpretation; concordance of diatom
genera and species is only 65 per cent 12
Diatom analysis is helpful when: cadaver is badly decomposed; body is found on land but suspicion of prior
drowni ng; bod y may have been put into the wa ter after death 12.79

Table 18.4 Relative rank order (per cent) of the modus of child Fresh Water
homicide in the USA in children under the age of 5 years

Modus Percentage Fresh water contains variable amounts of organic material,

dissolved salts and free and nascent gases. Most young
children who drown do so in chlorinated, freshwater swim­
Head injury 45
ming po ols; or in bath tubs with variable concentrations of
Drowning 23 dissolved soap or bath salts. 14
Non-drowning asphyxia 13 The chemical affects of chlorine and soap in fresh water
Body trauma 12 are not of consequence in the pathophysiology of the great
Other, e.g. neglect, poisoning 7 majority of fatal immers ions, and are certainly not of
From ref. 9, w ith permission. any practical significance in the pathophysiology of lung
 348 I Drowning and near drowning
syndromes in more than 95 per cent of survivors. Experi­
mental animal studies have not demonstrated significant
effects on the surface tension of lung surfactant when
chlorinated water inhalation is compared with that of pure
fresh water. However, any type of fresh water inhalation
elevates the minima l surface tension of tracheal and lung
surfaces very significantly when compared with the affects
that follow sa lt water aspiration.
Sea Water
Sea water contains a mixture of inorganic sa lts. Concentra­
tions vary widely but the relative prop0l1ions of the differ­
ent types of dissolved salts remain surprisingly constant. The
total concentration of sea-water salts may change over rela­
tively small distances, a phenomenon seen particularly in
estuaries and even during tidal flow. II Sa linity is defined as
the total mass of disso lved solids (in grams) in one kilogram
of water. Such is expressed either as parts per thousand (ppt)
or in grams per kilogram (g/kg). In instances where children
drown in in land seas, canals, rivers or delta regions (pa11ic­
ularJy during floods), the salinity of the inha led water may
be very low, approach ing that of fresh water.
In organic sea water, a typical salt ion profile is:
• sodi um , 10.5 g/ kg;
• magnesium, 1.3 g/kg;
• calcium, 0.4 g/kg;
• chloride, 18.9 g/kg;
• sulphate, 2.6 g/kg.
Oceanic sea water thus typically contains 34.5 g/kg of
dissolved salts, of wh ich 29.5 g/kg is sodium chloride; this
is equ iva lent to 2.9 per cent sodium ch loride compared
with 0.87 per cent sodium chloride that is the concentra­
tion in human plasma.
Irrespective of the salt content or osmolality of the inhaled
water, during the drownin g ep isode there is a one-way, mas­
sive transuda tion of water across the lung membranes into
the pu lmonary vasculature. Childhood drowning deaths are
exclusively due to cerebral hypoxia fo llowed by acute brain
death. Any acute changes in osmo la lity, consequent upon salt
concentration differences of the inhaled water, are not of
forensic relevance. In survivors, because of rapid compensa­
tory homeostatic changes - Starling's Law, the Bainbridge
Reflex and renal clearance - such diffe rences are not of prac­
tical relevance in the clinical management of near-drowned
victims in the emergency room or intensive care ward. This is
not to say that post-mortem electrolytes and osmolality
should not be measured, as the issue continues to be a focus
of legal debate in subsequent cou11 hearings.
How much fluid is needed to drown a child remains
unknown. Depth is no guide, as some infants drown in less
than 20 cm of water in ornamenta l pools, garden fish ponds,
pails or the family bath tub. Drug-intoxicated teenagers,
patients with epilepsy and children with spinal or head
--- - ----
injUlies (sometimes sustained dUling skylarking in the sur­
rounds of swimming pools) can drown in very shallow water.
In experimental mammals, as little as 1.0 mL/kg of fresh
water instilled into the trachea causes gross pathophysiolog­
ical responses in the lungs. Experimental studies, usin g dogs,
have shown that drownin g mammals may aspirate more than
22 mi/kg of water. The imp lication is that in the case of a
typ ical chi ldhood drowning (with a median age of approxi ­
mately 2 years in most current series) a \ 3-kg child may asp i­
rate as much as 300 mL of water. Such ac ute increase in blood
volume is not greater than the capacity of the nonnal healthy
child's heart to compensate for this potential flu id overload.
'Dry' Drowning
There exists an historic debate abo ut the proportion of cases
of immersion accidents that are due to 'dry' drownings, in
which laryngea l spasm is tho ug ht to be the primary cause
of fata l hypoxia. Alt hough laryngeal spasm occurs to some
degree in every case of fluid inhalati on, water enters the
peripheral airways in at least 90 per cent of cases. In child­
hood, autopsy findin gs of 'dry drownings' are rare. Such
evidence is supported also by clinical experi ence in the man­
agement of near-drowned survivors, where in every case
there is evidence of water inhalation. I n adult series, the fre­
quency of 'd ty drowning' has been va ri ously estimated to
occur in 10 per cent to 15 per cent of cases. However, a
recent review of the original studies from which such con­
clusions have been historically drawn has indicated that
such may be without firm foundation 15 and that card iac
arrest, sustain ed coincidentally while the victim is in the
water may be the true cause of death rather than asphyxia.
Despite this, there is persuasive experimental evidence to
SUpp0l1 the phenomenon of 'dty drowning'. In experimenta l
drowning with diatom monitoring, using deliberate immer­
sion deaths of aquatic mammals such as mink, muskrat and
beaver, it has been shown that not all fully submerged ani­
mals inhale significant amounts of water into their lungs. 16
When small amou n ts of water enter the larynx or trachea,
immediate laryngeal spasm results due to an efferent vagal
reflex. An immed iate outpouring of thick mucus occurs,
probably while the drowning victim is still conscious. Froth
develops and in some cases a physical mucous plug forms.
When such spasm relaxes preterminally, a significant
amount of water is thus prevented from enterin g the trachea
and lungs by the foam-froth plug, which acts as a physical
barrier. In such cases, loss of consciousness is caused by
anoxia or carbon dioxide narcosis. Death fo llows as a result
of cerebral anoxia.
Water Temperature
Almost all child ren who drown do so in water in a temper­
ature range of 10-25°(, Water temperatures ab ove 20'C do
~

not influence the brain-protectin g diving reflex, but low
temperatures augment it. The hypothermic, brain-sparing
effects of near-drowning in very cold water are mediated
through physical chilling of the body core and through an
augmented diving reflex. In those situ ations when children
drown in cold water, phys ical chilling of the body core is
the result of not only the conduction and convection
effects of cold water on the child's skin and through the
involuntary inhalation of cold water, but also from swal­
lowing relatively large amounts of cold water. All children
extracted from the water in immersion incidents have a
reduced body temperature, including those who respond to
cardiopulmonary resuscitat ion and who may survive the
incident. In boating accidents, in accidents involving falls
tho ugh ice and in cold water accidents involving older
children who can swim, immersion hypothermia may pre­
cede primary drowning asphyxia. Clouding of conscious­
ness occurs in a drowning child w hen the core body
temperature falls below 36°C; and consciousness is lost
,-,,,hen core body temperature falls below 34°C.
Immersion Asphyxia Occurring Other than
in Water
Rarely, children drown in non-water media such as paint, liq­
uid fertilizer or industrial chemica ls. Some children die after
falling into wheat or grain silos, or are smothered in falls of
earth or sand while playing in 'cubby houses', excavated
·Wendy' houses, or following the collapse of tunnels exca­
vated into the sides of cliffs at the beach or beside rivers or
creeks. In such cases the asphyxiating medium behaves as if
it were a fluid. Inhaled paint or sand within the airways, for
example, may be a dramatic macroscopic feature at autopsy;
the foreign material may be demonstrated histologically
(Table 18.3). However, in a ll such cases the mechanisms of
anoxia are secondary to foreign substance inhal ation. The
mechanisms of oxygen exclusion (first with brain death and
then with rapid somatic death) are identical to those encoun­
tered in cases of fatal water immersions.
THE PATHOPHYSIOLOGY OF DROWNING
submerged infant or child holds his or her breath volun­
.J,.
tarily until the 'break point' is reached. The 'break point' is
determined by both hypercarbic and hypoxic drives. High
arterial carbon dioxide levels combined with falling oxy­
gen concentrations are synergistic. For example, the 'break
po int' occurs at PaC0 2 levels of below 55 mmHg if there is
asso ciated hypo x ia; however, they may occur at Pa0 2 lev­
els of below 100 mmHg if the PaC0 2 exceeds 45 mmHg. In
the case of older children, particularly those indulging in
competitive or peer-demonstrative exhibitionism in swi m­
ming pools, the practice of pre-immersion hyperventilation
is dangerous. Such underwater endurance games or dives
Pathophysiology of drowning I 349
may alter the physiological setting of the breath-h olding
'break point'. If the hyp ercarbic drive to breathe is lost,
unconsciousness from cerebral hypoxia may occur before
the hypoxia-induced 'break-point' is reached. This is the
pathophysiological mechanism leading to drownin g in
teenagers or adults who, during snorkelling or free diving,
attempt maximum endurance dives. 11
In all cases when the ' break point' is reached and the
first involuntary inspiration is made, arterial hypoxaemia
has already developed. Tissue hypoxia and acidosis have
commenced. Involuntary gasp ing then occurs in a sub ­
merged subject, who may stiU be conscious. Glottal spasm
occurs. Even before consciousness is lost, vomiting with
aspiration of gastric contents may occur. Consci ousness is
lost within 3 minutes of invol untary submersion; and the
electroencephalogram (EEG) becomes flat within 4.5 min­
utes. Breathing movements w ith fluid aspiration , and often
the aspiration of sand , mud, g ra vel or vegetable matter,
continues after the drowning v ictim has lost consciousness.
Respiratory arrest follows. Dysrhythmias may follow and
hypoperfusion with hypoxic blood leads to brain death.
During the march of intra-immersion hypoxic events,
intense autonomic, catecholamine-meditated blood redis­
tribution occurs. The spleens of drowned victims are rela­
tively bloodless, as one manifestation of the reflex
constriction of sp lanchnic (including splen ic) vessels.
The duration of brain hypoxia necessary to cause the
death of an otherwise fit and healthy infant or child obvi­
ously remains unknown. Extensive research from the Bris­
bane Drowning Study,Il.l7 using a 'bracket method' and
involving an ana lysis of hundreds of immersion fatalities
and survivors, has indicated that (I) children who are
immersed for 3 minutes or less are likely to survive an
immersion incid ents and (2) children are unlikely to
respond to card iopulmonary resuscitation if the immersion
time is lon ger than 10 minutes in the most common water
temperatures in which they drown (10-20°C).
The Diving Reflex
The brain-sparing diving reflex may be demonstrated in
virtually all children, dramatically so in the case of
neonates and yo un g infants. The afferent arm of this reflex
is the skin surface of the face supplied by the trigeminal
nerve. The efferent arc involves t he vagus nerve and the
autonomic nerve supply to blood vessels of the skin and all
internal body organs except those of the brain and heart. In
the case of a submerged child, the diving reflex is manifest
by rapid-onset bradycardia and the shunting of blood from
cutaneous and splanchnic vascul ar beds to the cerebral and
coronary circulations. 18 Blood pressure starts to rise imme­
diately. The diving reflex is augmen ted by catecholamine
release and is probably inhibited by obtundant drugs such
as alco hol. It has been estimated t ha t the di ving reflex may
provide a drowning infant or young child with an extended
 350 I Drowning and near drowning
'window of salvage' of perhaps 30 seconds of immersion
hypoxia.
Lung Pathology
As soon as water enters the lungs, a chain of pathophysio­
logical events occurs, as follows.
PERIPHERAL AIRWAY RESISTANCE INCREASES
Aspiration of even small amounts of water (1.0 mL/kg) is
followed by pulmonary vasoconstriction, with immediate
development of pulmonary hypertension due to parasympa­
thetic reflexes. The composition of the inhaled water is
important in this context. Vasoconstriction occurs, particu­
larly following aspiration of fresh water into the mammalian
lung; it occurs to a lesser extent in the case of sea water aspi­
ration, but does not occur with aspiration of amniotic fluid.
LARYNGEAL SPASM
Laryngeal spasm follows, with an immediate outpouring
of thick mucus, followed by froth development. The conse­
quent degree of airway obstruction probably varies from
subject to subject; the larynx relaxes preterminally.
REFLEX PULMONARY VASCULAR VASOCONSTRICTION
This phenomenon leads to immediate pulmonary hyperten­
sion. Intrapulmonary reflexes then cause shunting of blood
through non-ventilated areas of the lungs. Such shunts, com­
bined with surfactant loss or inactivation and consequent
alveolar collapse, reduce mechanical compliance. Normally,
intrapulmonary shunting involves no more than 18 per cent
of the pulmonary vasculature; but within minutes of fresh
water inhalation, some 75 per cent of the lung volume of
such drowning victims manifests intrapulmonary shunting.
FLUID SHIFTS ACROSS ALVEOLAR MEMBRANES
A flux of inhaled water, ilTespective of its osmolality, occurs
across the alveolar epithelium, through the basement mem­
brane and, finally, across the endothelial lining into the cap­
illary lumen, where haemodilution occurs. This flux causes
rapid and severe distortion of pulmonary ultrastructure with
damage to both type I and type II pneumonocytes. 19 ,20 Elec­
tron microscopic studies reveal endothelial changes of cell
swelling, microvesical formation, cell detachment from the
base of the membrane and cell destruction.
SURFACTAI\lT CHAI\lGES
Surfactant, primarily produced by type II pneumonocytes,
is changed within minutes of water inhalation. Sea water
and water containing sodium and chloride concentrations
---~----
that are approximately iso-osmolar (e.g. 0.87 per cent
sodium chloride) with plasma, do not denature pulmonary
surfactant but may dilute it or wash it out of the alveolar
sacs. Fresh water, or inhaled fluid that is significantly
hypo-osmolar, causes acute degradation of surfactant
activity. Whether the lining surfactant is lost or denatured,
the end result is the same and alveoli collapse.
FROTH FORMATION
Exudate outpouring from the laryngeal and tracheal mem­
branes, surfactant washout from the alveolar sacs and dis­
rupted alveolar membranes result in the formation of froth.
The production of froth is characteristic, in variable degree, of
a high proportion of the lungs and upper airways of drowned
individuals. In some cases, it may be the only forensic mani­
festation that drowning has occurred. In cases of neonaticide,
froth in the air passages is a valuable sign; when interpreted
together with changes in the lungs it is indicative that a
newly delivered fehls was born alive. The forensic evidence
of post-immersion froth in the upper air passages is preserved
if freezing has occulTed, and can be demonstrated after
warming even following freezing temperatures as low as
-18°c. 21 It is important to appreciate that any prolonged
freezing of the body of a drowned subject will cause tissue
distortion owing to ice crystal formation. This compromises
the interpretation of subsequent histological examination.
Hypoxic Cascade
In all immersion accidents the cause of death is ultimately
irreversible cerebral anoxia. It is the end result of a hypoxic
march of events, 11 any step of which may be influenced by the
circumstances sUlTounding the immersion incident. This chain
of cardiorespiratory events follows an inexorable sequence:
• voluntary apnoea;
• the diving reflex ensues, pa11icularly in children;
• arterial hypoxaemia occurs;
• tachycardia and hypertension develop;
• tissue hypoxia ensues;
• tissue acidosis develops;
• inhalation with aspiration of water, followed by glottic
spasm leading to a phase of secondary apnoea;
consciousness is lost somewhere at this point of the
pathophysiological sequence;
• involuntary respiratory movements occur, continuing
under water until respiratory arrest occurs; the
diaphragm may continue to contract after intercostal
activity has ceased;
• hypotension occurs, with progressive loss of cerebral
and coronary perfusion;
• dysrhythmias may develop, and in the case of children
who drown the preterminal bradycardia mayor may
not be followed by ventricular fibrillation, before the
inevitable asystole;
- - - - - - - - - -

• cardiac arrest;
• brain death;
• somatic death.
Brain death after immersion accidents follows a primary
state of altered neuronal metabolism. Research studi es of
changes in the hypoglossal nucleus (in the brainstem) indi­
cate that compared with other quicker forms of asphyxia
(e.g. hanging, strangulation and choking) pre-mortem
damage to proteins such as the c-fos gene protein and the
72-kDa heatshock protein occur in this phase. 22
With increased percentages of the population trained in
cardiopu lmonalY resuscitation (CPR) skills and the evolu­
tion of better ambu lance services and more sophisticated
intensive care units, recent years have seen a number of
childhood immersion victims maintained on life support
prior to their delayed death due to drowning. The forensic
pathologist is involved in such cases and coronial autop­
sies are required in almost all jurisdictions. Exa mination
of the brains of these children may demonstrate any or all
of the signs of asphyxial brain death. Although cerebral
oedema may develop in survivors, it is not a feature of
autopsy studies in the drowned victim who dies within
minutes of immersion. Besides petechial haemorrhages
on the brain surface, there may be features of cerebral
oedema, sometimes to the point of hemispheric swelling
with resultant mass effect. Such massive oedema has not
been reported in 'CPR survivors' who died within 24 hours
after rescue. Delayed-death subjects, in whom there exists
the legal imperative of coronial autopsy, may also show a
wide constellation of lung changes. Such autopsy features
range from relatively normal lungs, perihilar pulmonary
oedema, generalized pulmonalY oedema, pneumonia, col­
lapse and consolidation, disrup tion of alveoli, alveolar
haemorrhage, foreign inhaled particles in the airways, and
signs of unilateral and bilateral pneumothorax. At post­
mortem , such 'delayed drowning' lun gs may manifest both
macroscopic and microscopic features of the adult respira­
tory distress syndrome (ARDS). 23 In such cases there are
often other features of pneumonia, due either to nosoco­
mial infection or infection with waterborne bacteria or
colonization by single-celled organisms. In the case of
children who have survived with intensive care or ve ntila­
tor support for hours or days following extraction from the
water, there may be uncommon (indeed, exotic) micro­
organisms found in pneumonia consolidates or lung
abscesses. In the case of salt-water near-drownings, when
death has occurred later in hospital, marine Vibrio bacteria
or algae may be demonstrable.
FORENSIC IMMERSION SYNDROMES
From the perspective of the fore nsic pathologist, the site and
circumstances of the immersion incident are of crucial
importance in the reconstruction of events as these details
Forensic immersion syndromes I 351
will fOlm the substance of the final medicolegal report.
Whereas the clinician is concerned reactively in combating
the clinical consequences of hypoxia, the pathologist is con­
cerned with the recreation of the events that led to the pri­
mary immersion anoxic episode. In more than 95 per cent of
childhood immersion incidents, unlike other forms of child­
hood trauma, the immersion incident is not witnessed. This
fact, combined with historic difficulties of confirming death
by drowning and the relative absence of specific forens ic
signs, places the pathologist in a position of great responsi­
bility. For these reasons, there is an imperative to approach
the pathological investigation against the background of all
possible syndromic templates. There are at least 16 such
age-site paediatric drowning syndromes, each with specific
types of antecedent, different risks for the potential of
unlawful child killing, different approaches to investigation
and, ultimately, different stratagems for prevention. The
importance of such syndromic definition is cognate to the
generation of a clinical and forensic differential diagnosis
when each new case is being assessed. Review of the child's
entire past medical history is essential as a prelude to inter­
pretation of the forensic autopsy results and indeed may
modify the extent of technical investigation (Table 18.3)
used in port-mortem analysis.
In many nations the crimes of neonaticide and infanti­
cide are distinct from those of filicide, child homicide or
unlawful child killing. Such jurisdictions recognize that the
syndromes of neonaticide and infanticide have specific
features not only in terms of aetiology, modus operandi
and sociocul tural implications, but also in prevention.
There exists also the rare problem of false confessions to
the drowning of children .24 In these circumstances the
recreating of events leading to the child's death is crucial to
the legal protection of a parent or other party who may be
psychiatrically disturbed.
In practice, the forensic examination of childhood
immersion victims is straightforward. However, in evelY
case judgement is required with respect to the choice of
post-mortem ancillary tests and investigations. Most juris­
dictions operate under conditions of scarce, and sometimes
inadequate, resources. This applies particularly to patholo­
gists or other medical examiners who are called upon to
investigate childhood immersion deaths away from sophis­
ticated fo rensic centres. 25 A small minority, less than 5 per
cent of all childhood immersion accidents, have the poten­
tial for criminal overtones. However, with the increasing
tendency for civi l litigation, the pathologist may become
enmeshed in highly complex and prolonged civil law suits,
in which grieving and aggrieved parents seek compensa­
tion or damages from parties who are the site owners of
water hazards in which children have drowned . For this
reason, there is an increasing tendency for forensic pathol­
ogists to undertake tests a nd investigations that will leave
no doubt as to the recreation of factual events that led to
death in the water, or after failed attempts at post-rescue
resuscitation.
352 I Drowning and near drowning
Classification of Drowning in Children
There are two paradigms by which child drowning may be
classified. The first of these uses the primary discriminator of
intentionali ty, thus separating accidental drownings from the
syndromes of unlawful child killing. The second approach is
to classify drownings by site, each with its site-specific syn­
dromic boundaries. Most medicolegal approaches to child­
hood immersion incidents use both approaches.
Child-killing Immersion Syndromes
NEONATICIDE
Neonaticide is defined va riously, in different jurisd ictions, as
the killing of a n infa nt in the first 24 hours of life or the
killing of an infant in the first 4 weeks of life. In the forensic
literature it is usually taken to mea n the unlawful child
killing of a subject delivered naturally and of sufficient
developmental maturity to have been capable of independent
survival, whose killing was perpetrated within the first 24
hours of life. Neonaticide by dro wni ng is not rare, but it is,
fOltu nately, decreasing in frequency in many Western soci­
eties. The syndrome is very specific and has been long recog­
nized because of its sad sociofamilial overtones. 26 In general,
neonaticide is in almost always perpetrated by the mother.
In some legal jurisdictions, the crime of neonaticide, with its
connotations of diminished responsibili ty, is only recognized
as a specific crime if perpetrated by the mother.
The mothers are almost always young (95 per cent of
cases), often teenagers. They are almost alw ays single.
Neonaticide by drowning usually occurs in the context of a
concealed pregnancy and in the context of a concealed,
solitary labour and delivery. The mothers are often mem­
bers of ethnic minority groups, often in religious or lan­
guage isolates living in Western society. Such families
typ ically are t hose with religious or tra ditional cultural
condemnation of premarital sexua l re lations . These tragic
incidents are also sometimes encountered in white or black
families, particularly in those of lower socio econom ic sta­
tus, in which there is a very strict, male-dominated ethos in
the microsociety in w hich the parturient mot her is trapped.
In these circumstances it is not rare for a Single teenage
girl or yo ung woma n to conceal the pregnancy. At the time
of the ensuing delivery a nd during labour, the victim will
seek a priva te bathroom or public toil et in which to deliver
her child secretly.26 The mother adopts the least uncomfort­
able position for such so li tary delivery. Und er these circum­
stances, the neonate is often born into a toilet bowl or, less
commonly, in to a bath. In Western and oriental societies this
usually occurs in a private or public toilet cubicleP
There is often doubt about the degree of intent of the
deliberate taking of the baby's life, allow ing for the dimin­
ished responsibility of a dist ressed, usually lonely and
always ignorant primigravida. In many such instances, the
young single teenager, ignorant of physiology and without
antenata l care, de livers herself of her infant. In the context
of fear, pain and ignorance, a proportion of such mothers
may not perceive the risk of immersion death. Having said
t his, neonaticide is the deliberative taking of the newborn's
life and t he related crimes of the deliberate exposure or
abandonment of t he newborn have been classic and not
uncommon crimes known since antiquity. In this context,
drowning is simply the modus of such unlawful child
killing, the incidence of which varies, as it is always done
with t he sociocul tural mores and the economic circum­
stances of the parents of the neonate concerned.
Forensic autopsy of such infants needs to establish
whether spo ntaneous breathing has occurred and whether
the airways are patent, and whether asphyxia by water or
other agents was the cause of death. As a proportion of such
new ly born infants a re found dead in places such as public
to ilets, DNA extraction and storage is essential. The tracing,
finding and identification of mothers is important in these
distressing circumstances, as post-neonaticide identification
of the biological father may ha ve important forensic or legal
implications. Such may occur if there is assa ult on putative
fathers by the girl 's family (particularly by her father or
brothers) and in the rare cases when putative or biological
fathers may be enjoined in the act of neonaticide itself.
INFANTICIDE
The crime of infanticide is the unlawful killing of a child
under 1 year of age by its mother. The designated crime of
infanticide has evolved as the judicial recognitio n that
there is a subset of unlawful killings that are the result of
diminished responsibility.
The deliberate killing of an infant, by a mother often dis­
abled by psychosis, occurs not at birth but in the weeks or
months following birth. Under these circumstances, drowning
is, in one sense, a non-specific modus, as the means of ending
the child's life. 28 Most such perpetrators are suffering from
post-natal depression, with a smaller proportion afflicted with
schizophrenia. 29 The intra family dynamics in cases of infanti­
cide by immersion 3o differ from those encountered in cases of
the deliberative, repeated , sub -fa tal U'auma, which is a feature
of the crescendo child ab use syndrome that, of course, may
ultimately lead to the death of the child concerned.
Post- nata l depression is a dangerous syndrome for both
infant and mo ther. When depressed mothers kill their
infants, the prox imi ty of the family bath tub or of was hing
machines, buckets or pails makes this cause of dea th well
recognized. The syndrome of infanticide-suicide is well rec­
ognized by all forensic pat hologists. Sometimes, a mother
will kill one, severa l or a ll of her children before taking her
own life. 3 1.32 Infanticide by drow ning may reveal, at forensic
autopsy, skin features or lesions co nsequent upo n the force
exerted by the perpetrator to drown the struggling child.
Some parents have attempted to drug their children before
immersing them as the fi nal act of killing.

The forensic pathologist has an important preventative
role in infanticide, not of course in realized cases but as
an advocate for increased surveillance and help to those
mothers afflicted by psychosis (either by endogenous bipo­
lar disorder or by schizophrenia) when the risk of infanti­
cide may be hi gh. Such risk remains high in subsequent
pregnancies, and the syndrome of sequential infanticide,
sometimes by drowning, is well recognized.
FILICIDE
Filicide is that crime in which the offender is a biological,
adoptive or de facto parent. The method of killing is culture­
specific. In European, Asian, Canadian and Australasian
societies the cause of such deaths are head injury, drown­
ing or suffocation. 33 In the United States, homicidal
asphyxia is less common in some regions, where gunshot
murder is more frequently employed. 34
Mothers (60 per cent) kill their children more often than
do fathers. Eighty per cent of such victims are between 1 and
5 years of age, with a median age of between 2 and 3 years. 35
Particular 'at risk' times for such immersion killings are in the
early hours of the evening, particularly during weekends. 36
Site-specific Immersion Syndromes
BUCKET OR PAll
A particularly difficult forensic immersion scenario is that
which involves a child who has drowned in a bucket or
pai1. 37 ,38 The majority of such incidents are undoubtedly
accidental, but there remains a subset, of unknown propor­
tion, in which attempted or realized infanticide or child
homicide has occurred.
The age range of such victims is 9-20 months. Toddlers
are at risk, but infants who are able to pull to stand , but
cannot yet wa lk , are also at risk. The infant or toddler may
become wedged in the bucket. No complete unselected
(fatalities and non-fatalities) series for this type of accident
has yet been published. A review of all published papers
indicates that the mortality rate approaches 60 per cent and
that the risk of post-accident neurological sequelae is high
amongst survivors. A related and rare type of immersion
involves those toddl ers who climb up beside and fall into
washing machines or washing tubs.
Buckets and pails are usually ofJ-S gallons (40-70 L) in
capacity. The contents of the buckets are almost always
water, soiled nappies, dirty mop water, bleach, detergent,
soap or antiseptic. 38 These fluids have the potential to
cause intense bronchospasm and life-threatening laryngeal
spasm , irrespective of the immersion time. In survivors of
such immersion incidents, there is always severe pneumo­
nia and often systemic complications that require the most
sophisticated management in the inten sive care ward over
subsequent days. In the case of fatalities, analysis of bucket
or tub contents and of lung tissue forms an important
Forensic immersion syndromes I 353
component of the chemical and toxicological approach
that is so important in these accidents.
BATH TUB DROWNINGS
Immersion incidents in the family bath tub and in domes­
tic spa pools have special implications for the forensic
pathologist. From both the clinical and the forensic point
of view there are eight defined syndromes of bath tub
drowning and near-drowning (Table 18.5). Of these specific
and definable bath tub or bath spa syndromes, accidental
immersion is the most common in childhood. In adult
series, suicides figure promin ently.39
A special challenge to both general and forensic patholo­
gists, and also to clinicians, is the generation of a differen­
tial diagnosis in respect of causation to encompass the key
prese nting feature - the fact that the bath tub is the site of
the immersi on incident. If the clinician or path ologist does
not consider the various syndromes that constitute such a
forensic differential diagnosis (Table 18.5) then it is certain
that lawyers in subsequent coronia I, civil action or criminal
courts will do so. The fact that perhaps 10 per cent of child­
hood bath tub drownings mask the tragic fin al result of
crescen do child abuse necessitates full post-moltem X-ray
skeletal survey in all such cases.
Al most all bath tub immersions involve fresh (tap) water,
chlorinated to the standard 1 palt per million (ppm). Diatom
analysis plays no part in the interpretation of forensic bath
tub analyses. By contrast, soap products are inhaled with
water in many cases of accidental bath tub immersions
involving infants and toddlers, unlike the situation in adult
suicides 39 and some homicides 4o for which the bath tub is
the site of death.
Bath salts are used in many children's baths, especially
in the USA and Japan 14 and in many affluent families of all
nations. Bath salts contain fluorescein, which is highly sensi­
tive to detection by high-performance liquid chromatography.
Confirmation of fluorescein in lung tissue (and other tissues),
in parallel with its demonstration in any residual bath tub
water samples, may be helpful and specific. Its demonstration
parallels the selected and rare use of diatom marrow demon­
stration following occasional fatal immersions in open water.
Accidental Child Drowning in Bath Tubs
Overall, 80-90 per cent of childhood bath tub drownings
are accidental. The usual scenario of such accidents is the
vely human and universal situation in which a mother, tired
at the end of the day, is attempting to coordinate the feed­
ing and bathing of a large and vigorous family. As part of
this, an infant who is in the bath tub becomes overlooked
for several minutes before his or her well-being is checked.
The syndrome of accidental infant bath tub immersion
is quite specific. Such fatalities and near-fatalities occur
only, or virtually only, in working class, labouring and
poorer families. 41
354 I Drowning and near drowning

Table 18.5 Site-syndromes of bath tub and spa pool drownings: forensic and clinical implications

Syndrome Incidence Notes

Accident 20 per 100000 infants aged 0-"
months, annually
8 per 100000 infants aged 12-23
months, annua lly
Child abuse 10-15 per cent of bath tub
immersions estimated to be
attempted or realized homicide
Child homicide Rare - 12 per cent of all child
homicides 10
Bath tub epilepsy Rare
Euthanasia Rare
Bath tub drowning with Very rare; sudden cardiac death
cardiac pathology (all causes) in apparently normal
children has a frequency of
1.2 per 100000 patient-years
Poorer and work ing class families
Large family size
One-th ird are single-paren t families 50
High successful resuscitation rates (60 per cent)
Child unsupervised due to sudden unexpected break in routine
Parenta l dichotomy of care, each believing the other is caring
for the infant
'Vu lnerable periods' such as acute ill ness or marital discord
Families often known to ch ild protection agencies
Step-parents or de facto relationships inc rease risk
Often other types abuse uncovered
Some perpetrators 'draw back' after the child loses consciousness
Sometimes precipitated by acute stress in a step-parent or de facto
partner who is left alone with a toddler at a time of acute
interpe rsonal discord
Pa rent psychotically dep ressed or with low 10
Child homicide-su icide syndromes fall in this group
High rate of realized death
Only a small percen tage survive
Older children and youn g teenagers
Victims may have uncontrolled epilepsy
Children with a history of epilepsy who lock themselves in the
bath room for ablutions are pa rticula rly at ri sk
A proportion of physically disabled toddle rs and yo ung child ren
Often not concealed from police, altruistic intent
May be a history of cardiac abnorma lities; especially those of bundle
of His and right bundle branch block or long 0-Tsyndrome 56
Although very rare, commonly raised as a defence against a charge of
homicidal drowning or in civil claims for compensation after drowning

These accidents affl ict infants and todd lers in a very
defined, age-specific window of 8-18 months; the modal
age of such accidents is 9-11 months.42 The mean number
of siblings of victims of this type of immersion is greater
tha n that of the population-matched average. The victims
are usually the you ngest or second-youngest child in the
family - often the second-youngest, when there is a new
baby claiming a significant part of the mother's attention.
This peculiar vulnerability to accidents, to which children
in higher birth ranks within the family are especially prone,
is known to app ly to other childhood accidents as well,
especially accidental poisoning.43
More than one-half of bath tub immersion in cidents
occur during a specific 'vulnerable period' when the family
routine is suddenl y or unexpectedly broken, such as that
which occurs during acute sickness afflicting either the
parents or children or in the context of martial strife.41 A
typical scenario is that which involves a stressed mother,
who is tired at the end of the day, attempting to cope,
unaided, with the control, bathing and feeding of several
high-spirited or fractious, but always vigorous, yo ung
children. The telephone rings, or an appliance breaks, or
someone calls unexpectedly at the door. The mother lea ves
two or three children, including an infant, in the family bath
tu b,44,45 leaving the youngest in the care of an older, but st ill
pre-school, child - the older children may hop out of the
bath when the mother is gone, leavin g the infant alone.
Another factor in some infant bath tub drownin gs is the
use of bath seats, wh ich confer inappropriate assurance of
safety,45.46 despite clear instruction from the manufacturers
that the baby should not be left unsupervised.
Unlike other childhood drowning accidents that all
show a preponderance of males, accidental bath tub
immersio ns show an equal sex ratio. Of those in fan ts and
toddlers found unconscious in the family bath tub, some
60 per cent respond to CPR, with residual intact intellect
and without neurological sequelae. This reflects the rela­
tively short immersion times invol ved in such incid ents.
Bath Tub Drownings and Child Abuse
All reviews of consecutive, unselected series of ch ildhood
bath tub immersions with detail ed follow-up have revealed
th at some 10-25 per cent of such cases are in fact the result

- -. -­

of non-accidental injury35,42,47 Forensic studies in Ger­
many have shown that in 5-10 per cent of all child homi­
cides in the under-5 age group, the modus operandi is
drowning, usually in the bath tub. 48
The immersion assault is usually one such in a
crescendo series of acts of child abuse perpetrated on the
infant or toddler. Step-parents and de facto partners are
almost exclusively the perpetrators. Some 'lose their nerve'
during the assault and draw back and may summon help
for resuscitation. A clue to the specific syndrome of child
abuse by bath tub immersion may be that the child is older
than the modal age for true accidental drowning in this
site (9-11 months); or ou tside the typical 8-18 months age­
specific window of vulnerability to bath tub immersion
homicide. The median age of normal infants to be able to
pull to stand is 9 months and the media n age for walking
in normal full-term infants is 13 months. Usually, children
older than these ages ca n support their heads out of water
in a bath tub, even if abandoned or left unattended for
short periods of time. There should be a high index of
forensic or coronial suspicion in all bath tub immersions
involving toddlers older than 11 months, particularly if
there is no history of developmental delay or epilepsy. The
median depth of water in bath tub immersion incidents is
approximately 20 cm.
Although deliberate hot water scalding by immersion in
the bath tub is not uncommon in reported series of child
abuse cases, the pathology of bath tub immersion incidents
is primarily that of hypoxia and the two injuries are not
reported coincidentally. It may be that the perpetrators do
not wish to scald their own hands.
Bath Tub Drowning and Child Homicide
There exists the separate and distinct syndrome of child
homicide, as a 'one-off' event, by bath tub drowning. As a
single act, not in the context of ongoing child abuse, this
assault is relatively uncommon. Syndromically, it almost
always involves a parent afflicted by psychiatric illness or
by low intellectual ability. Mothers are usually the perpetra­
tors, have a psychiatric history and are living in disadvan­
taged socioeconomic environments. Mothers who kill their
children prefer to strangle or drown their victims. Psychol­
ogists believe that in a large proportion of such cases the
psychodynamics are those of surrogate suicide. 49 It is some­
times the modus operandi of the familiar and tragic 'child
homicide-parent suicide' doublet. In this latter incident,
tragically familiar to all forensic pathologists, one parent is
almost always psychotically depressed, or in unremitting
despair, and kills his or her child or several children and
then immediately takes their own life. Survival rates for
children involved in such incidents are low; less than 10 per
cent sUlvive this tragic scenario. Child homicide by drown­
ing, in the context of sexual assault followed by homicide,
always perpetrated by a stranger or non-biological relative,
is rare. The mode of killing is almost always by physical
Forensic immersion syndromes I 355
violence or strangulation, but the body is sometimes dis­
posed of in a dam, creek or watercourse.
Infant homicide by bath tub drowning approaches the 'per­
fect crime' in that forensic skills usually cannot distinguish
non-accidental submersion from accident. Two series of bath
tub immersions have revealed that child homicide is a subset
of the bath tub drowning syndrome - known only because of
later, unsolicited confessions by the perpetrators. 24 ,50
Childhood Bath Tub Immersion in Epilepsy
A small proportion of children and teenagers drown, or
almost drown, in the bath tub as a result of an epileptic
seizure sl Enquiry about a past his tolY of epilepsy is impor­
tant in both the clinical and forensic workups of all immer­
sion accidents. There exist several specific issues relating to
this syndrome.
It is very rare for children to drown in the sea or swim­
ming pools as the result of in-water seizures. 52 This applies
even to children with uncontrolled epilepsy. The risks are
significantly greater if there is a sub-therapeutic serum
concentration of anticonvulsant dnlg present. The inci­
dence of post-seizure bath tub fatality is less than 0.1 per
cent of all childhood drownings.
There is a higher specific risk to epileptic children, espe­
cially teenagers with epilepsy, to fatal and near-fatal drown­
ings in the family bath tub. All parents (and children) in
families in whom uncontrolled epilepsy is present are
warned not to take private plunge baths in the family bath
tub, but rather to shower standing up. Several cases of flan­
nels or 'washers' occluding the drainage plug, even during
such 'stand-up' showering ablutions, have been repOlied in
forensic sel1es of fatal bathroom immersion acciden ts related
to epilepsy.
Teenagers, being what they are, are often insistent on
absolute privacy in the bathroom and insist on locking the
bathroom door before taking a plunge bath or shower. If a
seizure occurs, and the unconscious victim slides beneath the
water, the chance of a successful resuscitation is small. Such
cases are represented in all forensic series of childhood
immersions. It is believed that the (misguided) practice of
placing feverish infants into a hot bath - raising a high risk
of febrile convulsion - may be the cause of immersion in a
small proportion of cases of childhood bath tub fatalities.
Water impinging on the face, involving the sensory dis­
tribution of the trigeminal nerve, is the afferent trigger
for the diving reflex. 18 This leads to a complex series of
dynamic cardiovascular changes, including' increased
vagal tone. There is some evidence that this may be one
cause of 'bath epilepsy', initially described in the Indian
medical literature but known to occur in all races.
Bath Tub Drowning and Euthanasia
All series of unlawful child killings include cases of
euthanasia. 53 They are usually perpetrated by a parent but
356 I Drowning and near drowning
rarely by a grandparent or other close family member. One
method of euthanas ia is bath tub immersion. In the major­
ity of such cases the chi ld has a congenital disability, such
as a neural tube defect, mental retardation , cerebral palsy
or a chromosome abnorma li ty. In occasional cases, the par­
ent is psychotic, wrong ly believing that a normal child is
disabled or suffering when such in reality is not the case. 49
In instances of euthanasia by drowning, the motive is
always (albeit tragically) an altruistic one. The parent believes
that the child is suffering and t hat it would be kinder if the
child was put out of his or her pain or misery, or that they
could be helped from a t ragic black or wicked world. There
may be no attempt to conceal the crime. Some perpetrators
proactively contact the police to report the incident.
Bath Tub Drowning and Organic Pathology
A final bath tub drowning syndrome is one in which pre­
mortem pathology of the heart, recognized or unrecognized,
is the cause of loss of consciousness during bathing. This
syndrome is rare, but is included both for completeness and
because the question commonly arises in the legal defence
mounted by those charged w ith homicide by drowning. Sud­
den cardiac deaths in childhood and during adolescence
are we ll known to forensic pathologists. 54 Myocarditis was
demonstrated in 5 out of 22 cases of accidental drowning
and underlines the need for full investigation of drowning
deaths wherever they OCCUr. 55 Importantly, the demonstra­
tion of pre-existing cardiac abnormalities in a drowned child
may protect any innocent person on whom suspicion of cu l­
pable neglect has faUen. This applies not only to situations
where ch ild homicide is being considered in the differential
diagnosis, but also to situations such as deaths in public
swimming pools where lifeguards may be subject to cliticism
of neglect, sometimes of culpable neglect, in subsequent
legal proceedings.
Congenita l or acquired cardiac abnormalities are also
the antecedent to drowning in some cases of older victims
who drown in bathtubs,56 swimming pools or open water.
Anomalies of the bundle of His, ab normalities of the right
cardiac bundle branch and congenital abnormalities of the
coronary arteries have all been reported in cases of teenage
drowning, including bath tub drowning, in which toxicolog­
ical tests for alcohol and illegal drugs have proved negative.
In rare cases, genetic cardiomyopathies may be present. 57
The mechanism of such sudden and un expected death dur­
ing swimming or bath tub bathing is conjectural. One pos­
sibility is that the diving reflex (which follows facial
imm ersi on) induces cardiac dysrhythmia with heightened
sens itivity, i.e. a priori abnormal hearts. It h as also been
suggested that cardiovascular damage may follow reduced
venous return due to venous dilatation in hot water immer­
sions; if this is superimposed on an 'at risk' heart then car­
diac output may be insufficient to sust ain consciousness.
Under normal circumstances, the cardiac output wi ll be
sufficient to return brain function to co nsciousness if an
individual temporarily col lapses. However, if the individual
is under water, secondary vagal stimulation from laryngeal
irritation may compound the hypoxia that is already present.
In this co ntext, it is important to note that any drowning
episode itself may provide an outpouring of catecholam ines
(the 'sympathetic storm'), whi ch, w ith hypox ia , may result in
microscopic changes to the myocardium , including myocyte
contraction banding and foc al myocyte hypereosinophilia s 8
SWIMMING POOL DROWNINGS
In most temperate and tropical countries of the developed
world, the principal site of drowning in the childh ood years
is the backyard swimming pool. Salt-water imm ersions
involving children are now less common than fresh-water
immersions, even in littoral regions, for this reason. In the
USA alone there are now over 12 million pla st ic wading
pools and over 5 million surface swimming pools, of which
an estimated 2 million are of the more dangerous, in-ground
variety. Proportionate rates may be even higher in other
countries, such as Australia and New Zea la nd. In many
countri es, pool drownings constitute 70 per cent or more of
all consecuti ve unselected series of immersion deaths in
ch ildhood. In-ground pools cause 80 per cent of swimmi ng
pool fatalities. Motel , hotel , caravan and trailer park pools
are a particular hazard in all reported series. 59 Even in
countries such as Finland, as many near-drownings take
place in swimming pools as occur in lakes, and twice as
many in swimming pools as in the sea.
The current status of swimming pool drownings has been
a feature of forensic series only since the early 1970s. 47 At
t hat time, engineering technology and social affluence made
private in-ground pools widely accessible. In many commu­
nities in many nations, one in five homes now has an in- or
above-gro und swimming pool filled with water at least
throughou t the warmer months of t he year. Even in temper­
ate climates, where winters may be free zing, pool-house
ratios may be as high as 1: 10. In the State of California, some
100 toddler dro wnin g deaths continue to occur annually.
The age spectrum of such victims is between 12 and 40
months, with a modal peak between 18 and 24 months. The
social class 'risk curve' is U-shaped, with deaths occurring
disproportionately w ithin richer families who, most com­
monly, have in-ground pools, and in poorer families where
above-ground pools are often bought as an impulse purchase,
installed and then poorly maintained. Above-ground pools
are not as dangerous as in-ground pools. However, some
children in the 18- to 36-m onth age range will climb a pool
ladder, or will place a box or other object beside an elevated
pool to gain access. Overall, 70 per cent of toddlers who
drown do so in their own pools. Other 'at lisk' pools are those
of neighbours, motel, caravan or t rail er park pools and the
pools at homes of relatives whom children visit.
Toddlers and yo un g children in this age-vulnerable
window do not fear water. Many simply crawl into it, or
step into such pools if no barrier is present. The Brisbane

Drowning Study 47 showed that often such vIctIms were
attracted to the water by a floating toy, mbber ring or ball ,
or by an object lying on the bottom of the pool. The follow­
ing factors lead to higher survival rates following pool-side
CPR: clothing (with its buoyancy), the diving reflex (pre­
served in infants and toddlers) and children found floating
(with residual aeration) as opposed to those discovered on
the bottom.
Toddlers virtually never drown in fenced pools with
safety standard-approved self-latching gates with high,
hidden locks. Intense advocacy, often supported by pathol­
ogists, to introduce policed, safety legislation to protect
toddlers from such water hazards has been unsuccessful
in many communities. Most toddlers who drown in such
pools do so within 30 m of their own home. Immersion
times are almost always under 20 minutes. Successful
resuscitation rates for consecutive, unselected series of all
such swimming pool immersion accidents approach 60 per
cent and rise to 70 per cent if a trained first aider happens,
coinCidentally, to be involved in the attempted resuscita­
tion. Survivors do well, and some 97 per cent of children
who survive this near-drowning syndrome function nor­
mally. Some 30 per cent of such survivors have wide sub­
scale disparities on formal psychometric testing but do not
manifest clinical neurological signs.
Autopsy findings of such children are typically those of a
completely fit and well toddler who is alive at one moment
and dead 10 minutes later. In such cases there may be mini­
mal anatomical changes in the lungs. The body often shows
post-mortem signs of CPR trauma, including fractured ribs,
contused pericardium or injection needle marks from the
frantic use of cardiostimulant drugs by ambulance officers,
paramedics, firemen (in the USA) or physicians.
PUBLIC SWIMMING POOL DROWNINGS
A small proportion of children, almost always in the age
range of 3-1B years, drown in public access swimming
pools or aquatic parks. One US study draws attention to an
excess of deaths in this situation to black, adolescent
males 60 Such individuals are typically found on the bot­
tom of the pool by another casual swimmer. Very com­
monly, such drownings are followed by major medicol.egal,
civil, insurance and regulatory consequences. A high pro­
portion of such drownings becomes the focus of prolonged
court cases, often involving claims for large compensatory
sums of money; commercial pool owners, lifeguards, para­
medics, emergency room physicians and forensic patholo­
gists may all become enmeshed in complex discussions
about pathophysiology and the differential diagnosis of the
causes of drowning in general.
The majority of such cases are simply due to the fact that
individuals, usually poor swimmers, cannot stay afloat and
so they inhale and the 'drowning chain' ensues. Swimming
lessons and, perhaps, optimistic evaluation of swimming
abjJj!y, may engender a faJse sense of security amongst
Forensic immersion syndromes I 357
parents. They are no substitutes for parental vigilance and
close supervision when children are in the water. 61 •62 A small
proportion of such cases involve children who have been
skylarking in the pool surrounds, fall and hit their heads or
become winded and cannot extricate themselves when they
become submerged. Another subgroup is probably second­
alY to cardiac abnormalities, often congenital in nature but
hitherto unrecognized; this group includes those with con­
genital coronalY artery or valvular abnormalities or those
with hitherto unrecognized dysrhythmias, with mechanisms
similar to those described in the case of bathroom immer­
sions. Some rare cases are consequent upon cervical spine
injuries following diving accidents. A very small sub-group
are consequent upon epileptic seizures in the water. In this
latter context, however, it must be said that swimming pool
drownings due to proven epileptic seizures are extremely
rare. This is not to say that children do not have epileptic
seizures in the water but, in practice, most are recognized
and the victim is extracted without significant hypoxia and
without any subsequent post-hypoxic neurological defects.53
Another subset comprises those young teenagers who have
taken alcohol or drugs, and who lose coordination in the
water, or sometimes who are illegally or clandestinely swim­
ming in pools, often as members of group 'larks', sometimes
involving 'skinny dipping' (nude swimming) at night.
Another subset consists of children performing the danger­
ously underrated practice of hyperventilating before diving
and underwater swimming in such public pools.
Because of the extended differential diagnosis that is
inescapable in such cases, often dissected in meticulous
detail in subsequent court proceedings, it is absolutely
essential that the widest array of appropriate post-mortem
investigations be undertaken. This includes radiographic
studies (especially of the skull, chest and spine), alcohol and
drug assays and meticulous forensic autopsy techniques in
the examination of heart, lungs and brain (Table IB.3). In
practice, it is uncommon for such cases to result in positive
findings in any of these investigations, but the careers, rep­
utation and professional security of many individuals who
are involved in the rescue, the attempted resuscitation and
the management chain may well be dependent on the
pathologist's findings and defended opinion. Perhaps most
importantly of all, 'closure' of the grieving process by par­
ents and loved ones is facilitated by the forensic patholo­
gist's confident recreation of the events that have led to
post-immersion somatic death.
RIVERS, CREEKS AND LAKES
Children occasionally drown in rivers and creeks and in
lakes. 63 Those who do are almost always (90 per cent) boys,
and the modal age falls in the age window of between B
and 12 years. There is often an element of disobedience
involved in such cases, when children are swimming far
from supervision. The typical scenario is when a boy or
maJe teenage! experiences diff]cuJties in the water and his
358 I _ Drowning and near drowning
friends find that they are unable to rescue him and run
for help. Because of the distances involved, resuscitation
attempts are unsuccessful in more than 90 per cent of
cases. Estim ated immersion times are relatively lon g and
often exceed an hour or more before the body is extracted
from the water. The forensic autopsy often reveals gravel,
sand or aquatic flora in the air passages.
DRAINS, TRENCHES AND SEWERS
Children who drown in drains, trenches and sewers are
almost always boys, and sometimes more than one victi m is
drowned in the same episode. The victims are often playing
or swimming in forbidd en areas. Often, extraction from the
water is difficult and bodies may be wedged in outlet pipes
or municipa l watercourses. A series of such drownings
always include that subset of children who are playing in
fast-movi ng and dangerous waterways during floods. Under
such circumstances the novelty of the situation may mask
the threat of immersion and the bodies of the victims may
never be recovered. Other are found wedged in or entrapped
in pipes or outlet drains or under submerged trees.
SEA DROWNING
Many nations have hi gh shoreline-area ratios, with a pop­
ulation who are sea-oriented from infancy. In such regions,
children grow up in an environment in which respect for
the ethos of both water safety and water danger is a natu­
ral part of culture and folklore. Under these circumstances,
sea drownings involving toddlers are relatively uncom­
mon. 64 Older children are particularly at risk.
Child victims of sea drownings are found disproportion­
ately amo ng immigrants, tourists or other ethnic sub­
groupS.17,65 The children and their parents may not be aware
of the threat of the sea or surf. Occasionally, young teenagers
drow n during sailing excursions but, in general , boating and
the use of surfboards, in current practice, are lo w drowning
threats to children. Childhood drownings in the sea occur
almost always duri ng daylight hours and, as a manifestation
of the recreational use of the sea, most often occur durin g
weekends or holidays. Childhood salt-water immersion rates
are unaffected by tidal state. Most sea and surfing beaches
are separated by road, dune strip and a beach from residen ­
tial houses. The result of this is that it is very unusual indeed
for toddlers or preschool children to drown in the open sea;
the modal age for this type of sea drownings is 8 years. The
phenomenon of 'secondary drowning' is most often identi­
fied after salt-w ater drownings. In this latter syndrome an
indiv idual may be extracted, pulseless, from the water and
may respond well to cardiopulmonary resuscitation, only to
manifest life-threatening deterioration hours later due to
washout or denaturation changes in pulmon ary surfactant.
Children are sometimes envenomed by jellyfish or poi­
sonous fishes while swimming.66 Under such circumstances,
as in the case of adults, there is a very serious risk of
--
secondary drowning consequent upon pain, panic and the
ensuing incoordination even amongst proficient swimmers.
The victims of stonefish envenomation, which are com­
monly encountered throughout the Indo-Pacific littoral in
both tropical and temperate waters, often become mani aca l
with pain . If this occurs during snorkelling, scuba-diving
or reef walking, drowning is an ever-present threat. A sea rch
for, and identification of, barbs or other foreign material
from the venom apparatus is important in such cases. Severe
envenomation by cubomedusans can cause death by direct
envenomation while the victim is still in the water, or from
a combination of envenomation and drowning. Box jelly­
fishes (sometimes called 'sea wasps') are amongst the world's
most venomous creatures and many of the recorded victims
of such deaths, a disproportionate number of whom are chil­
dren, occur before the victim can be extricated from the
water. In the forensic examination of such victims, a strip of
adhesive tape applied to areas of the skin thought to be
affec ted and then examined under the microscope may
reveal not only the characteristic nematocysts, but also will
enable both genus and species identification to be made.
The foren sic autopsy of salt-water drowned victims usu­
ally shows no specific distinguishing features if the body is
retrieved w ithin several hours after the drowning episode.
As in adults, the bodies of long-immersed child victims
may show secondary changes due to crustacean or piscine
post-mortem damage. In such cases forensic identification
may be difficult and dental X -rays, post-mortem finger­
printing or DNA extraction and analysis are essential.
AFTERMATH
Only a minority of immersion incidents result in fatali­
ties. 67 ,58 An understanding of the pathophysiology of the
immersion sequence leading to somatic death, taken in con­
junction with the documentation of post-mortem changes,
will mean that both more efficient prevention and better car­
diopulmonary resuscitation will be possible in the future.
The documentation and subsequent compilation of child­
hood immersion statistics is very important from the point
of view of future prevention. In this context, case finding of
immersion fatalities undertaken for epidemiological research
is compromised because of the many different causes and
syndromes of drowning in childhood. As many as 1 in 10
cases of childhood immersions may not be retrieved for
analysis. For example, some are coded as motor vehicle
injuries in those cases when vehicles have crashed into the
water.59 Multiple-cause coding is essential if a community's
or nation's statistics are to be fully exploited in the context
of injury prevention. 70
The pathologist's professional and ethical responsibility
is to the dead child and to the determination of the circum­
stances of that death. The evid ential value of such autopsy
findings, however difficult may be their interpretation in the
current stage of knowledge, is of the greatest impOltance.
Ultimately, the integrity of not only the professions of
~
-

medicine and of law depends upon such a fo rensic service,
but also that of society. The deliberative killing or preve nta ­
ble death of a child defines, in part, the society in which such
tragedies occur. From the professional point of view, the
pathologist also has both a professional and an ethical
responsibility to any living siblin gs of children who have
died from immersion, whether accid ental or no t.
Much remains to be ach ieved in fos terin g closer coll ab ­
oration between clinici a ns ca rin g for the survivors of near­
drowning in cidents and the path ologists who investigate
fatal imm ersions. 7l •72 Paediatric a nd forensic pathologists
have an important role to play in preventative medicine
a nd in promoting advocacy for public health stratagems
that relate, in particular, to home and water safety in com­
munities in which drownings are common.
A child's death from drowning is, in one sense, from the
family's po int of v iew, the beginning of a new era. The role
of the pathologist is a crucial one in helpin g parents in
t he immediate aftermath of death. To be able to recreate
exactly what has happened often brings long-term resolu­
tion to the aftermath ofa child's drow ning, which is always
a t ragedy.
ACKNOWLEDGEMENT
I thank Dr Terry Sinton, a senior for ensic pathologist of the
John Tonge Cent re, Queensla nd Health Pathology Services ,
Brisbane, Australia, for mu ch encouragement.
REFERENCES
Centers for Disease Contro l and Prevention (C DC). Nonfa tal
a nd fata l d row nings in rec reatio nal wa ter settings - United
States. 200 1-2002. MMWR 2004; 5 3:447- 52 .
2 Meyer RJ, Theodorou AA, Be rg RA. Chil dhoo d drownin g.
Ped iatr Rev 2006; 27:16 3- 9.
3 Fin gerhut LA, Cox CS, Warner M. Internat iona l co mparative
ana lyses of injUlY mortality. Findings from the ICE on inj ury
sta tistics. International Collaborative Effort on Injury
Statistics. Adv Da ta 1998; 303: 1-20.
4 Peden MM , McGee K. Th e epidemiology of drowning
wo rld wide. Injury Cont Safety Promotion 2003; 10:1 95- 9.
5 Laraque D, Barlow B, Durkin IVI. P reve ntion of yo uth injuries.
J Natl Med Assoc USA 19 99; 9 1 :557-].
6 Aus tralian Institute of He alth an d Welfare. Australia's Hea lth
2000. The Seven th Biennial Hea lth Report of the A ustra lian
In stitute o f Health a nd Welfare. 4.2 Childre n a nd Young
People. Canb erra: Aust lnst Health Welfare (AIHW Ca t. No.
19), 2000, pp. 184-92.
7 Byard RW. Accidental childh ood death and the rol e of the
pa th o log is t. Pediatr Dev Pat/101 2000; 3 :405-1 O.
8 Pearn JH, Nixon J. A n analys is of the causes of freshwater
immersion accid ents involving chiJdren. Accident Anal Prell
1979; 11:173-8.
9 CoUins KA, Nic hols CA. A decade of pediatri c ho mi cid e: a
retrospecti ve study a t the Medical Uni versity of So uth
Carolina . Am J F01'ensic Med Pathol 19 99; 20:169 -72.
References I 359
10 Byard RW, Houldsworth G, J a mes RA, Gilbert JD .
Cha racte ri stic feat ures of sui ci dal drownings : a 20-y ear st udy.
A m J Forensic Med Pathol 2001; 22: 134-8.
11 Pearn JH. Drowning. [n Pearn J, Leditschke JF, Marsha ll V
et al. Th e Science of Firs t Aid. Theore tical and Experimental
Basis of Modem First Aid Practice. Canberra: St Joh n
Amb ulan ce Australia , 1996, pp. 138-47.
12 Ludes B. Coste M, NOlih N et al. Diatom analysis in victim's
tissues as an ind ication of the site of drowning. Int J Legal
Med 1999 ; 11 2 :163-6.
13 Pearn JH. The manageme nt of near drowning. In Aochi 0,
Ama ha A, Takesh ita H red s) Intensive Critica l Care Medicine.
Amsterdam: Excerpta Med ica, 1990, pp. 139 -46.
14 Muka ida M, Kimura H, Takada Y. Detect ion of bathsa lts in the
lungs o f a baby drowned in a bathtub: a case report. Forensic
Sci Int 1998; 93:5 -11.
15 Modell JH, Bellefleur M, Dav is JH. Drownin g withou t
asp iration: is this an appropriate diagn osis? J Forens ic Sci
1999; 44:1119- 23 .
16 Gi lbert FF, Gofton N. Te rm ina l di ves in min k, muskrat and
beaver. Physiol Beh av 1982; 20:8 35-40.
17 Pea rn JH. Dro wning. In Di cke rmann JD, Lucey JF red s) The
Critica lly III Child, 3rd edn. Philad el phia, PA: WB Saunders,
198 5, pp. 129 -5 6.
18 Gooden BA. Mechanism of the human diving response.
Integ rative Physio l Beha l! Sci 1994; 29: 6-1 6.
19 Torre C, Varetto L, Tappi E. Scann ing el ectron microscopic
ultrastructural a ltera tions of the pulmonary alveolus in
ex perimental drowning. J Forensic Sci 1983; 28: 1008-1 2.
20 Gu MN, Xia o JF, Hua ng YR et al. Study of di rect injury by
sea water in canine models. Di Yi JUI1 Yi Do X ue Xue Bao
2003; 23:201-5.
21 Tabata N, Morita M, Azu mi J. A frozen newbo m infant: froth in
the ai r-passage afte r thawing. Forensic Sci Int 2000; 24:67-74.
22 Gotoh da T, Kubo S, Kitamura 0 et aL HSP70 an d c-Fo s
expression of brain stem hyp oglossal nucl eus in drowning.
J Med Inl!est 2000; 47:76-9.
23 Sac hd eva RC. Near drowning. CritCare Gin 1999; 15: 28 1-96.
24 Heinemann A, Pus chel K. Discrepancies in homicide statistics
by suffo cation. Arch KrimilJo/1 99 6; 197:129-41.
25 Norman MG, Smialek JE, Newman DE, Horemba la EJ. The
post- mo rtem exam in ation o n the a buse chil d. Pathological,
radiograp hic, a nd legal aspects. Perspect Pediatr Pathol 1984;
8: 313 - 43.
26 Saund ers E. Neonati cides followin g 'secret' pregnanc ies: sev en
case reports. Public Hea lth Rep 1989; 104:368-72 .
27 Shiono H, Maya A, Tabata Net aL Med icolega l aspec ts of
infant ici de in Hokkaid o District, Japan. Am J Forensic !VIed
Path ol 1986; 7: 104- 6.
28 Scott PD. Fatal ba ttered baby cases. IvIed Sci La w 1973;
13:1 97 - 206.
29 Kondo T. Ohshim a T. Re trospective in vestiga tio n of medico­
legal autopsy cases involv ing mentally handicapped
individu als. Nippon Hoiga ku Zasshi 199 5; 49:478-0 3.
30 Geertin ger P, Voigt J. Death in the bath. A survey of bath tub
deaths in Copen ha gen, Denmark a nd Go th en burg, Sweden.
J Forel1Sic Med 1970; 17:136-47.
31 Bourget D, Gagne P. Matern al fili cide in Quebec~ ) Am
Psychia try Law 2002 ; 30 :345-51.
32 Rouge-MaiHart C, Jou sset N, Gaudin A et a l. Women who kill
their chil d ren. Am J Forensic Med Parhol 2005 ; 26:320-6.
33 Vanamo T, Kauppi A, Karkola K et aL Intra -familial child
homicid e in Finland 1970-1994: incid ence , caus es of death
and demographic chara cteristics. Forel1sic Sci Int 200 I;
ll7: 199-204.
34 DiMaio VJ. Hom icid a l asphyx ia. Am J Forel1Sic Med Patilo l
2000; 21: 1-4.
360 I Drowning and near drowning
35 Gillenwater JM, Quan L, Feldman KW Inflicted submersion in
childhood. Arch Pediatr Adolesc !\lIed 1996; 150:298-303.
36 Schmidt P, Grass H, Madea B. Child homicide in Cologne
(1985-94). Forensic Sci Int 1996; 31: 131-44.
37 Scott PH, Eigen H. Immersion accidents involving pails of
water in the home. J Pediatr 1980; 96:282 -4.
38 Wa lker S, Midde lkamp IN. Pail immersion accidents. Clill
Pediatr 1981; 20:341-3.
39 Copeland AR. Suicide by drowning. Am J Forensic Med Path
1987; 8: 18- 22.
40 Schmidt P, Madea B. Homicid e in the bathtub. Forensic Sci IIlI
1995; 72:135 -4 6.
41 Pearn JH , Brow n H, Wo ng R, Bart R. Bathtub drownings:
report of seve n cases. Pediatrics 1979 ; 64:68-70.
42 Kemp AM, Mott AM, Sibert JR. Accidents and child abuse in
bathtub submers ions. Arch Dis Child 1994; 70:435-8.
43 Pearn J. Predisposing fact ors leading to child trauma.
J Epidemiol COI11I11 Hlth 1978; 32:190-3.
44 Byard RW, deKoning C. Blackbourne B et al. Shared bathing
and drowning in infants and you ng children. J Pediatr Child
Health 2001; 37:542- 4.
45 Somers GR, Chiasson DA, Smith CR. Pediatric drowning: a 20­
year review of autopsied cases: III Bathtub drownings. Am J
Foren sic Med Patho12006; 27:113-16.
46 Sibert J , John N, Jenkins D et al. Drowning of babies in bath
seJts: do they pro vid e fal se reassurance? Child Care Health
Dev 2005; 3 1 :255-9.
47 Nixon J, Pearn J. Non-accidental immersion in the bath:
anoth er extens ion to the syndrome of child abuse and
neglect. Child Abuse Neglect 1977; 1 :455- 8.
48 Griest KJ, Zumwalt RE. Child abuse by drowning. Pediatrics
1989; 03:41-6.
49 Ma rl eau JD , Roy R, Laporte L et al. Infanticide committed by
the mother. Can J Psychiatry 1995; 40: 142 - 9.
50 Pearn JH, Nixon J. Swimming pool immersi on accidents. An
analysis fro m the Brisb ane Drowning Study. Med J Aust
19 77; 1:432- 7.
51 Brenin gsta ll GN. Mortality in pediatric epilepsy. Pediatr
Neuro12001; 25:9- 16.
52 Pearn J , Bart R, Yamaoka R. Drow nin g ri sks to epilep tic
children. A study from Hawaii. BMJ 1978; 4:1284-5.
53 Wilkey I, Pearn J, Petrie G, Nixon J. Neonaticide, infanticide
and chi ld homicide. Med Sci Law 1982; 22:31-5.
54 Wilkinson JL. Sudden cardiac death in childhood and
adolesce nce. J Paediatr Child Health 1994; 30:384-5.
55 Somers GR, Smith CR, Wi lso n GJ et al. Association of
drown in g and myocarditis in a pediatric popU lation: an
autopsy-bas ed study. Arch Pathol Lab J\!led 2005:
129:205-9.
56 Stumpp JW, Schn eider J, Bar W. Drownin g of a girl with
anomaly of the bundl e of His and the right bundle branch.
A 111 J Forensic !\lIed Pat/wi 1997 ; 18 :208 -10.
57 Schmidt P, Madea B. Death in the bathtub involving chi ldren.
Forensic Sci Int 1995; 72:147-55.
58 Lunt DWR, Rose AG. Pathology of the human heart in
drowning. Arch Parhol Lab iVIed 1987; 111:939-42.
59 Pearn JH , Brown J, Hsia EY. Swimmi ng pool drownings and
near-drownings involving chi ldren. A total population study
from Hawaii. Milit Med J980; 145 :15- 18.
60 Brenner RA, Trumble AC, Smith GS et a l. Where children
drown, United States, 1995. Pedia trics 200 1; 108:85-9.
61 Brenner RA, Saluj a G, Smith GS. Swimming lessons,
swimmin g ability and the risk of drowning. Injury Cont Safety
Promotion 2003; 10:211-16.
62 Moran K, Stanley T. Pa renta l perce ption s of toddler water
sa fety, swimm in g ability and sw immi ng lessons. lil t J Cont
Safety Promotion 2006; 13:139-43.
63 Editor. Drowning in a private lake, NOlth Carolina, 1981 -9 0.
MMWR 1992; 41:329-31.
64 Patrick M, Bint M, Pearn J. Saltwate r drowning and
near-drowni ng acc idents involving ch ildren. Med J Aust
1979; 1 :61-4.
65 Schul pen TW. Migration and child hea lth: the Dutch
experience. Eur J Pediatr 1996; 155:35 1-6.
66 Pearn JH. Th e sea, stingers a nd surgeons: the s urgeon's role in
prevention, first aid and management of marine
envenomations. J Paediatr Surg 1995; 30: 105-10.
67 Pearn JH. Surv ival rates after se rious immersion acc id ents in
childhood. Resuscitation 19 78; 6:271-8.
68 Henderson H, Wilso n RC. Water incident related hospiral activity
across England between 1997/8 and 2003/4: a retrospecti ve
descripti ve study. Bi\;lC Public Health 2006; 16:2 10.
69 Wintemute GJ, Kraus JF, Teret SP et al. Deaths resu lting from
motor vehicle immersions: the nature of the injuries, personal
and environmental contribu ting facto rs and potentia l
interventi o ns. Am J Publ Healrh 1990; 80:1068-70.
70 Smith GS , Langley JD. Drowning surveilla nce: ho w we ll do E
codes identify sub mersion fa ta lities. Injury Prev 1998;
4: 135-9.
71 Start RD , Delargy-Aziz Y, Dorries CP et al. Clinicians and the
coroni al system: ability of clinicians to reco gnise reportable
deaths. BMJ 1993; 306:1038-41.
72 James DS , Leadbeatter S. Detectin g homici de in hospital. J R
Coli PhYSicians Lond 1997; 3 1:296-8.
73 Fornes p, Pepin G, Heud es D, Leco mte D. Diag nosi s of
drowning by combined computer-assisted histo morphometlY
of lun gs with blood strontium determination. J Fore nsic Sci
1998; 43 :772-6.
74 Puschel K, Schulz F, Darrmann I, Tsokos M. Macromorphology
and histology o f intramuscular haemorrhages in cases of
drowning. Int J Legal ili/ed 1999; 112:101-6.
75 Hooper WD, Johnson LP. Antiepi leptic drugs: pharmacology
and therapeuti cs. In Eadi e MJ, Vajda FJ (cds) Handbook of
Experimental Pharma cology. Berlin: Sprin ger- Verlag, 1999.
pp. 173-83.
76 Azparren JE , Vallejo G, Reyes E et al. Study of the diagnosti c,
chloride, haemoglobi n and dia toms in imm ers io n cases.
Forensic Sci Int 1998; 9 1 :123-32.
77 Boles JM, Mabil le S, Scheydecker JL et at. Hyopglycaemia in
salt water near-drowni ng v ictims. Intens ilJe Care jvIed 1988;
14:80.
78 Zhu BL, Ishida K, Quan L et al. Post-m orte m urinary
myoglob in levels with refe rence to the causes of death.
Forensic Sci Int 2001; 115 : 183-8.
79 Pollanen MS. Diato ms and homicide. Forensic Sci Int 1998;
9:29-3 4.
Further Reading
DROWNING: PATHOLOGY AND PATHOPHYSIOLOGY
Colebatch HJH, Halm agyi DFJ. Reflex pulmonary hypoten s ion of
fresh-water aspirati on. J Appl P!JysiolI963; 18:179-8"5.
[A classic paper]
Good en B. Why some people do not drown. Med J Aust 1992;
157:629-32.
Modell JH. The Path ophysiology and Treatment of Dro wning and
Near-drowning. Springfie ld, IL: Charl es C. Thomas, 19 71.
[A classic book]
Pearn JH. Drowning and nea r-drowning. In Black JA (ed.). Paediatric
Emergencies, 2nd edn. London: Butterworths, 1987, pp. 40- 9.
Pearn JH. Pathophysi ology of drowning. iVIed J Aust 1985; 142 :586-8.


DROWNING: EPIDEMIOLOGY
Brenner BA, Smith GS , Overpeck MD. Divergent trends in childhood
drowning rates. ]AMA 1994; 271:1606-8.
Langley JD, Smeijers J. Injury mortality among children and
teenagers in New Zealand co mpared with the United States of
America. Injury PreIJ 1997 ; 3: 195-9.
Lifesa ving Society (Canad a). Tile National [Canada] Drowning
Report, 1997 edn . Ottawa : Lifesav ing Soc. (Canada), 1997.
Lindholm p, Steensberg J. Epidemiology of unintentional drownin g
and near-drowning in Denmark in 1995. Injury Pm; 2000;
6:29-31.
Nixon J, Pearn J, Wilkey [, Corcoran A. 15 years of child drownin g.
A 1967-81 analysis of all fata l cases from the Brisbane Drowning
Study and an J I-year study of consecutive near-drowning cases.
Accident Anal Prev 1986 ; 18:199- 20 3.
Pearn JH, Wong RYR, Brow n J et ai. Drowning and near-drowning
involving child ren. A five-year tota l population study from the
City and County of Honolulu. Am] Publ Hlth 1979 ; 69:450-4.
Royal Life Saving Society Australia. Th e National Drowning Report
2001. Sydney: Royal Life Saving Society Austra lia, 2001,
pp. 1-7.
References I 361
DROWNING: FORENSIC CONCOMITANTS
Boles JM, Mabille S, Scheydecker JL, Garo B, Garee M.
Hypoglycaemia in sal t water near-drowning victims. Il1tens Care
Med 1988 ; 14:80.
Cummings P, Qual) L. Trend s in unintentional dro wning: the role of
alco hol and medical care. ]AMA 1999 ; 281 :2198- 202.
Lunt DW, Rose AG. Pathology of the human heart in drowning. Arch
Pathol Lab Med 1987 ; 111:939-42.
Orlowski JP. Drowning, near-drowning, and ice-water submersions.
Pediatr Ciil1 North Am 1987; 34:75-92.
Pearn J. Aquatics for ep il eptic chi ldren. Aust Poediatr] 1982;
18:255-6.
Pearn J. Drowning and alcohol. Med] Aust 1984; 141 :6-7.
Pearn J. Bart R, Yamaoka R. Drown ing risks to epileptic children. A
study from Hawaii. 8M] 1978 ; 4:1284-5.
Wi lkinson JL. Sudden cardiac death in childhood and adolescence. ]
Paedintr Child Health 1994 ; 30:384-5.
I CHAPTER 19 I
SUDDEN DEATH OF CHILDREN
IN HOSPITAL
Jem Berry
Introduction 362
Definition and frequency 362
Deaths due to natural disease 363
Deaths due to failure to monitor 365
Therapeutic misadventures 366
Deaths due to drug treatment 366
Deaths due to medical devices and procedures 368
Deaths in the dental chair 371
INTRODUCTION
strictly hospital death s, but because they share many fea­
Investigation of the sudden an d unex pected death of a
baby or child in hos pital is a challenge that requires the
combined skills of paediatric and forensic pathology. The
death may be due either to the condition for which the
patient was bein g treated or to so me other unreco g nized
condition. Sometimes death is due to a complication of
treatment or failure of a medical device. Occasionally, sud­
den death is due to an error by professiona l staff. Rarely,
death is due to the deliberate act of a carer.
Sta ff are often upset and understandably defensive, but it
is the pathologi st's role to tease out the facts of what hap­
pened in a non-judgemental manner so that parents receive a
full account of their child's death, and if errors have occurred
then the hospital is able to lea rn from them. A pathologist is
an advocate for the child and his or her parents, not the insti­
tution. A post-mortem report indicating medical error seldom
results in a coronial findin g of negligence, but the penalties
for failing to highlight poor practice may be heavy.
The possible causes of sudden unexpected death in hos­
pital are even more numerous than the many diseases,
drugs and procedures th at children en counter th ere. No
attempt will be made to describe every eventuality that can
lead to sudden death in hospital. The ca uses are broadly
grouped in Table 19.1. Deaths in ' the dental chair are not
Sudden death in newborn babies 371
Accidents 372
Suicide 373
Filicide and homicide in hospital 373
Investigation of sudden unexpected death of
children in hospital 375
References 377
tures with other anaesthetic de aths they are incl uded in this
chapter. The causes of sudden death of newborn babies in
hospital are different from those of older babi es a nd chil­
dren and so will be discussed separately.
Sudden hospital deaths pose interesting lega l questions.
[s death due to a recognized complication of treatmen t for
a lethal disease, natural or unnatural? When is an error
negligent? When is an error so serious that the increasingly
us ed criminal charge of manslaughter is appropriate?
Whatever the answers to these questions, the post-mo rtem
report is likel y to be closely scrutinized , and so the investi­
gation and post-mortem examination must be canied out
to the highest stan dards. Studies have shown the value of
autopsy examina tion in revealing major unexpected find­
in gs t hat might have affected outcome in 9 per cent of chi l­
dren from paediatric inten sive care units (P[CUs), 7 per cent
from emergency departments, 2 per cent from gene ral
wards,I,2 25 per cent of pa ediatric oncology patients,J and
8.5 per cent of paediatric cardiac patients.4 ,5
DEFINITION AND FREQUENCY
The definition of sudden and unexpected death of children
in a hospital context is necessarily subjective because these

Table 19.1 Causes ofsudden death ofchildren in hospital
Due to natural disease
Un recogn ized natural disease
Complications of known natural disease
Haemorrhage
Septic shock
Embolism
Cerebral oedema
Ca rdi ac dysrhythmia
Other
Failure to monitor
Th erapeutic misadventure
Adverse drug events
Errors of drug, dose or rou te of administration
Anaphylaxis
Other adverse reactions
Complications of medical devices and procedures
Dea ths associated with surgery
Surgical dea ths
Anaesthetic deaths
Accidents unrelated to medical care
Suicide
Murder
Filicide
Homicide by carers and others
children are usually already suffering from sign ificant ill­
n ess. Buchino et al 6 considered 'sudden' to be death within
1 hour of clinically apparen t distress excluding time spent in
resuscitation, and 'u nexpected' to mean that no matter how
ill the child, death was no t considered imminent. They esti­
mated that a children's hospital with 200 beds wo uld have
two to four such sudden and unexpected deaths per year.
Li fe -threatening collapse is surprisingly common in
specialist children's hospitals. Two per cent of more than
6000 children admitted to a I22-bed university children's
hospita l during a I2-month period received cardiopul­
monary resuscitation; a susta ined circulation was restored
in two-thirds of these children, 33 per cent were still alive
at 24 hours and 15 per cent at 1 year. The most common
causes of cardiopulmonary arrest, which was not always
unexpected, were respiratory fa ilure and shock. 7
DEATHS DUE TO NATURAL DISEASE
Sudden unexpected death in h osp ital may be entirely nat­
ural. either as a result of a complication of an already diag­
nosed disease or due to a separate unrecognized disease.
In either case, post-mortem exami nation plays a major role
in determining the cause of death.
There are ma ny possible complicatio ns of natural disease
that can cause sudden unexpected death, but a limited
number of final common pathways including haemorrhage,
Deaths: natural diesase I 363
septic shock, embolism, cerebra l oedem a and cardiac dys­
rhythmia. For exa mple, Manto n et al8 described two child ren
who died suddenly and unexpectedly in hospital with
haemolytic uraemic syndrome complicated respectively by
septicaemi a and haemorrhag ic cerebral infarction. Perhaps
surprisingly, unexpected hospital deaths from conditions
associated with su dden death in the community such as
asthm a 9 and epilepsyl° appear to be rare. This is probab ly
because sudden death in these two conditions is associated
with poor disease control and inadequate management of
the acute attack, both of which are more likely to occur out­
side hospital. Similarly, the 'dead in bed ' syndrome in young
persons with type 1 diabetes ge nerally occurs at home and
may be due to untreated hypoglycaemia or ketoacidosis. I I - I)
Fatal Haemorrhage
Fatal haemorrhage as a complication of an underlying con­
dition may be overt or if it is internal it may be completely
unrecognized prior to autopsy. Haemorrhage can occur in
the course of haematological disorders, such as acute
leukaemia, or can fo llow erosion of a major vessel by an
abs cess, tumour, fore ign bodyl4 or vasculitis. IS Spontaneo us
rupture of the spleen with ensuing haemorrhage typically
occu rs in infectious mononucleosis, but ca n occur in acute
leuk aem ia,16 malaria 17 and other conditions where the
spleen is enlarged . Rarely death is due to concealed bleeding
into a large tumour. 18 Haemorrhage in hospital is usually
treatable but may be ovelwhelming, occasionally, resulting
in sudden death.
Septic Shock
Bacteraemia and fungaemia may complicate such recognized
infections as those of the urinary tract, meningitis, endo­
card itis, gastroenteritis, pneumonia, osteomyelitis, infective
arthritis or they may arise from a n unknown cause. There
may be predisposing factors such as chemotherapy, steroid
therapy, cyclic neutropen ia, HN infection, extensive burns,
eczema or aspleni a . Indwelling vascular catheters and pro­
longed infusions are other strong associations. The overall
mortality in one study of bacteraemia and fungaemia in
childhood was 19 per cent, a nd risk factors for death included
neutropenia, hospital-acquired infection, polymicrobial
infection and ineffective empirical antibiotic treatment. The
Single strongest association was septic shock, WHich had a
mortality of 60 per cent. 19
Post-mortem diagnosis rests on suspecting the diagnosis,
recognizing the source a nd co llecting fresh bacterial cultures
from several sites, especia lly a blood culture and a swab of
sp leen. Some treatments such as steroids can completely
suppress the symptoms of disseminated infection in children
and so cultures must be a routine part of the examination of
any child who dies suddenly and unexpectedly. 20
364 I Sudden death in hospital
Toxic shock syndrome occurs in children with burns 21
and after even minor surgical procedures. It has recently
been recognized as a complication of varicella infection
and is sometimes fatal. 22 ­ 25
Fatal Embolism in Childhood
Massive pulmonary thromboembolism in childhood is rare,
but lesser emboli are more common than is generally recog­
nized. 2G ,27 A study at the Hospital for Sick children Toronto
found only eight cases of massive pulmonary thrombo­
embolism causing sudden death in 17500 autopsies (0.05 per
cent),27 but Buck et al 26 found pulmonary emboli in approx­
imately 1 per cent of paediatric autopsies and considered
that it contributed to death in 31 per cent of these cases.
Pulmonary thromboembolism may complicate prolonged
immobilization, sepsis, dehydration,28 cardiac surgery, 29
cardiomyopathy, nephrotic syndrome,30-33 haematological
disorders with enhanced coagulation, occult malignancy
and large vascular malformations. 34 -36 Vascular malforma­
tions may also be the explanation for reports of pulmonary
embolism in conditions such as Proteus syndrome. 37 In chil­
dren, pulmonary emboli may arise from cardiac, mesenteric
or cerebral veins and from the superior and inferior vena
cava, as welJ as from pelvic and leg veins, which are the
usual sources in adults. Vascular catheters are a potent
source of thromboemboli, which may be infected.
Inherited thrombophilias such as deficiency of protein C,
protein S or the presence of a Iupus anticoagulant are very
common in children with thrombotic events, whether or not
there is another condition predisposing to the thrombosis. 3B
Paradoxical emboli or quite small emboli arising in the
left side of the heart, for example in bacterial endocarditis
and rheumatic fever,39 can cause sudden death by coronary
artery occlusion and will be missed unless the coronary
arteries are carefully examined. Fatal coronary artery
embolism is also described in newborns. 40 .41
Tumour embolism typically occurs in nephroblastoma,
either spontaneously42 or as a complication of surgery,43.44
but may also complicate other childhood renal tumours
and tumours with large vein involvement. 45 .4G
Fat embolism occurs after trauma , including non­
accidental injury47 and orthopaedic procedures such as
scoliosis surgery.48 Circulating fat is frequently found in
venous blood after fractures and bone operations without
apparent ill effects, but once cerebral and cardiovascular
symptoms have appeared the mortality approaches 90 per
cent. 49 Reports describe sudden death due to fat embolism
after closed limb lengthening and also after accidental
fractures during manipulation of the hips in cerebral palsy,
muscular dystrophy and severe dermatomyositis. 49 - 52 Fatal
fat embolism appears to be a particular risk for such chil­
dren who have severely porotic bones and expanded
marrow cavities as a result of immobilization or steroid
therapy.
Fat embolism occurs in sickle cell disease, and may be a
cause of the acute chest syndrome in that condition. 53-55
Gray et al 56 cited a 9-year-old boy with sickle cell disease
and acute chest syndrome who was found dead in his
hospital bed.
A related disorder occurs from lipid overload during
intravenous feeding, causing respiratory compromise,
hepatosplenomegaly, abnormal liver function and coagu­
Jopathy.57.58 The condition is generally reversible, but may
be fatal. Several studies have demonstrated intravascular
lipid in the Iungs post mortem in infants who had been
receiving parenteral lipid emulsion, but its significance in
the absence of clinical signs is not known. 59- 62 Fatal lipid
accumulation in the brain has also been reported. 63 Salmon
pink discolouration of the spleen at autopsy is a clue to
lipid overload - the fat can be stained using Sudan black in
routinely processed paraffin sections. However, this syn­
drome should be rare with modern preparations of lipid
and clinical awareness of the hazard.
Gas embolism is mentioned here for completeness, but is
usually a therapeutic misadventure resulting from catheter
accidents,64 open heart surgery, mechanical ventilation 55 or
neurosurgely, especially in the head-up position. Small
amounts of venous air embolism are seldom fatal, but much
smaller volumes of air are necessary to cause death in
infants than in adults and older children. Venous air
embolism has been demonstrated in more than 80 per cent
of children undergoing craniosynostosis repair, although it
resulted in hypotension in only one-third of cases and there
were no deaths in this study66 Topical application, acciden­
tal ingestion or infusion of hydrogen peroxide can result in
fatal embolism of oxygen produced by the action of blood
and tissue catalase.57 ,58 Buchino et al 69 describe how a nurse
mistook a central arterial catheter for a feeding tube and
injected a bolus of air with fatal consequences. Cardiovascu­
lar gas may be seen by computerized tomography (CT) after
failed cardiopulmonary resuscitation, and is thought to be
introduced during venous catheterization. 7o
Diagnosis of fatal gas embolism depends on prompt
autopsy before oxygen can be absorbed,?1 radiology, and
correct technique including inspection of the large veins
for bubbles, aspiration of the heart in situ under water, and
examination of the cerebral vessels before removal of the
brain.
Cerebral Oedema
Cerebral oedema may complicate diffuse hypoxic-':ischaemic
brain damage due to any cause, and so may be erroneously
assumed to be the consequence of cardiorespiratory arrest
rather than its cause. However, it may be a very significant
finding in children who have collapsed suddenly and without
obvious explanation.
Cerebral oedema is a common concomitant of severe
head injury, but it also occurs occasionally in young

children following relatively minor head injuIY and can
cause unexpected death after a symptom-free interval. 72
The mechanism of delayed cerebral oedema in children is
unknown, but in rare older patients it has been associated
with familial hemiplegic migraine. 7J Delayed deterioration
after head injury may also occur as a result of the develop­
ment of subdural or epidural haematoma .74 The effect of
space-occupying lesions such as tumours and abscesses
may be exacerbated by local cerebral oedema, and they may
cause sudden death in hospital from acute hydrocephalus,
brain swelling and coning 75,76
Cerebral oedema is a well-recognized but poorly under­
stood complication of the treatment of ketoacidosis in
young children with diabetes in whom it is the most com­
mon cause of death. 12 ,77-79 Cerebral oedema develops in
about 1 per cent of paediatric cases of diabetic ketoacido­
sis despite what is considered 'best practice' management,
and in one study had a mortality of 21 per cent. 79 About
one-half of these deaths follow sudden respiratory arrest in
hospital. Traditionally, the development of cerebral oedema
has been attributed to rapid changes in the blood glucose
level and administration of intravenous fluid, causing
osmotic swelling of the brain. A case-control study did not
confirm these associations and showed hypocarbia, raised
serum urea and administration of bicarbonate to be risk
factors. 78 Evidence for and against is discussed by Glaser 80
Post-mortem examination occasionally shows that acute
deterioration of consciousness and death in childhood dia­
betic ketoacidosis is due to other cerebral catastrophes, such
as subarachnoid haemorrhage, basilar aliery thrombosis or
multiple thrombi. 81
Cerebral oedema also occurs in other metabolic disorders
such as acute liver failure in which it is a major cause of
death,82-84 and in some inherited metabolic diseases such as
maple syrup urine disease when it can cause sudden death
in hospital. 85
Cerebral oedema should be sought in any hospitalized
child who dies after an unexpected deterioration in con­
scious level. Signs of herniation and coning are a better
guide than brain weight, and in babies these signs may dis­
appear if the brain is put to one side before examination.
Cardiac Dysrhythmia
Cardiac arrest is the dramatic common endpoint of numerous
processes leading to hypoxia or electrolyte imbalance. Some­
times it is due to primary disease of the heali. The manage­
ment of cardiac arrhythmias in children is not always
straightfolVVard and adverse events resulting in death or
brain damage may become the subject of litigation. 86
Sudden death from cardiac dysrhythmia can occur in
myocarditis, cardiomyopathies, myocardial infarction, con­
genital heart disease, and in abnormalities of the conducting
system and electrolyte imbalance. Unrecognized myocarditis
can cause sudden death in patients who are hospitalized
Deaths: failure to monitor I 365
with other illnesses87 and during anaesthesia,88 Myocardial
infarction is more common in childhood than is often appre­
ciated,89.9o occurring in congenital heart disease, anomalous
origin of the coronary arteries, cardiomyopathy, as a result
of cardiac surgery, as a complication of myocarditis and vas­
culitis such as Kawasaki's disease and in cystic t1brosis. 91
Sudden death may follow unrecognized cardiac contusion
due to trauma. Histiocytoid cardiomyopathy affects infants
and young children, especially girls, and, characteristically,
causes recurrent arrhythmias or sudden death, sometimes in
hospitaJ. 92 .93 Dysrhythmias and cardiac conduction defects
are common features of some of the disorders of fatty acid
oxidation, but not typically medium-chain acyl Co-A dehy­
drogenase deficiency.94 Recurrent dysrhythmia as in the long
QT syndrome may be mistaken for breath-holding or
epilepsy95 and the author has encountered an example of
this latter situation causing sudden unexpected death in hos­
pital. Sudden death in anorexia nervosa may be due, inter alia,
to sudden dysrhythmia, and has been associated with
acquired prolongation and dispersal of the QT interval. 96 - 98
The cause of cardiac arrest may be immediately apparent
at the time of post-mortem examination. If it is not and
death appears to have been due to cardiac dysrhythmia then
the heart should be retained with the necessary authoriza­
tion for further detailed examination of the myocardium,
conducting system and coronary arteries. Appropriate sam­
ples should be sent for virology, vitreous electrolyte values,
and retained for possible metabolic and toxicological stud­
ies. The genetic basis of many of the inherited defects of
cardiac conduction is now becoming understood, and
retained genetic samples may sometimes prove as useful as
examination of the heart itself. 99
DEATHS DUE TO FAILURE TO MONITOR
It may not always be easy for a pathologist to recognize
that there has been a failure to anticipate complications
and adequately monitor a sick child, because this some­
times requires intimate knowledge of standards of clinical
care. Failure to monitor is not always due to negligence ­
rather it may be due to lack of facilities or equipment fail­
ure. As a rule of thumb, when intensive care facilities are
adequate an acutely ill child should not die on a general
paediatric ward.
Failure may involve simple observations such as pulse,
temperature, blood pressure, level of consciousness and
fluid balance so that a patient's gradual deterioration is not
appreciated. Failure may also involve more complex mon­
itoring, for example of blood gases, electrocardiogram or
intracranial pressure. Many routine treatments require
monitoring using laboratory tests such as electrolyte values
during intravenous fluid therapy.
In a retrospective review of case notes, Arieff et al 100
described hyponatraemia due to hypotonic infusions after
surgery in 0.34 per cent of children, with a consequent

366 I Sudden death in hospital
mortality of 8.4 per cent. In a 6-year prospective study
they also identified 16 children who were electively hosp i­
talized for prima ry care and who developed severe sympto­
matic hyp onatra em ia. Leth argy, headache, nausea and
sudden respiratory a rrest occurred from 3 to 120 hours
after hypotonic fluid administration, at which time the
average serum sod ium level had fallen from 138 to
115 mmol/L. Out of t hese children 10 died, 5 were left in a
vegetative state, a nd 1 became mentall y handicapped; all
16 of them had radiological or post-morte m signs of cere­
bral oedema.
When a child suddenly dete riorates and is found to have
a low plasma sodium level, ano ther consideration is the
syndrome of inappropriate antidiuretic hormone (ADH)
secretio n. This can foll ow head injury and other intracranial
patholo gy, neurosurgery including spin al surgery, pneumo­
nia , in tens ive chemotherapy, bone marrow transplantation,
anti-epileptic and psychotropic drugs, social use of t he drug
'ecstasy' and many other conditions, and is occasionally
fatal. The key to unravelling this and other electro lyte dis­
orders is expert chart review, supported by measure ments of
post -mortem body weight and the osmo lality/el ectrolyte
levels of vitreous humour, pl asma and urine.
Sometimes, hospital staff do n ot recognize how ill a
child has becom e until too late. Exa mples in myex pelience
includ e a newb orn baby who died suddenly in hospital of
fluid and electro lyte imb alan ce as a result of undiagnosed
duod enal atresia, an institu tionalized child w ith epilepsy
wh o died un exp ectedly from undi agnosed acute .li ver fail­
ure att ributable to sodium valproate, an adolescen t thought
to have gastroenteritis but who died of undia gnosed Addi­
son's disease (a similar case has bee n reported by others), 101
and a child being treated for a head injury who died of
bleeding from an unreco gn ized splenic rupture. It is now
accepted th at so-called delayed splenic rupture is more
often due to delayed dia gnosis rather than delayed haem­
orrhage. 102 ,103
Careful chart review may clarify the sequence of events
lead ing to death , but it is importan t not to include con­
te nti ous judgements best left to others in the post-mortem
report.
THERAPEUTIC MISADVENTURES
Complications of care are an important cause of morbidity
and mortality in hos pitalised patients. A stud y of 1035
consecutive admissions to a PICU showed 2.7 complica ­
tio ns as a resu lt of medical care per 100 PlCU days; overall,
42 per cent were rated as major a nd 36 per cent involved
human error. Complications were ventil ator- , drug- and
procedure-related, involving infectious or involved inva­
sive devices, mostly vascular catheters, and affected 8 pe r
cent of all patie nts admitted to the PICU . Cardiopulmonary
resuscitation was required in six patients and two deaths
were directl y due to complications of treatment. 104
DEATHS DUE TO DRUG TREATMENT
Death may occur as a result of g iv ing the wro ng dru g, errors
in dosage, an inapp ropliate route of admin istrat ion, or
adverse reactions to the medicat ion. Adverse reacti ons to
drugs may be du e to an excessive therapeutic effect (toxic­
ity), in to lerance, side-effects, idiosyncrat ic reactions and
allergy or drug in teractio ns. Severe adverse reactions within
the scop e of this chapter are likely to be due to toxicity,
idiosyncratic reactio ns, anaphylaxis or drug interactions.
The Wrong Drug or the Wrong Route
of Administration
Medication is occasionally given by an inappropri ate route
with disastrous consequences. Sudden death h as fo llowed
intravenous administration of hy pelio ni c sa line,105 hydro­
gen peroxide, G8 and connection of an oxygen cylinder to an
intravenous camtl a.lO GAlcohol injected in mistake for saline
into a line misplaced in the subclavian artery caused the
death of a newborn baby. lol Narsinghani et al l08 describe
the life - threaten ing effect of infusion of polyethylene glycol
electrolyte solution via a misplaced naso gastric t ube into
the trachea. Deaths have also followed errors in in travenous
adm inistration of potassium chloride in children just as in
adul ts. 109 In Phil adelphia, three neonates died when nurses
flu shed their vascular access cathe ters wi th dilu te heparin
that pharmacy personnel h ad accidentally prepared with
bulk potassium chloride sol ution instead of 5 per cent
dextrose. llo Multiple incid ents of inadvertent intrathecal
injectio n of v incristine or daunorubicin, when mistaken for
methotrexate, have been repo rted." I - 1I7 These usually lead
to a fatal myel oe ncephalopat hy desp ite intens ive salvage
th erapy, but not su dden death. Some incidents have resulted
in climinal prosecutions for manslaug hter and so these
deaths are of particular forens ic importance." 8,11 9 Cases in
which doctors h ave been accused of manslaugh te r involv­
ing children are sh own in Table 19.2.
Toxicity Due to the Wrong Dosage
Mista kes in dosage are common in paediatrics because of
errors in calculating and dispensing the correct dos e fo r
body weight in very small babies. 120- 122 This is exacerb ated
by the necessity of using adult formulations not packaged
in paediatric doses ' off-label' because they are not licensed
for use in children. 123 Errors are more likely ' in very ill
patients, very small babies, in em ergency department or
in tens ive care se ttings, an d by trainee medical staff. 124 In a
study of 10 778 medicatio n orders there were errors in 5.7
per cent, and 26 adverse drug events of which five (19 per
cent) were co nsidered preventable. 120 In their study Kozer
et a l 125 reported 20 errors by a factor of 10 in 22 500 doses
in a tertiary children's hospital, potentially resulting in
- ---
- -
 Deaths: drug treatment I 367

Table 19.2 Doctors occust:d of monsloughtt:r involving childrt:n following drug or onot:stht:tic t:rrors: UK 1970-99

Year Patient's age Place of death Nature of problem Outcome of case

in years (sex)

1974 9 (male) Hospital Nasotracheal tube inse rted orally Convicted

and kinked during appendectomy
1978 4 (male) Hosp ital Intravenous dose of methotrexate Acquitted
given intracerebrally
1991 16 (male) Hospita l Vincristine given intrathecally instead Two doctors convicted but
of intravenously acqu itted on appeal
1994 1 (female) Paediatric department Over-correction fo llowing insulin Case abandoned at committal
hypog Iycaem ia test
1994 9 (male) GP surgery Diamorphine given instead of Pleaded guilty
dihydrocodeine during circumcision
1999 14 (male) Den tist's surgery Nitrous oxide given instead of oxygen Convic ted
postoperatively (wrong connection)
1999 12 (male) Paediatric department Vincristine given intrathecally No evidence offered
instead of intraveno usly

From ref. 119.

death or life-threatening toxic effects in 15 children. Koren
et al 121 describe the sudden death of a newborn baby due to
a tenfold error in a digoxin injection, and cite serious
effects from tenfold overdoses of phenobarbitone, pal1­
curonium and salb utamol. Buchino et al 6 quote the case of
a 4-month-old infant who received about 10 times the
intended dose of ketamine because the verbal order was
given in cubic centimetres rather than milligrams. Medica­
tion errors in paediatrics have been reviewed by Sullivan
and Buchino. 126
Adverse Reactions to Drugs
Sudden deaths due to correctly administered drugs are
uncommon in infancy and childhood. However, infants may
react in ways not seen in adults because of differences in
drug metabolism. Immature drug metabolism explains the
grey baby syndrome due to chloramphenicol in neonates,
and probably why children who are under 3 years old
are particularly susceptible to sodium valproate-induced
hepatotoxicity.) 27
Dysrhythmias can occur during treatment with many
drugs including tricyclic antidepressants and as a complica­
tion of cardiotoxic drugS. 128 ,129 For exam ple, dysrhythmia
may follow the intravenous administration of cardiotoxic
chemotherapy such as doxorubicin. Although the great
majority of dysrhythmias in this situation are benign, sudden
deaths have been reported both acutely and as a late effect of
such therapy,130-1 33 Tumour lysis syndrome caused by rapid
destruction of tumour by chemotherapy may also cause
sudden death by hyperkalaemia-induced dysrhythmia,1 34,1J5
Cisapride, widely used for the treatment of gastroesophageal
reflux, has recently been implicated in sudden deaths due
to dysrhythmias,136 although the relationship has been
chaJlenged, 1)7, 138 Dysrhythmia and even death has occurred in
children und ergoing chemical skin peeling using phenol. 139
Respiratory depression caused by sedation used for
minor procedures also causes sudden death in children,
Adverse outcomes from sedation were more likely in non­
hospital settings, using multiple agents and when there was
no monitoring by pulse oximetry, There was no relationship
to the type of sedative or the route of administration, 140 A
syndrome of metabolic acidosis, lipaemia, rhabdomyolysis
and fatal myocardial failure has been linked to long-term
infusion of propofol for sedation of children in intensive
care units,141 - 143 It has recentl y been suggested that this
syndro me is due to impaired fatty acid oxidation caused by
the drug,144,145
Anaphylactic Reactions to Drugs
----------------
Fatal anap hylactic reactions are very uncommon in children.
Pumphreyl46 estima ted that there are about 20 anaphylactic
deaths of all ages per year in the UK, of which one-half are
iatrogenic. A study of anaphylactic reactions in children
gave a fatality rate of 2 per cent. 147 Most anaphylactic reac­
tions in children occur out of hospital and are due to food,148
Anaphylactic reactions in children may be biphasic in 6 per
cent of cases, leading to relapse after apparently sllccessful
treatment. 147,1 49
Almost any drug given by any route can cause a poten­
tially fatal anaphylactic or anaphylactoid reaction, even
something as apparently innocuous as a bovine gelatine­
containing chloral hydra te suppository, 150 Other causes include
vaccines, 151, 152 anti-venom,1 53 leukocyte infusions,154 skin
prick tests, 155-157 cryoprecipitate, 158, 159 gammaglobulin,160
anaesthetic agents, radiological contrast medium,161 ,162 and
parenteral alimentation,1 63 Children with chronic illnesses
368 I Sudden death in hospital
and those requiring multiple procedures including bladder
catheterization may develop hypersensitivities. For example,
anaphylax is on exposure to latex and ethylene oxide occurs
in spina bifida 164,165 and hypersensitivity to ethylene oxide
used for sterilization has been reported in renal dialysis
patients. 166,167 Children with tumours may become sensitized
to chemotherapeutic agents such as asparaginase.'68.169
Patients with cystic fibrosis may become sensitized to anti­
biotics, pancreatic extract and vascular catheters. 170, 171
Post-mortem findings in anap hylaxis are often non­
specific, but in children may include oedema of the respi­
ratory tract and asthma-like changes in t he lungs. 146 .172 , I73
Deaths due to an ap hylaxis are very rare, but failure to
diagnose anaphylaxis post mortem is often due to failure to
think of the diagnosis and retain the appropriate specimens
for measure ment of specific immunoglobulin E (IgEl and
mast cell tryptase, altho ugh the latter may give both false­
positive and false-negative results. 173 - 17 6
DEATHS DUE TO MEDICAL DEVICES
AND PROCEDURES
Vascular Catheters and Infusions
The insertion of central venous lines is associated with signif­
icant mortality and morbidity at all ages, despite the use of
softer catheters and attempts to educate doctors about the
risks. m ,178 Sudden death may occur as a result of cardiac
tamponade by blood, intraveno us fluid or parenteral alimen­
tation fluid, or due to pneumothorax, hydrothorax or
haemothorax. I07 Sudden death has also been caused by dys­
rhythmia provoked by a guide wire 107 and fatal coronary sinus
thrombosis du e to a misplaced central venous catheter. 179
Symptoms may occur almost immediately after the catheter is
inserted or be delayed. They are so frequently miSinterpreted
that tamponade carries a high mortality. 180 The frequency of
pericardial effusion or tamponade from long lines in neonates
found in a UK-wide survey was 1.8 per 1000 lines with a mor­
tality rate of 0.7 per 1000 lines. Two-thirds of the fatal cases
were not diagnosed until post-mortem examinat ion. 181 In
neonates the catheter does not necessarily have to perforate
the atria l or ventricular myocardium to produce tamponade,
fluid being abl e to permeate the loose cardiac muscle fibres if
the tip becomes wedged by thrombus.182-184 The location of
catheters should be documented before autopsy by routine
radiology, and the identity of unexpected fluid in the pericar­
dial sac or pleural cavity should be confirmed by chemical
analysis. Complications of intravascular catheters in neonatal
intensive care have recently been reviewed. 185
Small children are particularly vulnerable to fluid over­
load if an infusion device fails or is incorrectly set and
delivers perhaps a day's fluid in just a few minutes. Mobile
phones have been said to interfere with a number of med­
ical devices including infusion pumps, but the risk for cur­
rent generation phones has probably been exaggerated
unless the device is very close.186- '90 Two-way radios used
by emergency crews are more likely to interfere with med­
ical equipment than cell phones. '91 Addi tio nal hazards of
infusion include administration of the w rong fluid 105 or
fluid contaminated by bacteria. 192 Culture negative intra­
venous fluids contaminated with endotoxin caused the
death of 36 neonates in an incident in Brazil. I93 Blood
transfusion can be complicated by sudden dea t h due to
transfusion reactio ns, either as a result of faulty grou ping
and cross-matching or administration of blood intended
for another patient. Patients who have received multiple
transfusions of blood products are particularly at risk of
adverse reactions to further tra nsfusions. Frontela et al 194
describe an immunofluorescent technique developed to
establish the diagnosis of mismatched blood transfusion
using paraffin-embedded tissue from a child at more than
2 years after death . Sudd en death has followed exchange
transfusion usin g blood damaged by overheating.
Nasogastric and Orogastric Tubes
There are few reports of death following attemp ted gastric
intub atio n in children despite the potential hazards of acci­
den ta l tracheal intubation 108 and creating false passages. A
nasogastric tube that entered the cranial cavity via a frac­
ture was probab ly a peri mortem event, rather than the cause
of death.'95 Gastric rupture has been described both from
gastric intuba tio n and washout l96 and after emesis induced
with ipecac. 197 Saline used as an emetic or for gastric lavage
can cause fatal salt poisoning.198.1 99 Induced em esis is no
longer recommended as a routine treat ment for poisoning
in children. 20o
Endotracheal Tubes and Tracheostomies
Unexpected death can occur as a result of fail ed tracheal intu­
bation, accidental placement of the tube in the oesophagus,
blockage of the tube with mucus, accidental extubation 201 or
rupture of the trachea. 202,2o3 In one stu dy in a PICU the rate of
accidental extubation was 1.4 per 100 ventilator days. 104 Tra­
cheal tubes may become displaced into a tracheo-oesophageal
fistula and cause sudden collapse during surgery.69 Tra­
cheostomy tubes are also prone to blockage and displacement,
which very rarely cause death even in hospi ta l.204-2o9
Lumbar Puncture
Sudden de ath may occur as a result of cerebellar herniation
and brainstem compression if lumbar puncture is carried out
in t he presence of raised intracran ial pressure. Tbis was a
particular hazard when lumbar puncture was carried out for
the diagnosis of bacteria l meningitis, but this is no longer
recommended as a routine emergency investigation.210-2 12
---

Conin g in meningitis may sti ll occur despite a noml al CT
scan 21J or without lumbar puncture. 214 Herniation may be
reversed by vigorous treatment with hyperventilation and
m annitol, a nd so it may be hard to confirm post mortem
unless there a re irreversible signs, such as necrosis of the
cerebella r tonsi ls. 2l1 Coning may also cause sudden death in
hospi tal when there is an undiagn osed space occupying
intracrani al les ion, such as a tumour or abscess,l5,76 and has
been describ ed in childhood acute leukaemia with central
nervous system involvement. 21 5
Ventriculoperitoneal and Ventriculoatrial
Shunts
Shunt failure in treated hydrocephalus has been a significa nt
problem and continued to be so in the 1990s.216 Children
with shunted hydrocephalus may die sudden ly, sometimes
during in vest igation in hospital for a pparently minor symp­
toms.217 Acute deterioration due to obstruction of the sh u nt
is often preceded by a history of headaches, and the child
may be found de ad in bed.216 However, symptoms may be
non-specific, and shunt failure is not invariably accompa­
nied by v entricu lar enlargement. 218 Death can also occur
due to recurrent pulmonary embolism from vent riculo atrial
shunts a nd is usually heralded by respiratory sym ptoms or
pulmonary hypertension .217 •219 A ventriculo atria l shunt has
caused coronary sinus thrombosis, myocardial infarction
and fatal coll apse in hospital. 22o Byard 221 describes sudd en
death in a child with a ventriculoperitoneal shunt that had
perforated the transverse colon, leading to meningitis.
Post-mortem examinatio n should include careful ex am­
inatio n of the s ite of th e shunt for swellin g - both ends of
the sh unt in situ and the va lve for signs of obstruction,
examin ation of any intravascular component for t hrombus
and careful cultures to rule out infection. The help of a
neurosurgeon may be necessary to fully eval uate the func­
tion of the shunt and valve.
Other Devices
Pacemaker failure is a rare cause of sudden death in hospi­
tal but in all cases when a pacemaker is present it should be
saved, with its lead intact for expert ex am in ation, and the
site of the tip should be examin ed histologically.222-225 It is
important not to let the bare tip of the wire touch the bare
body of the dev ice because electrical settings may be lost.
Deaths Associated with Surgery
and Anaesthesia
Children are affected by the same surgical and anaesthetic
accidents as adults but have a higher operative mortality
because their sma ll size and immature physiological
~
-
Deaths: medical devices and procedue s I 369
responses allow a smaller margin for error. Problems posed
by their vulnerability to heat loss, narrow a irway, small
extracellular vo lume and age-depe nd ent responses to anaes­
thetic agents make paediatric a naesthesia a field for special­
ists. Paediatric surgery often involves children with multiple
complex problems, thus addin g to the operative risk.
Un expected deaths associated with surgery may be due
to the patient's original condition. In that case questions
arise about the appro priateness of the decision to operate,
and the adequacy of attempts to prepare a nd stabilize the
child for surgery. The 1999 report of the UK Nationa l
Confidential Enquiry into Perioperative Deaths found that
most dea ths within 30 days of sur gery in children were in
those with congenital anomalies, necrotising enterocolitis,
tumours or trauma. These children gene rally had respira ­
tory or cardiovascu lar problems in addition to their under­
lying diagnosis prior to surgery. In total, 16 per cent of
deaths took place on the day of surgery. 226 Lack of experi­
ence and inad equate superv ision of junior staff are recu r­
rent themes in studies of anaesthetic deaths.
When unexpected death is due to the surgery itself, the
cause is often obvious at post-mortem examination. In con­
trast, anaesthetic deaths a re usually du e to respiratory failure
or cardiac arrest, whic h may leave no trace, and so the ca use
often ca nnot be determined by autopsy examination alone.
Pathologists cannot have the expertise to assess all aspects of
an anaesthetic and should seek independent expert advice.
Deaths Due to Surgery
Deaths due to the surgery itself are either due to general haz­
ards of surgery, such as acute haemorrhage, or specific com­
plications of pal1icular operations beyond tile scope of this
chapter. Out of 289 paediatlic pelioperative cardiac arrests,
48 per cent were considered to be unrelated to tile anaes­
tIletic; of these, 22 per cent were cardiac surgery patients who
could not be weaned off by-pass, anotller 17 per cent suffered
ca rdiac arrest as a result of u ncontroll able surgica l haemor­
rhage, 22 per cent had miscellaneous ca rdiovascular eve nts, 4
per cent had air embolism and 4 per cent had complicatio ns
of central venous li nes.227 Complications due to vascular
cath eters inserted prior to the commencement of surgery are
similar to those already described (see 'Vascula r catheters and
infusions', above), but misplaced catheters may have disas­
trous effects duling surgery if tIley cause intrave nous anaes­
thetic agents or resuscitative fluids to be delive red into the
wrong compartment.
Cardiac arrest may follow su rgical traction o~ the viscera
(the viscerocardiac reflex) or inte rference with the eye (the
occulocardiac reflex). 228 Bietti 229 reported such a case of a
child operated on for recess io n of the medial rectus w ho s uf­
fered a fatal cardiac arrest. Pathologists investigating indi­
vidual cases should not accept such possible explanations
uncri tica lly, and w itllout excluding every other possi ble
ca use o f death.

370 I Sudden death in hospital
Examination of children who die after cardiac surgery
requires specialist knowledge of cardiac morphology and the
surgical techniques used to treat individual malformations.
Cardiac catheterization in infants and children carries a small
(about 0.14 per cent overall) risk of death, both from diagnos­
tic and interventional procedures, as a result of complications
such as perforation of cardiac chambers and bradycar­
dia. 2JO-232 As a general rule, surgical mortality increases with
the length of time on cardiopulmonary by-pass, and a long
by-pass time is often an indication that problems were
encountered during the surgery. Difficulty weaning the
patient off the pump usually reflects myocardial damage due
to prolonged cardioplegia,233 although if the patient dies
intraoperatively or soon postoperatively there may be no
morphological evidence of myocardial necrosis. Myocardial
necrosis may also be present preoperatively in congenital
heart disease. The risks of sutures obstructing branches of the
coronary al1eries or impinging on the conducting system are
common to several procedures. Allwork 234 gives helpful
advice about likely surgical difficulties encountered in spe­
cific ca rdiac operations and their recognition post mortem.
Complications of Anaesthesia
In one US teaching hospital 40 per cent of all anaesthesia­
related perioperative cardiac arrests were medication­
related events, 20 per cent due to complications associated
with central venous access, 20 per cent were attributed to
problems with airway management, and 13 per cent were
due to unknown or possibly vagal reaction. The risk of
death related to anaesthesia-attributable perioperative car­
diac arrest was 0.55 per ]0 000 anaesthetics. 235 A study of
infants and children found an anaesthetic-related cardiac
arrest rate of 1.4 per 10 000 anaesthetics, with 55 per cent of
all arrests occurring in babies who were under] year of age
and medication and cardiac causes accounting for 69 per
cent of all cases. Cardiovascular depression from halothane,
alone or in combination with other drugs, was responsible
for two-thirds of all medication-related cardiac arrests 22 7 In
a study of over 700 000 anaesthetics, Morita et al 236 found
the incidence of critical peri operative events such as cardiac
arrest and severe hypoxaemia to be greatest in babies who
were less than 1 month old, and that these babies also had
the highest incidence of critical events related to anaesthetic
management. However, although their mortality during and
within 7 days of anaesthesia was also the highest of any age
group, this was predominantly due to coexisting disease,
and no death was attributable to anaesthesia in children
under 5 years of age. Causes of anaesthetic deaths in older
age groups included anaesthetic overdose, toxic effects of
local anaesthetics, improper airway management, incom­
patible blood transfusion and errors in spinal anaesthesia.
Another study found 'major' perioperative events, which
included cardiac arrests and death, to be most frequent in
babies who were less than 1 month of age, although these
were the sickest group.237 The remarkable safety of paedi­
atric anaesthesia in a specialist setting is shown by a report
of 24165 paediatric anaesthetics in which there was a rela­
tively high incidence of adverse events in neonates, but no
anaesthetic-related deaths in any age group.2JB
Cardiac arrest during anaesthesia or surgery is most often
due to hypoxia, which has many causes. Induction of anaes­
thesia may be complicated by broncho- and laryngospasm,
failed intubation, or leakage around the endotracheal tube,
causing inadequate ventilation. The tube may be kinked, or
become blocked by a defective inflatable cuff or mucus.
Accidental oesophageal intubation is not necessarily fatal
unless spontaneous respiration is prevented by muscle relax­
ants, and may sometimes be suspected from erosions in the
oesophageal mucosa even when the tube has been removed
post mortem. Delayed detection of accidental oesophageal
intubation is rare, but in one study was more common in
infants and in association with emergency surgery.239 Many
anaesthetic agents cause respiratory depression, either in
therapeutic doses or in over-dosage.
Pulmonary aspiration of gastric contents is a much­
feared complication of general anaesthesia. However, it
occurred in only 0.04 per cent of 56] 38 consecutive
patients who were less than 18 years old and who under­
went 63 180 general anaesthetics for procedures performed
in all surgical specialities at the Mayo Clinic over a period
of 12 years. The risk of aspiration was highest in emer­
gency procedures, and those involving children who were
younger than 3 years of age with bowel obstruction or
ileus. Of the 24 patients who aspirated, only three required
mechanical ventilation and none died .240
Faulty anaesthetic equipment can cause fatal hypoxia.
Incorrect plumbing of piped gases, failure of the oxygen
supply, incorrect f10w rates or leaks in the anaesthetic cir­
cuit, faulty connections and accidental disconnections have
all caused unexpected anaesthetic deaths.
Cardiac arrest may also result from dysrhythmias due to
anaesthetic agents such as halothane, electrolyte abnor­
malities or vagal inhibition caused by stimulation of the
respiratory tract by irritant gases or passage of an endo­
tracheal tube. Ventricular fibrillation results from release of
adrenaline in patients who are inadequately sedated.
Collapse or death during anaesthesia and surgery in chil­
dren may be the first manifestation of an unsuspected inher­
ited or developmental abnormality. Death has been caused
by perioperative fasting in a child with unrecognized very­
long-chain acyl Co-A dehydrogenase deficiency2 4J Anaes­
thesia may precipitate cardiac arrest in children with
unrecognized congenital myopathy242-244 or cardiomyo­
pathy.227.245 Cucchiaro and Rhodes 246 reported a 9-year-old
boy who was found to have long QT syndrome after he
suffered ventricular fibrillation and cardiac arrest following
accidental intravascular injection of bupivocaine. Surgery is
a potential hazard in children with sickle cell disease.247
McGarry and Duncan 248 described four sudden deaths asso­
ciated with surgery due to red blood cell sickling in children

with undiagnosed sickle ceLl trait. Acute intermittent por­
phyria can present for the first time in childhood as a result
of an anaesthetic and cause death.249 Cardiac abnormalities
have recently been highlighted as a cause of otherwise unex­
plained sudden deaths during anaesthesia and surgery.250
Malignant hyperthermia is a condition characterized by
sudden onset of tachycardia, acidosis, muscle stiffness, and a
rapid rise in temperature, which, in susceptible individuals,
is precipitated by suxamethonium and certain inhalation
anaesthetics, including halothane. It is inherited as an auto­
somal dominant gene with variable penetrance, and recent
work has shown mutations of the RYRl gene in 30-50 per
cent of susceptible subjects. 251 The incidence of malignant
hyperthermia in children is greater than that in adults, per­
haps because of the anaesthetic agents used. Dantrolene is a
specific treatment. 252 At post-mortem examination, the mus­
cles may be paJe. Microscopy shows focal muscle necrosis
and myoglobin in renal tubules, which may be necrotic. 253 ,254
The fulminant form still occurs, although it represents a fail­
ure of diagnosis and treatment. 255 Anaphylactoid reactions to
neuromuscular blocking agents are quite distinct from malig­
nant hyperthermia, but may also be fata1. 256
Deaths in the recovery room are uncommon and are
likely to be due to the delayed effects of surgery such as
haemorrhage or failure to reverse the anaesthetic, resulting
in respiratory depression. The airway may become
obstructed by mucus, blood clot or a forgotten throat pack.
About 1 in 1500 children have an inherited deficiency of
cholinesterase, which greatly prolongs the effect of the
muscle relaxants succinyl choline and mivacurium, thus
delaying the onset of spontaneous respiration after anaes­
thesia. 257 ,258 Postopera tive cardiac arrests in recovery or
the intensive care unit are rarely fata1. 259
Excessive volumes of intravenous fluid cause postoper­
ative pulmonary oedema, which most commonly presents
without warning as cardiorespiratory arrest with normal
electrolyte values. Post-mortem examination shows pul­
monary oedema only, with no other cause of death.260
Postoperative hyponatraemia and cardiorespiratory arrest
due to brain swelling is often caused by administration of
intravenous fluids that are low in sodium, possibly exacer­
bated by inappropriate secretion of ADH, and autopsy
shows cerebral oedema. 100,261 A meta-analysis confirmed
the association with hypotonic solutions,262 and a survey
of anaesthetists has shown that children in the UK are still
at risk of this complication, especially in non-specialist
hospitals. 263 Irrigation of the operative site can lead to sub­
stantial unrecognized absorption of fluid.
DEATHS IN THE DENTAL CHAIR
Most deaths in the dental chair are anaesthetic related. The
public expectation of general anaesthesia as an option for
minor dentistry and the historical willingness of UK
dentists to administer anaesthetics while also carrying out
Sudden death in the newborn I 371
dental procedures resulted in 100 deaths in the UK from
1970 to 1979 264 and 42 from 1980 to 1989,265 one-half of which
occurred in children. 266 The number of dental anaesthetics
administered in surgeries and those given by operator­
anaesthetists has declined as a result of a flurry of reports
and stricter guidelines. 267 ,268 In the late 1990s approxi­
mately 300000 patients underwent general anaesthesia for
minor dental procedures, with around 2- J deaths per year,
usually in fit young adults or children who presented with
no particular anaesthetic risk,269
Historically, about one-half of deaths in the dental chair
were due to respiratory causes and one-half were due to
sudden cardiac arrest. Causes included unobserved respira­
tory depression due to premedication and multiple anaes­
thetic agents, sometimes in excessive doses, Cardiac arrest,
due to either hypoxia or anaesthetics, such as halothane, is
hard to manage in a dentist's surgery.
The use of local anaesthetics in dental surgery, as in
general surgery, can cause collapse and even death due to
rapid absorption, overdose, intravascular injection or ana­
phylaxis. The frequent simultaneous injection of a vaso­
constrictor such as adrenaline can also cause adverse
effects. Accidental over-dosage with local anaesthetic is
particularly easy in children, especially in those who are
undergoing multiple procedures on the same occasion. This
possibility is exacerbated by the use of multiple agents
with names ending in '-caine' requiring different doses.
Overdose of local anaesthetic can be fatal in both children
and adults and causes nausea, anxiety, excitement, vomit­
ing, convulsions and cardiorespiratory arrest. 270 Most cases
of 'anaphylaxis' that are attributed to local anaesthetic
agents are in fact probably due to inadvertent intravascu­
lar injection.
Fortunately, death in the dental chair has become rare
since stringent standards have been set for the administra­
tion of general anaesthetics in dentists' surgeries, but each is
particularly tragic because it affects a fit child undergoing
relatively trivial procedures. Post-mortem examination
should include not only toxicological studies, but also the
possibility of unrecognized cardiac, skeletal muscle or meta­
bolic disorders that were exacerbated by the anaesthetic.
Very rarely, oral surgery itself may cause death. Fatal
intracranial haemorrhage has followed accidental perfora­
tion of the base of the skull during surgery on the temporo­
mandibular joint in a child.271
SUDDEN DEATH IN NEWBORN BABIES
A detailed discussion of the causes and investigation of
sudden unexpected deaths in the early neonatal period is
beyond the scope of this chapter and is covered in standard
texts. 272 ,273 As most such deaths will turn out to be due to
natural causes they are best dealt with by perinatal pathol­
ogists, with a forensic pathologist in attendance only if
there are forensic concerns.
372 I Sudden death in hospital
There have been few systematic studies of the incidence
and causes of sudden death in the first week of life. Early
neonatal sudden hospital death of newborns considered pre­
viously healthy acco unted for 11 per cent of total neonatal
mortality in a Swedish study;274 a more recent study gave
the rate of 'early neonatal sudden death syndrome' as 0.14
in 1000 live births, with approximately one-third of infants
dying within the first hour after birth.275 Quite similar fig­
ures were found in a French study that noted that many
deaths occurred at night, and that some babies had signs of
illness that were not recognized or acted upon. 276
Sudden death of newborn babies, either in special care
baby units or maternity wards, is usually due to age group­
specific causes, or those related to labour, delivery and the
transition from intrauterine to extrauterine life. In term
babies such likely causes include delayed effects of birth
asphyxia or trauma, infection, genetic metabolic disease
and malformations, especially ductus-dependent cardiac
malformations such as hypoplastic left heart syndrome. 277
In pre-term babies who are receiving intensive care, addi­
tional consideratio ns are complications of prematurity and
adverse effects of procedures and treatment.64.278
Early neonatal death due to frank birth trauma is now
rare, thankfully, but minor trauma such as a fractured clavi­
cle, excess moulding of the skull or minor tearing of the falx
may indicate that the delive ry was not quite as simple as the
clinical records suggest. Unrecognized birth asphyxia may
result in sudden unex pected death in the 'lying-in ' ward.
Most fatal malformations are easily recognized at post­
mortem examination but careful technique is necessary if
subtler anom alies, such as those of the upper airway 279 and
larynx, coronary arteries or pulmonary venous drainage,
are not to be overlooked. Minor atrial septal defects a nd
probe patency of the ductus arteriosus are not causes of
death in the neonatal period.
Infection may be acquired before, during or after birth .
The effects of infection are exacerbated by lack of mater­
nally acquired specific antibodies and immaturity of the
immune system in newborns. Bacterial infection is tradi­
tionally divided into early and late sepsis, the former prin­
cipally caused by enterovaginal organisms and the latter
by nosocomially acquired bacteria. Group B streptococcal
infection in newborns is usually an example of the former,
and can cause sudden collapse and death with only mini­
mal signs at post-mortem examination. Viruses such as
herpes simplex and the enteroviruses cause devastating
disease and sudden death in newborn babies, especially
when they have no passively acquired maternal antibodies.
Severe infection in small babies is not always accompanied
by symptoms obvious to the new mother.
Numerous inborn errors of metabolism can cause death
in the neonatal period and are listed in appropriate text­
books. z8o Some, such as galactosaemia, are precipitated by
the intro duction of particular comp ounds into the diet. It is
worth noting that medium-chain acyl Co-A dehydrogenase
deficiency (MCAD), which is familiar as a cause of sudden
death later in infancy, may also cause collapse and death in
the early newborn period.
Sudden unexpec ted death in hospitalized infan ts with
bronchopulmonary dysplasia after prolonged mechanical
ven tilation has been reported as a significant cause of late
mortality, despite appropriate monitoring and prompt car­
diopulmonary resuscitation. 281
When a newborn infant is found dead in hospital in bed
with its mother then the possibility of accidental asphyxia
arises. A familiar scenario is tha t of a mother, perhaps
exhausted by childbirth, who falls asleep while breast­
feeding her new baby and wakes up to find it lifeless. 282
Byard and Burne1l 283 have demonstrated the ease with
which the huma n breast can cause accidental asphyxial
episodes, but such a possibility should be approached with
extreme sensitivity in the context of a neona tal death.
Similarly, the possibility of neonaticide has far-reaching
consequences, both for the mother and hospital staff, and
should be approached with care. Neonaticide is more com­
mon outside hospital but remains a consideration when
there is no medical explan atio n for a newborn's death in
hospital , and there are other social and medical pointers
such as concealment or denial of pregnancy.
Although no explanation may be found in a small pro­
portion of babies who die suddenly and unexpectedly in
the early neonatal period, such unexplained cases are
exceptional in the practices of specialist perinatal patholo­
gists. Careful review of the pregnancy and labour, com­
bined with a detailed post-mortem examination, including
appropriate laboratory investigation and examination of
the placenta, will provide an explanation in most cases.
ACCIDENTS
Fatal, non-medical accidents are seldom reported in hospi­
talized children and are probably rare, although patients
and visitors are exposed to many of the same hazards as
exist at home. For example, a 21-month old child was elec­
trocuted in hospital by an uncovered lamp switch which he
took into his mouth,284 and an infant aspirated the nipple
of a makeshift pacifier in a hospital nursery.28S
A study of eight hospitals, includin g specialist children's
hospitals, over an 18-month period reported 781 non­
iatrogenic accidents involving patients and visitors under
16 years of age. Overall, 41 per cent of accidents to inpa­
tients happened when their parents were present. Most
involved falls, slips, striking, and scalds; three were seri­
ous, resulting in fractures or bruising in children with
bleeding diatheses 286 There were no fatalities, although
one entrapment was potentially lethal. Falls from beds
and cots in hospital usually result in only minor injuries
even though beds are often higher than those at home,
although there are occasional skull fractures.287.288 Toys are
second only to beds and cots as a cause of equipment­
related hospital accidents, and attention has been drawn to

th e dan ger to other children of toys brought into the hos­
pital by visitors. 284
Finberg et af89 described the death of 6 out of 14
infants in a hospital nursery, who were accidentally poi­
soned when their formula feed was made up using salt
instead of sugar. 289 ,290 The most severely affected survivo r
had a serum sodium of 274 mmol/L and made a good
recovery. A similar episode occurred in Austria in which
five newborn babies died. 291
SUICIDE
Of the 30 000 suicides per year of all ages in th e USA, 5-6
per cent occur in hospitals. 292 Suicidal ideation and action s
are well recog nized in children and yo ung people in
psychiatric wards, a lthough actual suicid es app ear to be
uncommon. 293 ,294 Noren et al 295 describe a 15-year-old
mental health patient who jumped from the third floor of a
hospital and was found to have incidental lymphocytic
myocarditis at post-mo rtem exa mina tion. Children with
chronic or terminal diseases such as cystic fibrosis may
occasionally take their own lives. Repo rts of such deaths in
hospital are rare, but the possibili ty should be kep t in mind.
Bya rd reported 296 the case of a 14-year-old girl with cystic
fibrosis who injected hair conditioner via an intravenous
infusion pump intended for systemic antibiotics, although
suicidal intent was not proven in this case.
FILICIDE AND HOMICIDE IN HOSPITAL
Babi es and children in hospital may be harmed by their
paren ts, hospi tal staff or visitors. The motive may be
uncl ear, but examples include mercy killing, abuse due to
Mun chausen 's syndrom e by proxy, and frank psychosis .
Children may be admitted to hospital as a result of
deliberate harm that mayor may not have been recognized
for what it is, such as physical injury by a parent or simu­
lated illness. Physical abuse may then continue in hospital,
and has been clearly documented by covert video su rveil­
lance.297 Rarely, a paren t will go to extraordinary lengths
to cause real or simulated illness in their child in hospital,
using methods such as suffocation , administration of drugs
and poisons, tampering with samples, interference with
equipm ent and introducing infected material into infusion
fluid s.298-301 It has been said that the techniques used are
limited only by the intell igence or imagin ation of the per­
petrator. Some of these methods are capable of resulting in
sudden unexpected death, but it is believed th at this is not
usually the parent's objective and that such deaths in hos­
pital are rare. Very rarely a parent is implicated in the
'mercy ki lling' of a severely handicapped child . The scope
for filicid e in hosp ital has been increased by the greater
availability of rooms in which parents can stay overnight
with their child.
Filicide/homicid e in hospital I 373
In recent years it has been recogni zed that rare devian t
medical professional s may target vulnerab le pati ents caus­
ing serial collapses and deaths. Frank discussion of this
topic is hampered by the historical difficulty of obtaining
sustainable conviction s. How ever, children have fea tured
prominently amon g these so-called carer- assoc iated serial
killings (CASKs) because, like the elderly and tho se in
intensive care units, they are vulnerable and un able to
speak for themselves. Some cases invo lv in g children are
de ta iled in Table 19.3.
In a comprehensive review on which much of this
account is based, Forrest 302 esti mated the incidence of
CASKs to be approximately 1 per million carers per year.
Nurses are the professional group most often involved, per­
haps because they are the most numerous; however, there
are many instances of murderous doctors. 303 Park and
Kh an 304 list 13 cases involving nurses from the UK and USA
between 1975 and 1998. Perpetrators may be particularly
attracted to medical work because of the opportunities it
affords, and may be motivated by a desire to control events
or to show themselves in a good light in emergency si tua­
tions. In one adult series the motive appears to have been as
trivial as wishing to win an unofficial sweepstake predicting
the time when patients would die.
These deaths are extremely difficult to recognize because
medical professionals often have knowledge of, and access
to, methods that are almost undetectable. The victims are
pati ents a nd so post-mortem examination shows natural
disease that might have caused death and does not ra ise
concern. The alarm is usually raised by cli nical staff, but
typically not until the number of patients affec ted by col­
lapses or death has reached double fig ures. The trigger may
be the rec ognition of a train of unusual events or increased
mortality, or a member of staff may become suspicious after
witn essing some strange behaviour by the perpetrator.
Yorker3 05 suggested a number of clues to early recognition:
• a significant increase in cardiopulmonary arrests, or
deaths or both , particularly in patients not tho ught to
be in immedi ate danger;
• an unusually high rate of successfu l resuscita tion;
• multiple events in the same patient;
• events occurring more often in a particular shift.
It is notable that in so me cases, the staff themselves
have id entified a 'death shift' or a particular member of
sta ff involved in a large number of events, but their con­
cerns have not been taken serio usl y by senior personnel.
As soon as con cerns are raised they must be shared with
senior medical and mana geria l staff, w ho will carry out
further investigations and take the decision to inform the
authorities. It would be wise to draw on the experience of
independent experts and those who have investigated sim­
il a r cases before at an early stage.
Possi ble action s that may be taken when such a death
occurs are shown in Table 19.4. The over-riding immediate
priority after protecting other children is to secure the
374 I Sudden death in hospital

Table 19.3 Some well-documen ted corer-associated serial killings involving children

Cas.e Location Setting Likely agent(s) Deaths Outcome

Prelimin ary Toro nto Paediatric cardiac Dig ox in Uncertain; increased Case against nurse Susan
hearing 313 ward of children's mortality (x 4) on unit Nelles dismissed
hos pital over 9 months; nurse No proceedings taken
charged with four murders aga inst another known
only as nurse 'A'
Roya l Commission found
that out of 33 deaths 8
were certainly due to
digoxin and 15 others were
suspicious
Texas v. San Antonio Paediatric inten sive Hep arin, Su bsta nti a I exc ess Convicted and sentenced to
Genene ca re unit and suxa methonium mortality in PICU; 99 yea rs and 60 years,
Jones3 14 paedi at ric pri vate charged with one murde r respect ive ly, on each cou nt
practice an d one assault
R v. AllitP15 Grantham, UK Paed iat ri c wa rd s of Lign oca ine, in sulin , Charged with four murders Convicted
general hospital potassium ch loride and nine assau lts Clothier Inquiry316

Table 19.4 Some immediate actions that may be taken when a
child dies suddenly and unexpectedly in hospital in suspicious
circumstances to preserve evidence prior to the post -mortem
examination
Control and limit access to the room
Institute log of perso ns entering the room
Secure all samples and equipment (includes sharps containers
and waste)
Ensure that no items on body are removed (vascular lines,
end otrach eal tubes, pacemakers, etc.) and postpone
'la st offices'
To preserve settings and memory, do not switch off
electronic equ ipment
Secure clinical records, fluid charts, drug records and records
of resuscitation
Ask all staff present at the event to make an immediate
detailed written record of the event
Notify laboratories to save all earlier samp les (haematology,
blood bank, chem istry, microb iology, etc.) until
notified otherwise
Keep any videotape from security cameras and any
op erative procedures
Inform coron er, medical director, management or police
as approp riate
Consider taking immedi ate sa mpl es, subject to
coroner's permission
scene, sampl es and equipm ent, to prev ent anyo ne remov­
ing any tubes or other equipment from the body, and to
record who was present and invi te them to write down
their recollection of events as so on as possible. The imp or­
tance of preserving sa mples can not be overemphasized.
Vital information may be obtained from used syringes,
'empty' ampoules, and discarded samples for bl ood gas
analysis, all of wh ich may have to be recovered from the
sharps bin, taki ng appropriate precautions against injury.
Old samples already in the laborato ry and those from other
patients who may have been affected must also be secured
against being routinely disca rd ed.
Additional samples must be taken at post-mo rtem
examination including tissue from any possible injection
sites. Toxicological analysis will be informed by circum­
stances, also remembering that some of the agents used in
prev ious cases, such as insulin and heparin, are no t
included in a routine toxicological screen . Some cases have
involved inj ection of air.
Confinnation that serial killin gs have occurred and iden­
tification of the perpetrator often depend heavily on ep idemi­
ologica l evidence of increased death rates in a particular part
of a unit at particular times when an individual was known
to be on du ty (so me drugs with a delayed action, such as
digoxin, may be used to produce clinical deterioration and
death after the perpetrator has left the wa rd). When finally
identified, the perpetrato r is often found to have history of
medical and employmellt problems. Such is the difficulty of
identifying these individuals that they might never be discov­
ered if they move around. In retrospect, some are found to
have offended before in another centre.
Prevention of such rare events as CASKs is 'extremely
difficult. The Clothier report into the case of Beverley AJlitt
noted that 'a determi ned and secret criminal may defeat the
best regulated organization in the pursuit of his or her pur­
pose '. Nevertheless, sou nd practices such as daily ch ecking
of ward drug stocks, ce ntral preparation of drugs adminis­
tered by intravenous infusi on, exclusion of ampoules
of potassium chloride from general wards, the use of two

----~ -~ .
- -

witnesses for all injections and the use of swipe cards
to limit access to critical areas, not only reduces the
omnipresent risk of accidents, but also makes the very rare
occurrence of malfiscence more difficu lt. James and
Leadbeatter 306 suggest that hospital policies could include:
• Detailed recording of incidents of unexpected collapse
to include those present at or around the time of
collapse.
• Taking and storing of a blood sample following such a
collapse for possible future analysis.
• Regular mortality review including analysis of place of
death and persons in atten dance.
• Detailed monitoring of potentially harmful drugs on
wards (drugs such as insulin, potassium chloride and
other arrhythmogenic drugs should be monitored as
closely as controlled drugs).
Pathologists may contribute to early detection by being
scrupulous in assigning causes of death. It was notable in
the Allitt case that 'status asthmaticus' and 'status epilepti­
cus' were used in circumstances in which they did not
really fit the facts. In another case sudden infant death
syndrome (SmS) was used inappropriately.
INVESTIGATION OF SUDDEN UNEXPECTED
DEATH OF CHILDREN IN HOSPITAL
When an infant or child dies suddenly and unexpectedly in
hospital, parents will be understandably distressed and may
be angry. Hospital staff will also be upset and sometimes
defensive. In the UK all such deaths should be reported to the
coroner or procurator fiscal. Although it is often easier for
pathologists to suggest that the post-mortem examination
should be carried out by someone from outside the institu­
tion, an in-house pathologist is more likely to obtain full
clinical cooperation and wi ll be better placed to provide
feedback to the parents and the hospital. When there is real
concern about serious error or negligence then it is wise to
seek an independent pathologist.
The investigation follows the same pattern as in adults,
but with additional considerations due to young age. In par­
ticular, small children are more likely to suffer accidental
over-dosage of drugs, and unexpected deaths may be due to
unrecognized congenital or inherited conditions, such as
inborn errors of metabolism, myopathy and cardiac conduc­
tion defects exacerbated by therapy for another condition.
The vulnerability of small children occasionally makes it nec­
essary to think the unthinkable and consider deliberate harm.
The scope of the examination varies depending on
whether a natural cause for death is very likely, an accident
is suspected or there is suspicion of foul play. Deaths clearly
due to natural causes can be deal t with in the usual way, but
suspicion of accident or other unnatural death should trigger
a graded response adopting some or all of the measures
in Tab le 19.4, depending on the degree of suspicion. This
Investigation of sudden unexpected death in hospital I 375
judgement requires a very detailed clinical history and it is
important to speak to several members of staff before
accepting the history at face value. The parents' account of
events may con tain important details unknown, overlooked
or even suppressed by staff. Start and Cross 307 provided
general guidance for pathological investigation of deaths
following surgery, anaesthesia and medical procedures.
Pathologists seldom have the opportunity to carry out a
scene investigation with the body still in situ and so the
best chance of establishing what went wrong may be lost
long before he or she is involved if the hospital does not
have procedures in place for securing specimens and
equipment after any sudden and unexpected death. The
natural urge to clear up the scene and complete 'last
offices' should be resisted in favour of a careful appraisal
of the scene by a senior member of staff. Where indica ted,
used syringes, ampoules, infusion solutions and blood bags
should be preserved for possible labora tory analysis. Waste
bins and sharps containers may contain useful samples, but
should be searched with great care to avo id injury. Infusion
pumps, temporary pacemakers, monitors, ventilators,
anaest hetic machines and other equipment should be set
as ide for testing, but e lectroni c equipment should not be
switched off or unplugged until any settings or memory
have been checked by an expert . All tubes and catheters
should be left in the body, the reasons for so doing care­
fully explained to the parents. Staff should be asked to
write down their recollection of events, who was present,
t he position of the body as it was found and other details.
Th e post-mortem examination itself should only take
place when there is a full history and, if the cause of death
is truly unknown must be a hybrid of the best of paediatric
and forensic practice (Table 19.5). It is essential that the
pat holog ist is fully inform ed about the child's illness and
any procedures undergone. In the past, coro ner's autopsies
have been carried out by pathologists who seemed unaware
of the complexities surrounding sudden death in the par­
ticular disease concerned. 56 A clinician who looked after
the child in life should be present to explain treatment,
procedures and any special con ce rns. In surgical deaths the
anaesthetist should be consulted and the surgeon should
attend in order to explain the procedure and ass ist (but not
carry out) the examination of the surgica l site.
Radiographs obtained soon afte r death are necessary to
document a ny gas in vessels, and are useful as an objective
re cord of t ubes and catheters. The value of routine photo­
gra phs to supplement a full externa l examination and
description may only become apparent in retrospect. How­
ever, photographs are not a substitute for a cohtempo rane­
ous written and drawn record on a pre-printed body chart
(those used in clinical departments for recording injuries to
children are suitable). All punctures and injection sites
should be recorded with an estimate of their age. Standard
measurements ane! body weight should be measured
accurately. When compared with the chi ld 's we ight on
admission to hospital, the post-mortem body weight is
376 I Sudden death in hospital

Table 19.5 A checklist for post-mortem examination of a child who has died suddenly and unexpectedly in hospital

Before the event

Check that the institution has a suitable protocol to be followed by staff in the event of sudden unexpected death approved
by the coroner/procurator fiscal.

Before the post-mortem examination

Ensure that relevant equipment and ante-m ortem samples have been secured
Obtain a full clinical history
Decide whether the exam ination wou ld be better carried out by or with another pathologist (forensic, sub - specia list, independent,
double doctor. etc.)
Refresh own knowledge about clinical aspects of the child's disease, its patho logy, complications and treatment
Discuss the case with relevant clinicians (paediatrician, surgeon, anaestheti st, etc.)
Obtain parents' account if present before or at the time of death
Discuss relevant samples with , for example, the clinical biochemist or toxicologist in ad vance
Consider the need to take immediate samples
Obtain radiographs within 12 hours of death if gas embolism is a possibility

During the post-mortem examination

Document all tubes, lines and catheters (whole body radiograph s, photograph s, drawings, text)
Document all puncture sites and incisions
Document any injuries
Record the body weight and standard measurements
Full post-mortem with particular care to exclude and record:
Pneumothora x

Gas embolism

Thromboembolism

Cerebral oedema and herniation

Exa mine any operation site with the surgeon present

Save any implanted devices
Save appropriate samples:
Histo logy of all major organs, any lesion and operative site

Vitreous humour for electrolytes, and glucose if hyperglycaemia relevant

Culture blood, spleen, lung, cerebrospinal fluid, any infective lesion, thrombus on catheter tips, etc. for bacteria

Consider virolog ica l cu ltures

Frozen tissue samples, e.g. skeletal muscle, kidney, myocardium, liver [possible enzyme, muscle or toxicological studies)

Fibroblast culture or other source of DNA

Frozen lung if fat embolism possible

Serum, urine, stomach contents, bile, injection sites, etc., for possible toxicology as advised by local toxicolog ist

Serum for mast cell tryptase if anaphylaxis possible

Consider saving heart and/or brain for specialist examination

After the post-mortem examination

Discuss findings with coroner/procu rator fi scal and clinicians
Give cause of death when possible
Inform relatives of tissue, organs and samples retained, and any progress towards finding the cause of death, perhaps via
the legal authority
Attend or instigate an institutional death review meeting (Have there been simi lar deaths, and what is the mortality rate of that
particular clinical unit7)
Ensure that any adverse event has been appropriately documented and reported
If the cause of death remains unknown, then seek an independent detailed clinical review or take other appropriate action to exclude
adverse event, error, negligence or malfiscence

sometimes a useful clue to excess or inadequate adminis­ examin ation will include all body cavities (any excess fluid
tration of intravenous fluids. should be measured an d saved for analysis and comparison
The body should be opened with care, noting the posi­ with infusion fluids) an d all major organs. Special atten­
tion and condition of the tips of all tubes and catheters. The tion should be taken during dissect ion to exclude gas

• -

embolism, pneumothorax and cerebral oedema. Trauma,
such as chest wall contusion, retroperitoneal haematoma,
stomach perforation and even rib fractures may sometimes
be attributable to vigorous resuscitation. 308 ,309 Subdural
haemorrhage may be enco untered at post-mortem exami­
nation if there has been unsu ccessful open heart massage
during card iac surgery.
In addition to standard histological sampling, it is nec­
essalY to sample any lesion and any operative site. Snap­
frozen samples of urine, serum, li ver, myocardium, skeletal
muscle an d kidney will prove valu ab le if a n inherited
metabolic disorder or myopathy becomes a consid eration
late r, bu t when metabolic disease is suspec ted it is adv is­
ab le to con sul t a clinical biochemist over sa mplin g in
advance and to save a source of DNA.
Bacteriological samples should include blood and cere­
brospinal fluid as a minimum. Vitreous humour biochem­
istry may point to an unexpected electrolyte disturbance.
Samples must be saved for toxicology in case a drug error is
subsequently suspected. Sub-optimal specimens hampered
one study of anaesthetic-related deaths an d so a standard set
of samples that will include blood, urine and gastric contents
should be agreed in advance with the lo cal toxicologist.
Samples taken in life an d already sto red in the labo ra­
tory may prove invaluabl e for metabolic, toxicological and
serological examination . In problematic cases the patholo­
gist must take steps to secure these from routine disp osal.
If the death remains unexplained after the gross post­
mortem examination and was consistent with sudden car­
diac arrest then the heart should be saved with authorization
or, at the very least, extensively sampled, with the conduct­
ing system saved for possible serial sectioning and samples
saved for 'molecular autopsy'.99 Any implanted pacemaker
should be submitted for testing with its leads, although in a
series of 38 surgically induced cases of complete heart block
in children treated by permanent pacemaker implantation
there were no hospital deaths.223 If a cerebral cause of sud­
den death is possible from the history then the fresh brain
should be examined closely for fla tten ing of gyri, herniation
and coning and authorization sought for fixation and formal
neuropathological examination.
Ifthe cause of death is still undetermin ed when all inves­
tigations are complete then it is often appropriate to seek an
independent clinicopathological review within t he institu­
tion in the first instance. It is not acceptable or safe to
ascribe death to the patient's original illness if it does not fit
the facts. The question may arise whether SID S can occur in
a hospitalized patient. Peterson et a1310 calculated that if the
risk of SIDS is the same for hospitalized in fants as fo r those
in the community then SIDS would be rare even in a busy
children 's hospital. However, they suggested that the rate
might be higher in sick infants, citing cases that we might
find hard to accept as SIDS today. 'Undetermined' is not
usually a popular conclusion, but the truth is preferable to
obfuscation a nd fabrication. Inappropriate use of the term
SIDS may lead later to suspicion of a 'cover up'.
References I 377
In England an d Wales a co roner's pathologist is often
asked to determine w hether the death was natural or
unnatural. There is unfort una tely no definition of 'na tural
causes'. Deaths after surgery or medical treatment often fall
into the g rey area between natural and unnatural causes of
death. 311 Coroners themselves do not agree on how to use
these terms, and so it is as well to discuss borderline cases
individually. A majority of coro ners will accept natural
causes for deaths following appropriate med ica l interve n­
tion for natural disease from which the patient was likely
to die, provided tha t there was no element of negligence
and the relatives do not object.312 Great care should be
taken in assigning the cause of death and in any accompa­
nying comment not to go beyond the pathologist's expert­
ise. A si gnificant propOJ1ion of medical negligence claims
are attributable to inaccurate post-mortem diagnoses and
speculative comments in post-mortem reports.
ACKNOWLEDGEMENT
[ would like to acknowledge the assistance of Professor
ARW Forrest with the section on carer-associated serial
killings.
REFERENCES
Go ld stein B, Metlay L, Cox C, Rubenstein JS. Association of
pre morte m di agno sis and autopsy findin gs in pediatric
intensive care unit versus emergency department versus ward
patients. Crit Care iVIed 1996 : 24:683-6.
2 Stambouly JJ, Kahn E, Boxer RA. Correlation be tween clinical
diagnoses and autopsy fin dings in critica ll y iLl children.
Pediatrics 199 3; 92 :248-51.
3 Koszyca B, Moore L, Toogoo d I, Byard RW Is postmortem
examinat ion usefu l in pediatric oncology? Pediatr Pat/wi
1993; 13 :709- J 5.
4 Gatzoulis MA, Sheppard MN, Ho SY. Value and impact of
necropsy in paediatric card iology. Heart 1996; 75:626-31.
5 Russell GA, Berry P J. Postmo rtem audit in a paediatric
card iol ogy unit. J Clin Path 01 1989; 42:9 J 2-18.
6 Buchino JJ, Corey TS, Montgo mery V. Sudden unexpected
death in hospitalized children. J Pediatr 2002; 140:46 1-5.
7 Reis AG, Nadkarni V, Perondi MB et a!. A prospect ive
investigation into the epidemiology of in-hospital pediatric
ca rdiopulmonary resuscitation usin g the internation al Utstein
reporting style. Pediatrics 2002 ; 109:200-9.
8 Manton N, Sm ith NM, Byard RW Unexpected childhood death
due to hem olyti C uremic syndrom e. Am J Forensic iVIed Pathol
2000; 21 :90-2.
9 Champ CS, Byard RW Sudden death in asthma in,childhoocl.
Forensic Sci Int 1994; 66: 117-27.
10 Breningstall GN. Mortality in pediatric epilepsy. Pediatr
Neurol 2001; 25 :9-16.
II Dahlquist G, Kall en B. Mo rta lity in childho od -onset type I
diabetes: a population-based study. Diobetes Care 2005;
28:2384 - 87.
12 Ed ge JA, Ford-Adams ME, Dunger DB. Cau ses of death in
chi ldren with insulin-dependent diabetes 1990-96. Arch Dis
Child 19 99; 81:318-23.
378 I Sudden death in hospital
13 Thordarso n H, Sovik O. Dead in bed sy ndrome in
youn g diabeti c patients in Norw ay. Diab et Med 1995;
J 2 :782-7.
14 Grey TC, Mittleman RE, Wetli CV, Horowitz S.
Ao rtoes ophageal fistul a and sudden dea th . A rep olt of two
cases and review of th e literature. Am J Fore nsic Med Fath ol
1988 ; 9:19-22.
15 Cohle SO, Colby T. Fatal hemoptysis from Behcet's disease in
a child . Cardi ollClSC Pathol 2002 ; 11 :29 6-9.
16 Lash ley PM , Ramesh J , Ferdinand S. Spontaneous rupture of
th e spleen. West Indian Med J 1998; 47:172-3.
17 Vidyashankar C, Basu A. Kulk arn i AR, Chou dhury RK.
Spontaneous rupture of sp leen in falc iparum malaria. India'i J
Gastroentero/2003; 22:10 1- 2.
18 Koszyca B, Moore l, Byard RW. lethal manifestations of
neurofibrom atos is type I in ch ildhood . Pediatr PatllOl 199 3;
13:573-81.
19 Ashkenazi S, leibovici l, Samra Z et al. Risk factors for
morta lity due to bac te remia and fungemia in child hood . Clin
I'~fect Dis 1992; 14:949 -51.
20 lascari AD , Pearce JM, Swanso n H. Sudden death due to
disseminated cly ptococcos is in a child w ith leukem ia in
remission. South Med J 1997; 90: 1253-4.
21 Brown AP, Khan K, Sinclair S. Bacterial toxico sis/toxic sho ck
synd rome as a contributor to morbidity in children with burn
injuries. Bums 2003; 29:733-8.
22 Doctor A, Harper MB, Fleisher GR. Group A beta- hemolytic
st reptococcal bacteremia: hi sto rical overview, changing
incidence, and recent association with varicell a. Pediatrics
1995; 96:4 28 -33.
23 Hidalgo-Carballal A, Suarez - Mier MP. Sudden unexpected
death in a chi ld with varicell a caused by necrotizing fasc iitis
and strepto co ccal toxic sho ck syndrome. Am J Forensic Med
PatllOl 2006; 27 :93-6.
24 Tseng HW, Liu Cc, Wang SM et al. Complications of
varicella in children: emphasis on skin and cent ral nervous
system disorders. J Microbiol Immunol Infect 2000;
33:248-52.
25 Yang YJ, Liu CC, Wang SM et al. Streptococcal toxic shoc k
synd rom e complicating varicella in children. J Formos Med
A ssoc 1997; 96:749-53.
26 Buc k JR, Co nnors RH, Coon WW et al. Pulmonary embolism
in children. J Pediotr Surg 1981; 16:385 -91.
27 Byard RW, Cutz E. Sudden a nd unexpected death in infancy
and childhood due to pulm ona lY thromboembolism. An
autopsy study. Arch Pathol Lab Med 1990; 114:142-4.
28 de la Grandmaison Gl, Durigon M. Pulmonary embolism: a
rare cause of sudden infant death. Am J Forensic ivIed Pathol
2002; 23 :257 -9.
29 Baudet EM, Puel V, McBride JT et al. long- term resu lts of
valve replacement with the St. Jude Medical pro sth esis.
J Thorac Cardiovasc Surg 1995; 10 9:8 58-70.
30 Citak A, Emre S, Sairin A et al. Hemostatic problems a nd
thromboembolic com pli cat ions in nephrotic ch ildren. Pediatl'
Nephrol2000; 14:138-42.
31 Jones CL, Hebert D. Pulmo nalY thrombo-embolism in the
nephrotic syndrome. Pediatr Nephro! 199); 5:56-8.
32 Mehls 0, Andrassy K, Koderisch J et al. Hemostasis an d
thromboembolism in children with nephrotic syndrome:
differences from adults. J Pediatr 1987; 110:862-7.
33 Zimmerman Rl, Novek S, Chen JT, Roggli Y. Pulmona ry
thrombosis in a lO-yea r-old child w ith minimal change
disease and nephrotic syndrome. A clinical, radiologic, and
pathologic correlation with literature review. Am J Ciin Patho!
1994; 101:230- 6.
34 Baskerville PA. Thromboembolic disease and congenital
venous abnormalities. Phlebologie 1987; 40:531 - 6.
35 Bohm N, Bre ndlein F, Niederhoff H, Sutterle H. Diffuse
hemangiomatosis of the sm all pelvis. Death from lung
embo lism in an ll-yea r-old girl. Med Welt 1978; 29:839-43.
36 Machin GA, Kent S. PulmonalY thromboembolism from a
large hemangioma in a 4-week-old infant. Pedjatr Pathol
198 9; 9:7 3-8.
37 Slavo tinek MI, Vacha SJ , Peters KF, Biesecker LG. Sudden
death caused by pulmonary thromboembolism in Proteus
syn drome. Clin Genet 2000; 58:386-9.
38 Nuss R, Hays T, Manco-Johnson M. Childhood thrombosis.
Pediatrics 1995; 96:291-4.
39 Sta hl J, Santos lD, Byard RW. Coron ary artery
thromboembo lism a nd un expec ted death in childhood and
adolescence. J Forensic Sci 1995; 40:599-601.
40 Muraskas J, Besinge r R, Bell T et al. Perinatal myocardial
infarction in a newborn with a structurally normal heart.
Am J Perinatal 1997; 14:93- 7.
41 Tillett A, Hartley B, Simpson J. Paradoxical embolism causing
fatal myocardial infarction in a newborn in fant. Arch Dis
Child Fetal Neol1atal Ed 2001; 85:F 137 -8.
42 Zakowski MF, Edwards RH, McDonough ET. Wilms' tumor
presenting as sudden death due to tumor embolism. Arch
Pathol Lab Med 1990; 114:605-8.
43 Moore L. Byard RW. Fatal paradoxical emb ol ism to the left
carotid artery during partial resection of Wilms' tumor.
Pediatr Parhol 1992; 12:551 -6.
44 Shurin SB, Gaud erer MW, Dahms BB, Conrad WG. Fatal
intraoperative pulmonary embolization of Wilms tumor.
J Pediatr 198 2; 101 :559-62.
45 Arthur OS, Stephe ns CA, Brummitt WM et al. Fatal tumor
embolism during examination under anesthesia. Surgery
1973 ; 74:466-8.
46 Chan KW, Fry er CJ, Fra ser GC, Dimmick JE. Sudden cerebral
death in malignant presacral teratom a. Med Pediatr Oncol
1985; 13:395-7.
47 Nich ols GR, Corey TS, Davis GJ. Nonfracture-associated fatal
fat embol ism in a case of child abuse. J Forel1sic Sci 1990;
35:493-9.
48 Shapiro G, Green OW, Fatica NS, Boachie-Adjei O. Medical
compli ca tions in scoliosis surgery. Curl' Opin Pedi atr 2001;
13 :36-41.
49 Shulm an ST, Grossman BJ. Fat embolism in childhood. Review
with report of a fatal case related to physica l therapy in a ch ild
with dermatomyositis. Am J Dis Child 1970 ; 120:480-4.
50 Edw ards KJ, Cummings RJ. Fat embolism as a comp li cation of
clo sed femoral shortenin g. J Pediatr Orthop 1993; 12:542-3.
51 Ganel A, Israeli A, Horoszowski H. Fatal compli cation of
femoral elongation in an achondroplastic dwarf. A case
repo rt. Clin Orthop 1984 ; 185:69-71.
52 Pender ES, Pollack CV Jr, Evans OB. Fat embolism syndrome in
a chi ld with muscular dystrophy. J Emerg JvIed 1992; 10:705-11.
53 Horton DP, Ferriero DM, Mentze r We. Nontraumatic fat
embolism sy ndrom e in s ickl e cell anem ia. Pediatr Neurol
1995; 12:77-80.
54 Vi chinsky E, William s R, Das M et al. Pulmonary fat
embolism: a distinct cause of severe acute ches t syndrome in
sick le cell anemia. Blood 1994; 83:3107-12.
55 Vichinsky EP, Neum ay r lD, Ea rles AN et al. Causes and
outcomes of the acute chest syndrome in sickle cell disease.
National Acute Chest Syndrome Study Group. N Eng! J Med
2000; 342:1855-6 5.
56 Gray A, An ionw u EN, Davies S(, Brozovic M. Patterns of
mortality in sickle cell disease in the United Kingdom. J Clin
Pathol 19 91; 44 :459-63.
57 Haber lM, Hawkins EP, Seilheimer OK, Saleem A. Fat
overload syndrome. An autopsy study with eva luation of the
coagulop athy. Am J Clin Patho! 1988; 90:22 3-7.

58 Landry BA, Melhem RE. Pulmonary nodules seco ndary to
total parenteral alimentation. Pediatr Radial 1989 ; 19:456-7.
59 Leve ne MI, Batisti 0, Wigglesworth JS et al. A prospective
study of intrapul mon ary fat accumulation in the newborn
lun g following intra lipid infusion. Acta Paediatr Scand 1984;
73 :454-60.
60 Puntis JW, Rushton DJ. Pulmonary intravasc ular lipid in
neonatal necropsy specim ens. Arch Dis Child 199 1; 66: 26-8.
61 Schroder H, Paust H, Schmidt R. Pulmonary fat embo lism
after intralipid therapy: a post-mortem artefact? Light and
electron microscopic invest igations in low-birth-weight
infants. Acta Paediatr Scand 1984; 73:461- 4.
62 Shulman RJ , Langston C, Schanler RJ. Pulmonary vascu lar
lipid deposition after ad ministra tion of intravenous fat to
infa nts. Pediatrics 1987 ; 79:99-102.
63 Schulz PE, Weiner SP, Haber LIvI et al. Neurological
complicatio ns from fat emulsion therapy. Ann Neural 1994;
35:628-30.
64 Vald es -D ape na M. Iatrogeni c disease in th e perinatal period as
see n by the pathologist. Monog r Pathol 1981 ; 22:382-418.
65 Lau KY, Lam PK. Systemic air embolism: a comp lication of
ventilator therapy in hyalin e membrane disease. Ciin Radial
1991; 43:16-18.
66 Fab erowski LW, Black S, Mi ckl e JP. Inciden ce of venous air
embolism during craniectomy for craniosynostos is repair.
Anesthes iology 2000; 92:20-3.
67 Cina SJ , Do wns JC, Conradi SE. Hydrogen peroxide: a so urce
of lethal oxygen embolism . Case report and review of the
literature. Am) Forensic Med Pathol 1994; 15:44-50.
68 Lubec B, Hayn M, Denk W, Bauer G. Brain lipid peroxidation
and hydroxy rad ical attack fo ll owi ng the intravenous infusion
of hydrogen peroxide in an in fa nt. Free Radical Bioi Med
1996 ; 21 :2 19- 23 .
69 Buchino JJ, Kee nan WJ, Pi etsch JB et at. Malpos itioning of
the endotrac heal tube in infants with tracheo esophageal
fistula. ) Pediatr 1986; 109:5 24-5.
70 Shiotani S, Kohno M, Ohashi N et al. Cardiovasc ular gas on
non-traumatic postmortem co mputed tomography (PMCT):
the influence of card iopulm onary resuscitation. Radiat Med
2005; 23 :22 5-9.
71 Rudd PT, Wiggl esw orth JS. Oxygen embo lus during
mechanical vent ilation with di sappearance of signs afte r
death. Arch Dis Child 1982; 57:237-9.
72 Denton S, Mileusnic D. Delayed sudden death in an infant
followin g an accide ntal fall: a case report with rev iew of the
literature. Am) Forensic Med Pa thol 2003 ; 24:371-6.
73 Cox C. When ca n nurses be charged with manslaughter? Nurs
Times 2001; 97 :21.
74 Hahn YS, McLone DG. Risk fact ors in the outcome of children
with minor head injury. Pedia tr Neurasurg 1993 ; 19: I 35-42.
75 Ch un CH, Johnso n JD , Hofste tter M, Ra ff MJ. Brain abscess.
A study of 45 co nsecuti ve cases. Medici ll e (Baltimore) 1986;
65:4 15-31.
76 Shemie S, Jay V, Rutka J, Armstrong D. Acute obstru ctiv e
hydrocephalus and sudden death in children. Ann Em erg Med
1997; 29:524- 8.
77 Edge JA. Cereb ral oedema durin g treatment of diab eti c
ketoacidosis: a re we any nearer finding a ca use? Diabetes
Metab Res Re/J 2000 ; 16:3 16- 24.
78 Glase r N, Barnett P, McCaslin I et al. Risk factors for cerebra l
edema in children with diabetic ketoacidos is. The Pedi atric
Emergency Medicine Co llaborative Researc h Committee of th e
American Academy of Pediatrics. N El1gl ) Med 2001;
344:264-9.
79 Marcin JP, Gl ase r N, Barnett P et al. Facto rs associated with
adv erse outcomes in children with dia betic ketoacidosis­
related cerebral edema. ) Pediatr 2002; 141 :793-7.
References I 379
80 Glase r N. New perspec tives on the pathogenes is of cerebra l
edema co mplicati ng diabetic ketoacidosis in children. Pediatr
E17do crinal Re/J 2006 ; 3:3 79-86.
81 Rose nbloom Al. Intracerebral crises during treatment of
di abetic ketoacid osis. Diabetes Care 1990; 13:22-33.
82 Alper G, Jarjour IT, Reyes JD et al. Outco me of children with
cerebral edema ca used by fulminant hepat ic failure. Per/ia tr
Ne ll 1'01 1998; 18 :299 -304.
83 Ee LC, Shepherd RW, Cleghorn GJ et al. Acute liver failure in
chi ldre n: a regional experience.) Pa ediatr Child Health
2003; 39:107-10.
84 Liu GT, Urion DK, Volp JJ. Cerebral ede ma in acu te h epatic
fai lure: clinicopathologic co rrelati on. Pediall' Neural 1993;
9:224- 6.
85 Riviel lo JJ Jr, Rezvan i I, DiGeorge AM, Foley CM. Cerebral
edema causing death in children with map le syrup urine
di sease. ) Pediatr 199 1; 119:42-5.
86 Garso n A Jr. Medicolegal problems in the management
of cardiac arrhythm ias in children. Pediatrics 1987;
79 :84-8.
87 Sm ith NM, Bourne AJ, Clapton WK, Byard RW. The spectrum
of prese ntation at autopsy of myocarditis in infancy and
childh ood. Pathology 1992; 24: 129- 31.
88 Critchley LA. Yet ano ther report of anesthetic death due to
unsus pe cted myo carditis.) Ciin Anesth 1997; 9:676-7.
89 Celermajer DS, Sholler GF, Howman-Giles R, Celermajer JM.
Myoca rdi al infarction in childhood: clinical ana lysis of 17
cases and medium term follow up of survivors. Br Heart)
1991; 65:332- 6.
90 John srude CL, Towbin JA, Cecc hin F, Perry Jc. Postinfarction
ventricular arrhythmias in children. Am Heart) 1995;
129: 1171 -7 .
91 Zebrak J, Skuza B, PogorzelskJ A et al. Partial CFIR
genotypin g an d characterisation of cystic fibro sis patients
with myo ca rdial fibrosis and necrosis. C/in Genet 2000;
57 :56-60.
92 Sheh ata 8M, Patterson K, Thomas JE et al. Histiocytoid
ca rdi omyopathy: three new cases and a review of the
literature. Pediatr De v Pat/JOI 1998; 1:56-69.
93 Stahl J, Couper RT, Byard RW. Oncocytic cardiomyopathy: a
rare cause of unexpected early childhood death associated
with fittin g. Med Sci Law 199 7; 37:84-7.
94 Bonnet D, Martin D, Pascale DL et a1. Arrhythmi as and
conduction defects as present in g sy mptoms of fatty acid
oxidation disord ers in children. Circulation 1999 ;
100:2 248-53.
95 Allan WC, Gosp e SM Jr. Seizures, sy nco pe, or breath-h olding
presenting to th e pediatric neurol og ist - when is the etio logy
a life-threatening arrhythmia? Semil1 Pediatr Neural 2005;
12:2-9.
96 Cooke RA. Chambers JB. Anorexia nervosa and the hea rt. Br
) Hasp Med 1995; 54:313 -17.
97 Swen ne I, Larsson PI. Heart risk associated with weight loss
in anorexia nervosa and eating diso rders: risk factors for
QTc interval prolongation and di spersion. Acta Paedia tr
1999; 88:304-9.
98 Va nn acc i A, Baronti R, Masini E et al. Anorex ia nervosa and
the ri sk of sudde n death. Am ) Med 2002; 112:3 27-8.
99 Teste r DJ, Ackerm an MJ. The role of molecular a~topsy in
unexplained sudden ca rdiac death. Curr Opin Cardiol 2006;
2 J: 166-72.
100 Arieff AI, Ayus JC, Fraser CL. Hyponatraemia and death or
permanent brain dama ge in healthy children. BM) 1992;
304: 1218- 22.
101 al Sab ri AM, Smith N, Busu ttil A. Sudd en death due to auto ­
immune Addison's disease in a 12-yea r-old girl. Int) Legal
Med 1997; I JO :278- 80.
 380 I Sudden death in hospital
102 Berlatzky Y, Shiloni E, AnneI' H, Weiss Y. Delayed rupture of
the spleen or de layed diagnosis of the splenic injury? lsI')
lVled Sci 1980; 16:659-64.
103 Olsen WR, Polley TZ Jr. A second look at delayed splenic
rupture. Arch Surg 1977 ; 112:42 2- 5.
104 Sta mbouly JJ, McLaughlin LL, Mandel FS, Boxer RA.
Co mplications of ca re in a ped iatric intensive ca re unit: a
prospective stud y. Intensive Ca re Nled 1996; 22: 1098- 104.
10 5 Yo ung RS , Tru ax BT. Hypernatremic hemo rrh agic
enceph alopathy. Ann Neurol 1979; 5:588-91.
106 Rouse DA, Hargrove R. An unusual case of gas embo lism.
Am) Forensic lVled Parhol 1992; 13 :268-70.
107 Bagwell CE, Salzberg AM, So nnin o RE, Hay nes JH.
Potentially letha l comp lications of central ve nous catheter
placement.) Pediar Surg 2000; 35:709- 13.
108 Na rsingha ni U, Chadha M, Farrar HC, Ana nd KS. Life­
threatening respiratory fa ilure followin g accidental infusion
of polyethylene glycol elec tro lyte so luti on into the lu ng.
) Toxicol Ciin Toxicol 2001; 39 : 105-7.
109 We therton AR, Corey TS, Buchino JJ , Burrows AM. Fata l
intravenous injection of potassiu m in hosp italized patients.
Am) Forensic lVled Patho12003; 24: 120 - 31.
110 Cohen MR. Potassium chloride injection mix- up. Am ) Hasp
Pharm 1990; 47:2457-8.
III AJ Ferayan A, Russell NA, AJ Wohaibi ]\II et al. Cerebrospinal
t1uid lavage in the treatment of in advertent intrathecal
vincristine injection. Childs Nerv Sysr 1999; 15 :87 - 9.
112 Alcaraz A, Rey C, Concha A, ]\Iledina A. Intrathecal
vincristine: fatal myeloencephalopathy despite cerebrospi nal
t1uid perfusion.) Toxicol Ciin Toxicol 2002; 40:557- 61.
113 Dettmeyer R, Dri eve r F, Becker A et al. Fatal
myeloencephalop athy due to acc id ental intratheca l
vincristine administration: a report of two cases. Forensic
Sci Int 2001; 122 :60-4.
114 Fernandez CV, Esau R, Ham il ton D et al. Intrathecal
vincristin e: an analysis of reaso ns for recurrent fata l
chemotherapeutic error with reco mm end ations for
prevention. ) Pediatr Hemarol Oncol 1998; 20 :587 - 90.
115 Kwack EK, Kim DJ, Park TI et al. Neural toxicity induced by
accidental intrathecal vincristine administration . ) Korea n
Med Sci 1999; 14 :688-9 2.
116 Seale JR. Not again! Erroneous intratheca l. inj ect ion res ul ts
from a problem with protocols. BM) 200 1; 322:548.
117 Walter AW. Errors involving pediatric patients rece iving
chemotherapy: a literature review.1Vfed Pediatr Oncol 1997;
29: 155-6.
118 Dyer C. Doctors face trial for manslaughter as criminal
charges against doctors continue to rise. BM) 2002; 325:63.
119 Ferner RE. Med ication errors that have led to manslaughter
charges. BlVl)2000; 32 1:1212 - 16.
120 Kausha l R, Bates DW, Landrigan C et al. Medication errors
and adverse drug events in pediatric in patients. )AMA 2001;
205:2 11 4- 20
121 Ko ren G, Barzil ay Z, Gree nwald M. Tenfo ld errors in
adm inistratio n of drug doses : a neg lec ted iatroge nic disease
in pediatri cs. Pediatrics 1986; 77:848- 9.
122 Koren G, Haslam RH . Ped iatri c med ication errors: predic tin g
and preventing tenfo ld disas ters. ) Ciin Pharmacal 1994;
34: 1043-5.
123 Conroy S. Unlicensed and off- label dllJg use: issues and
recommendations. Paediatl' Drugs 2002; 4:353-9.
124 Kozel' E, Scolni k D, Macpherson A et al. Variables assoc iated
with medication errors in pediatric emergency medicine.
Pediatrics 2002; 11 0:737 - 42.
125 Kozer E, Scolnik D, Keays T et al. Large errors in the
dosing of medications for chi ldren. N Engl ) Med 2002;
346:1175-6.
-------- - - - -
- --- .
126 Sullivan J E, Buchino JJ. Med ic ation errors in pediatrics: the
octopus evading defeat.) Surg 011co12004; 88:182 - 8.
127 Johnson TN. The development of drug metabolising enzymes
and their influence on the susceptib ility to adverse drug
reactions in children. Toxicology 2003; 192:37- 48.
128 Swanson JR, Jones GR, Krasselt \IV et al. Death of two
subjects due to imipramine and deSipramine metabolite
accumulation during chronic therapy: a review of the
li terature a nd possib le mechanisms.) Forellsic Sci 1997;
42 :335 - 9.
129 Varley CK. Sudden death of a child treated with imipramine.
Case study. ) Child Adolesc PsychopiJarmacol2000; 10 :32 1-5.
130 Gup ta M, Thaler HT, Steinherz L. Presence of prolonged
dispers ion of QT in terva ls in late surv ivo rs of childhood
anthracycl ine the rapy. Pediatr Hematol Ollcol 2002;
19 :533 -42.
13 1 Krischer JP, Epste in S, Cuthbertso n DD et al. Clinica l
card iotoxicity fo ll ow in g anthracycline treatment for
chil dhood cancer: the Pediatric Oncology Group experience.
) Clill Ol1col 1997 ; 15: 1544-52.
132 Ma ssin MM, Dresse MF, Sc hmi tz V et al. Ac ute
arrhyth mogen icity of fir st-dose chemo therap euti c age nts in
chil dren. Med Pediatr 011co/2002; 39:93-8.
133 Wortman JE, Lucas VS Jr, Schuster E et al. Sudden death
dUJi ng doxorub icin administration. Cal1cer 1979; 44:1588-9 1.
134 Arseneau ]C , Bagley CM, Anderson T, Canellos GP.
Hyperkalaemia, a sequel to chemothera py of Burkitt 's
lymphoma . Lancet 1973; 1: 10-14.
13 5 Chasty RC. Liu-Yin JA. Acute tumour lysis syndrome. Br)
Hasp Med 199 3; 49 :488-92.
136 Anon. Severe cardiac arrhythmia on cisapride. Prescrire JJ1f
2000; 9: 144-5.
137 Levy J, Hayes C, Kern J et al. Does cisapride influence
card ia c rhythm? Results of a United States multicenter,
double-blind, placebo-controlled pediatric study. ) Pediatr
Gastroellterol NutI' 2001; 32:458-63.
138 Ramirez-Mayans J, Garrido-Garcia LM, Huerta-Tecanhuey A
et al. Cisapride and QTc interval in children. Pedia trics 2000;
106: 1028-30.
139 Unlu RE, Alagoz IVIS, Uysal AC et al. Phenol intoxication in
a child.) CmJ1ioJac Surg 2004; 15:1010-13.
140 Cote CJ, Karl HW, Notterman DA et al. Adverse sedation
events in pediatrics: analysis of medications used for
sedation. Pediatrics 2000; 106 :633-44.
141 Cannon ML, Glazier SS, Bauman LA. Metabo li c ac idosis,
rhabdomyolysis, and cardiovascular collapse after pro longed
propofol infusion. ) NeuroSL/rg 2001; 95: 1053 - 6.
142 Parke TJ, Stevens JE, Rice AS et al. Metabolic acidosis and
fatal myocardial failure after propofol infusion in children:
five case reports. BM) 199 2; 305:613 - 16.
143 Vasile B, Rasulo F, Candiani A, Latronico N. The
pathophysiology of propofol infusion syndrome : a simple
name for a complex syndrome. Intensive Care Med 2003;
29: 1417-25.
144 Wo lf A, We ir P, Segar P et al. Impaired fatty acid oxidation
in propofo l infu sion sy ndrome. Lal1cet 200 1; 357:606-7.
145 Wo lf AR, Potter F. Propofo l infusio n in ch il dren: when do es
an anesthetic too l become an intensive care liability?
Paediatr Al1aestil 2004; 14 :435- 0.
146 Pumphrey RS. Lesso ns for manageme nt of a nap hylax is from
a study of fata l reac tions. Clin Erp Allergy 2000;
30: 11 44-50.
14 7 Lee JM. Greenes DS. Bip hasic anaphylactiC reactions in
pediatrics. Pediarrics 2000; 106:762- 6.
148 Novembre E, Cianferoni A, Bernard ini R et al. Anap hy laxis
in ch il dren: clinical and all ergologic features. Pediatrics
1998 ; 101 :EO.
"""'""­

149 Novembre E, Calogero C, Mori F et al. Biphasic anaphylactic
reaction to ketorolac tromethamine. Int J Immul1opathol
Pharmacal 2006; 19:449-50.
150 Sakaguchi M, Inouye S. Anaphylaxis to gelatin-containing
rectal suppositories. J Allergy Ciin Immunol 2001;
108: 1033-4.
151 Bohlke K, Davis RL, Marcy SM et al. Risk of anaphylaxis
after vaccination of children and adolescents. Pediatrics
2003; 11 2:815-20.
152 Ponvert C, Ardelean-Jaby D, Colin-Gorski AM et al.
Anaphylaxis to the 23-valent pneumococcal vaccine in child: a
case-control study based on immediate responses in skin tests
and specific IgE determination. Vaccine 2001; 19:4588-91.
153 Singh A, Biswal N, Nalini Pet al. Acute pulmonary edema as
a complication of anti-snake venom therapy. Il1dian J
Pediatr 2001; 68:81-2.
154 Zylberberg R, Schott RJ, Fort J et al. Sudden death following
white cell transfusion in a premature infant. J Perinatal
1987; 7:90-2.
155 Devenney I, Faith-Magnusson K. Skin prick tests may give
generalized allergic reactions in infants. Ann Allergy Asthma
Immunol 2000; 85:457-60.
156 Lockey RF, Benedict LM, Iurkeltaub PC, Bukantz Sc.
Fatalities from immunotherapy (IT) and skin testing (SI). J
Allergy Ciin Immunol 1987; 79:660-77.
157 Reid MJ, Lockey RF, Iurkeltaub PC, Platts-Mills IA. Survey
of fatalities from skin testing and immunotherapy
1985-1989. J Allergy Ciin Immullol1993; 92:6-15.
158 Burman D, Hodson AK, Wood CB, Bmeton NF. Acute
anaphylaxis, pulmonal)' oedema, and intravascular haemolysis
due to cl)'oprecipitate. Arch Dis Child 1973; 48:483-5.
159 Harper JL, Gill ]C, Hopp RJ et al. Induction of immune
tolerance in a 7-year-old hemophiliac with an anaphylactoid
inhibitor. Thromb Haemost 1995; 74: 1039-41.
160 Shemin ER. Anaphylactic reaction to gamma-globulin.
JAMA 1968; 203 :59.
161 Lasser EC, Lyon SG, Berl)' Cc. Reports on contrast media
reactions: analysis of data from repolts to the U.S. Food and
Drug Administration. Radiology 1997; 203 :605-10.
162 Shehadi WH. Death following intravascular administration
of contrast media. Acta Radio/ Diagn (Stockh) 1985;
26:457-61.
163 Pomeranz S, Gimmon Z, Ben Zvi A, Katz S. Parenteral
nutrition-induced anaphylaxis. J Parenteral Enteral NutI'
1987; 11:314-15.
164 Bimlingham PK, Suresh S. Latex allergy in children: diagnosis
and management. Il1dial1 J Pediatr 1999; 66:717-24.
165 Moneret-Vautrin DA, Laxenaire MC, Bavoux F. AJlergic
shock to latex and ethylene oxide during surgel)' for spinal
bifida. Anesthesiology 1990; 7J :556-8.
166 Bommer J, Ritz E. Ethylene oxide (EIO) as a major cause of
anaphylactoid reactions in dialysis (a review). Artif Organs
1987; 11:111-17.
167 Hurley RM. Anaphylaxis during hemodialysis. Int J Pediatr
Nephro11984; 5:53-4.
168 Dellinger CI, Miale ID. Comparison of anaphylactic
reactions to asparaginase derived from Escherichia coli and
from Erwil1ia cultures. Cancer 1976; 38:1843-6.
169 Zanotti KM, Markman M. Prevention and management of
antineoplastic-induced hypersensitivity reactions. Drug
Safety 2001; 24:767 -79.
170 Ganier M, Lieberman P. IgE mediated hypersensitivity to
pancreatic extract (PE) in parents of cystic fibrosis (CF)
children. Clil1 Allergy 1979; 9: 125-32.
171 Haworth CS, Niven RM, Moorcroft AJ et a!. Acute
anaphylaxis following midline catheterisation in a patient
with cystic fibrosis. Thorax 1999; 54:747.
References I 381
172 Pumphrey RS. Fatal anaphylaxis in the UK, 1992-2001.
Nouartis Found Symp 2004; 257:116-28.
173 Pumphrey RS, Roberts IS. Postmortem findings after fatal
anaphylactic reactions. J Clin Patho/ 2000; 53:273-6.
174 Caughey GH. Il)'ptase genetics and anaphylaxis. J Allergy
Clin Immunol2006; 117:1411-14.
175 Horn KD, Halsey JF, Zumwalt RE. Utilization of serum
tryptase and immunoglobulin E assay in the postmoltem
diagnosis of anaphylaxis. Am J Forensic Med Pathol 2004;
25:37-43.
176 Yunginger JW, Nelson DR, Squillace DL et al. Laboratol)'
investigation of deaths due to anaphylaxis. J Forensic Sci
1991; 36:857-65.
177 Co.llier PE, Blocker SH, Graff DM, Doyle P. Cardiac
tamponade from central venous catheters. Am J Surg 1998;
176:212-14.
178 Collier PE, Goodman GB. Cardiac tamponade caused by
central venous catheter perforation of the heart: a
preventable complication. J Am Call Surg 1995; 181 :459-63.
179 Suarez-Penaranda JM, Rico-Boquete R, Munoz JI et al.
Unexpected sudden death from coronal)' sinus thrombosis.
An unusual complication of central venous catheterization.
J Forensic Sci 2000; 45:920-2.
180 Pezzati M, Filippi L, Chiti G et a!. Central venous catheters
and cardiac tamponade in preterm infants. Intensive Care
Med 2004; 30:2253-6.
181 Beardsall K, White DK, Pinto EM, Kelsall AW. Pericardial
effusion and cardiac tamponade as complications of
neonatal long lines: are they really a problem? Arch Dis
Child Fetal Neonatal Ed 2003; 88:F292-5.
182 Byard RW, Bourne AJ, Moore L, Little KE. Sudden death in
early infancy due to delayed cardiac tamponade
complicating central venous line insertion and cardiac
catheterization. Arch Patitol Lab Med 1992; 116:654-6.
183 Leibovitz E, Ashkenazi A, Levin S, Nissim F. Fatal pericardial
tamponade complicating total parenteral nutrition via a
silastic central vein catheter. J Pediatr Gastroel1terol NutI'
1988; 7 :306-7.
184 Rogers BB, Berns SD, Maynard EC, Hansen IW. Pericarclial
tamponade secondal)' to central venous catheterization and
hyperalimentation in a vel)' low birthweight infant. Pediatr
Pathol 1990; 10:819-23.
185 Hermansen MC, Hermansen MG. Intravascular catheter
complications in the neonatal intensive care unit. Ciin
Perinatoi2005; 32:141-56.
186 Anon. Radiofrequency interference with medical devices. A
technical information statement. IEEE En9 iVIed Biol Mag
1998; 17:111-14.
187 Betts IR, Simpson lAo Inhibition of temporary pacing by a
mobile phone. Heart 2002; 87:130.
188 Robinson MP, Flintoft lD, Marvin AC. Interference to medical
equipment from mobile phones. J Med Eng Technol 1997;
21 :141-6.
189 Iat FH, Wah KC, Hung YH. A follow-up study of
electromagnetic interference of cellular phones on electronic
medical equipment in the emergency department. Emerg
Med (Fremantle) 2002; 14:315-19.
190 Iri JL, Severson RP, Firl AR et al. Cellular telephone
interference with medical equipment. Mayo Ciin Proc 2005;
80: 1286-90.
191 Boyle J. Wireless technologies and pati,ent safety in
hospitals. Telemed J E Health 2006; 12:373-82.
192 Ihong ML, Iay LK. Septicaemia from prolonged intravenous
infusions. Arch Dis Child 1975; 50:886-9.
193 Garrett DO, McDonald LC, Wanderley A et aJ. An outbreak of
neonatal deaths in Brazil associated with contaminated
intravenous fluids. J Infect Dis 2002; 186:81-6.

239 Chinachoti T, Surase ranivongse S, Pengpol W, Valairucha S.
Delayed detecti on of esophageal intubation: Thai Anest hesia
Incidents Study (THAI Study) d atabase of 163,403 cas es.
J Merl Assoc Thai 2005; 88(Suppl. 7):S69-75.
240 Warner MA, Warner ME , Warner DO et a1. Periopera tive
pulmon ary aspiration in infants and chi ldren. Anesthesiology
1999; 90:66-71.
241 Roe CR, Wiltse HE, Sweetman L, Alvarado LL. De a th caused
by perioperative fasting and sedat ion in a child with
unrecognized very lo ng chain acy l- coenzyme A
dehydrogen ase deficiency. J Perlia tr 2000; 136 :397-9.
242 Breucking E, Rei mnitz P, Schara U, Mortier W. Anesthetic
co mplication s. The incidence of severe anesth etic
com plication s in patients and families with prog ressive
muscular dystrophy of the Duchenne and Becker types.
Anaesthesist 2000 ; 49 : 18 7-95.
243 Larach MG, Rose nberg H, Gron e rt GA, Allen Gc. Did
anesthetics tr igge r cardiac arrests in patients with occult
myopathies? Anesthesiology 2001; 94:933- 5.
244 Oudesluys-Murphy AM, van Berke l JH, van den Bran d HM.
Unexplain ed sudden deatb und e r genera l anaesthesi a. Lancet
1985 ; 1 :696-7.
245 Schmidt GN , Burmeister MA, Lilje C et a l. Acute b ea rt failure
during spin a l s urgery in a boy with Duchenn e muscul ar
dystrophy. Br J Anaesth 2003 ; 90:800-4.
246 Cuccbiaro G, Rbodes LA. Unu s ual presenta tion of long QT
sy ndrome . Br J A naesth 2003; 90:804 -7.
247 Lagard e MC, Tun e ll WP. Surgery in patients w ith
hemoglobin-S disease. J Pediatr Surg 1978; 13 :605-7.
248 McGarry P, Dunca n C. Anesthetic risks in sickle ce ll trait.
Pediatrics 1973; 51:507 - 12 .
249 Dover SB, P le nderleith L. Moore MR, McC oll KE. Safety of
ge ne ra l anaesthes ia a nd s urgery in acute he pat ic porphy ria.
Gut 1994; 35: 111 2-15.
250 Tabib A, Lo ire R, i\Iliras A et a l. Unsuspected card iac lesion s
associated with sudden un ex pec ted periop e rative death. Eur
J Anaesthes iol 2000; 17:230- 5.
25 t Galli L, Orrico A, Loren z ini Set al. Frequen cy a nd
localization o f mutation s in the 106 exo ns o f the RYR J gene
in 50 indiv idu als w ith malignant hyperthermia. Hum Mulat
2006; 27:830.
252 Hallid ay NJ. Malignant hyperthermia. J Cran iojac Surg
200 3; 14 :800-2.
253 Karger B, Te ige K. Fatal m a lig nant hyperthe rmia: delayed
onset and aty pical course. Forensic Sci lilt 2002 ;
129:1 87 - 90.
254 Prescott RJ, Roberts SP, Wi.lIiams G. Malig nant
hyp erpyrexia: a rare cause of postoperative death. J Clin
Path o11 992; 45:361-3.
25 5 Ellis FR. Malignant hype rpyrex ia . Arch Dis Child 1984 ;
59: 1013-1 5.
256 Watkin s 1. Adverse rea ct ion to neuromusc ular blockers :
frequ e n cy, investiga tion, a nd epidemi ology. Acta
Anaesthes iol Scand 1994; 102(Suppl.):6-10.
25 7 Fox MH, Hunt Pc. Prolonged neuromusc ular block
associated with mi vacurium. Br J Anaesth 1995;
74: 237-8.
258 Pasqu ariello CA, Schwartz RE. Plasm a ch olinesterase
defici ency in a neon ate. Ca n J Ana esth 1993; 40:529- 3 1.
259 Hoffman TM, Wern ovsky G, Wieand TS et al. The in cid ence
of a rrhythmias in a pedia tric ca rdiac inte nsive care unit.
Pedia tr Cordial 2002; 2 3:598-604.
260 Arieff AJ. Fatal pos toperative pulmonary edema: pathogenesis
and litera ture revi ew. Chest 1999; 115:1 371-7.
261 Brouh Y, Paut 0, Tsim a ratos M, Camb o ulives J. Postoperative
hyponatremia in chi ld re n: pathophysiology, diagnosis an d
treatme nt. Ann Fr Anesth Reanim 2004 ; 23 :39-49.
References I 383
262 Choong K, Kho ME, Menon K, Bohn D. Hypo to nic v ersus
isoto nic saline in hospitalised children: a systematic revi ew.
Arch Dis Chi ld 2006; 91 :828-35.
263 Way C, Dhamrait R, Wade A, Walker 1. Perioperat ive fluid
the rapy in c hildren: a s urvey of cu rrent presc ribing pract ice.
Br J Anaesti1es ioi2006; 97:37 1-9.
264 Copl ans MP, Curson I. Dental anaesthesia a nd mOliality.
Br Med J (Ciin ResEd) 1982; 284:661 -2.
265 Copl a ns MP, Curs on I. Dea ths associate d w ith dentistry an d
d en ta l di sease 1980-19 89. Anaesthesia 1993; 48:43 5- 8.
266 Worthin g ton LM, Flynn PJ, Struni n L. Dea th in the de nta l
chair: a n avo idab le ca tastrophe? Br J Anaesth 1993;
80: 13 1-2.
267 De nta l Anaesthesia Comm ittee of the Royal College of
Anaesthetists. Standards and Guideli nesjor General
Anaesthesia jar Del/tistry. London: The Roya l College of
A na esthe tis ts, 1999.
268 Poswillo D. General Anaes thesia Sedation a nd Resusc itat io n
in dentist ry: Repon of a n Expert Working Party for the
Standing Dental Advisory Commi ttee. London: Dep artment
of Health, 1990.
269 Cartvl'right DP. Dea th in the dental ch a ir. A naesthesia 1999;
54: 105-7.
270 Malamed SF. Morb idity, mortality and local anaesthesia.
Primary Dent Care 1999; 6:11-1 5.
271 Lifsc hultz BD, Kenn ey JP, Sturgi s CD , Donoghue ER. Fatal
intrac ranial hem orrhage follo win g ped iat ric oral s urgica l
proced ure. J Fo rensic Sci 199 5; 40: 13 1-3.
272 Keelin g JW. Fetal and Neonatal Pa thology, 4th edn. Lo ndon:
Springer, 2007.
273 WiggleswOlih JS. Perina tal Path ology, 2nd edn. Lond o n:
WB Saund ers, 1996.
274 Po lberger S, Svenningsen NW. Ea rly neonatal sudden infant
dea th a nd near death of full-term infa nts in mate rni ty
wards. Acta Pa ediatr Scand 198 5; 74:861-6.
275 Rodriguez-Alarcon J, Melchor Jc. Linares A et a1. Early neo natal
sudden death or near death syndrome. An epidemiolo g~cal study
of 29 cases. Acta Paediatr 1994; 83:704-8.
2 76 Deh a n M, Imbert MC, Benisvy C et a l. [Sudden newbo rn
infa nt deat h in mate rnity. Anatomo-c1inica l study of 3 1
casesJ. Arch Fr Ped iatr 1992; 49: 609- 16.
277 Pina r H. Postm onem findings in term neonates. Sel11in
Neonato l 2004 ; 9 :289-302.
278 Keeling JW. Iatrogenic disease in the newborn. Vi rchoUis
Arch A Pathol A nat Histo1 198 1; 394: 1-29.
279 Byard RW, Kennedy JD. Dia g nost ic difficulties in cases of
s udden death in infants with mandibular hyp op la sia.
Am J Forensic Med Pathol 1996; 17 :255-9.
280 Wigglesworth JS. Performan ce o f the perin atal a utopsy. In
Perinatal Pathology, 2nd edn . Lo nd on: Saund ers, J 996, p. 36.
28 1 A bman SH, Burc he ll MF, Schaffer MS, Rosenberg AA. Late
s udden unexpected deaths in hospitalized infants with
bronchopulmonary dyspl asia. A m J Dis Child 1989;
143:815- 19.
282 Espagne S, Hamo n I. Thiebau geo rges 0, Hascoet JM. [Sudd e n
death of n eo n ates in the delive ry roo mJ . Arch Pediatr 2004 ;
11 :436-9.
283 Byard RW, Burn ell RH. Appare nt life-threatening events and
infant h olding practices. A rch Dis Child 199 5; 73 :502-4.
284 Frank DJ, Drobish NL. Toy safety in hospi tals: or beware of
parents bea rin g gifts. Ciin Pediatr (Phila) 1975; 14:400-2.
28 5 M illun ch ic k EW, MeAno r RD. Fatal a sp ira tion o f a makesh ift
pacifier. Pediatrics 1936; 77 :369-70.
286 Levene S, Bo nfield G. Accidents on hospital wards. Arch Dis
Child 199 1; 66: 1047-9.
287 Lyons TJ, Oates RK. Falling out of bed: a relatively benig n
occurrence. Pediatrics 1993; 92:125 -7 .
384 I Sudden death in hospital
288 Nimityongskul P, Anderson LD. The likelihood of injuries
when children fall out of bed. ) Pediatr Orthop 198 7 ;
7: 184-6.
289 Finberg L, Kiley J, Luttrell CN. Mass accidenta l sa lt
poisoning in infancy. )AlI!LA J963; 184:187-90.
290 Finberg L. Unfo rgettable patients.) Pediatr 1992;
121 :323-4.
291 Walter GF, Maresch W [Acc id ental sa lin e poisoning in
newborn infants. Morphologic findings and pathogenetic
discussio n]. Klin Padiatr 1987; 199 :269-73.
292 Busch KA, Fawcett J, Jacobs DG. Clinical correlates of
inpatient suicide. ) Ciin Psychiatry 2003; 64: 14-19.
293 Pfeffer CR, Plutchik R, Mizruchi MS, Lipkins R. Suicidal
behavior in child psychiatric inpati ents and outpatients and
in nonpatients. Am) Psychiatl)' 1986; 143:733-8 .
294 Wo lfersdorf M. [Patient suicide in the psychiatric hospita l:
selected results of the Clinic Suicide Wo rking Group Study
1/11 1970-1992 of the 'Su icidal Behavior and the Psychiatric
Hospi ta l' Study Group]. Psychiatr Pra_r 1996; 23 :84-9 .
295 Noren GR, Staley NA, Bandt CM, Kaplan EL. Occurrence of
myocarditis in sudden death in children. ) Forensic Sci
1977; 22:188-96.
296 Knight DM, James RA, Sims DN et at. Sudden death due to
intravenous infusion of hair conditioner. Am ) Forensic Merl
Pathol 1998; 19:252-4.
297 Southall DP, Plunkett MC, Banks M et at. Covert video
recordings of life-threatening child abuse: lessons for child
protection. Pediatrics 1997; 100:735-60.
298 Feldman KW, Hickman RD. The central venous catheter as a
source of medical chaos in Munchausen syndrome by proxy.
) Perliatr Surg 1998; 33:623-7.
299 Goldfarb J, Lawry KW, Steffen R, Sabella e. Infectious
diseases presentations of Munchausen syndrome by proxy:
case report and review of the literature. Clin Pediatr (Philo)
1998; 37:179-85.
300 Kamerling LB, Black XA, Fiser RT. Munchausen syndrome by
proxy in the pediatric intensive care unit: an unusual
mechanism. Perliarr Crit Care Med 2002; 3:305-7.
301 Meadow R. Munchausen syndrome by proxy. Arch Dis Child
1982; 57:92-8.
~~ --------_.
302 Forrest ARW The Investigation and Prosecution of Health
Care Workers Who Systematically Harm Their Patients. LL.M
thesis, Uni versity of Wales, 1992.
303 Kinnell HG. Serial homicide by doc to rs: Shipman in
perspective. BM) 2000; 321: 1594-7.
304 Park GR, Khan SN. Murder and the ICU. Eur) Anaesthesiol
2002; 19:621-3.
305 Yorker Be. Nurses accused of murder. Am ) Nurs 1988 ;
88: 13 27-8, 1332.
306 James DS, Leadbeatter S. Detecting homicide in hospital.
) R Call Physicia11S LOlldol1 1997 ; 31:296-8.
307 Start RD , Cross SS. AC P. Best practice no 155. Pathological
investigatio n of deaths following surgery, anaesthesia, and
medical procedures.) C/il1 Patlwl 1999; 52:640-52.
308 Bush CM, Jones JS, Cohle SD, Johnson H. Pedia tric injuries
from cardiopulmonary resuscitation. Anl1 Emerg Med 1996;
28 :40-4.
309 Maguire S, Mann M, John Net al. Does cardiopulmonary
resuscitation cause rib fractures in children? A systematic
review. Child Abuse Negl 2006; 30:739-5 1.
310 Peterson DB, Beckwith JB, Benson EA. The sudden infant
death syndrome in hospitalized babies. Pediatrics 1974;
54: 644 - 6.
3 II Gill JR, Goldfeder LB, Hirsch CS. Use of 'therapeutic
complication' as a manner of death. ) Forensic Sci 2006;
51 :1127-33.
3 12 Roberts lSD, Gorodkin LM, Benbow EW What is a natural
cause of death? A survey of how coroners in England and
Wales approach borderline cases.) Clin Pathol 2000;
53 :367 - 73 .
313 Buehler JW, Smith LF, Wallace EM et al. Unexplained deaths
in a children'S hospital. An epidemiologic assessment. N
El1gl) Med 1985; 313:211 -1 6.
314 Istre GR, Gustafson TL, Baron RC et al. A mysterious cluster
of deaths and cardiopulmonary arrests in a pediatric
intensive care unit. N El1gl) Med 1985; 313:205-11.
315 Hunt J, Goldring J. The case of Beverley AJlitt. IVIed Sci Law
1997; 37:189-97 .
316 Clothier e. The AliitT InquilY. London: HMSO, 1994.
- ------ ~
- -
 I CHAPTER 20 I
ROAD TRAFFIC ACCIDENTS
IN CHILDREN
Anthony Busuttil

Road traffic fatalities 385 Diffuse axonal injury 393

Investigation of a fatal road traffic collision 387 Whiplash injuries 393
Vehicular collisions 390 Injuries to children in utero 393
Other supervening problems in collisions 391 Other vehicular accidents 393
Pedestrian injuries 392 References 394
Child cyclists 392

ROAD TRAFFIC FATALITIES Table 20.1 International comparisons: road accident fatality
rates per 70 6 papUlation year - 2000 (aged 0-14 yearsj4s
Road traffic deaths are still common, although rates among
countries vary considerably (Table 20.1). They deserve the Iceland 0 Ireland 27
same level of investigation as any other unexpected and Sweden 12 Australia 29
violent death; however, this does not always occur because Japan 13 Belgium 29
of prioritization of other fatal ities. In the thorough investi­ United Kingdom 15 Canada 30
gation of any road traffic death , particularly one that Italy 17 Spain 31
involves a child, a number of sources of inform ation must Germany 19 Czech Republic 32
be considered and integrated to enable a useful and faith­ Netherlands 19 France 33
ful reconstruction of the individual incident and, perhaps Austria 20 Greece 33
more importantly, the development of appropriate preven­ Finland 21 Luxembourg 37
ta tive measures. The detailed information obtained about Scotland 21 Poland 37
the collision from each potential source must be carefully Norway 22 USA 40
collated to ensure that the incident can be reconstructed in Switzerl and 23 New Zealand 47
detail - not only to assist the bereaved in coming to terms Denmark 26 Portuga l 52
with the death, but also for court inquiry (criminal and Hungary 26 Republic of Korea 58
civil) and insurance purposes. 1, 2 If the death can be shown
to have been preventable then steps should be taken to
ensure that similar deaths in t hat particular location or
manner could be avoided in the future. from a high-income family to be killed ,on the road . Chil­
In the UK nearly two out of three road accidents happen dren from ethnic minorities are involved in up to twice as
when children are walking or playing and almost two­ many accidents as the national average while walking or
thirds of child accident victims are boys. As a child gets playing. Also, the risk of being involved in a road accident
old er, the risk of a road accident increases. A child from a when walking or pl ay ing is more than 10 times greater for
low-income family is five times more likely than a child a child with hearing difficulties. Child pedestrian deaths

others are in use and no restraint is available. Children
must not sit in the front seat withou t the right child seat.
• In France, children under 10 years of age are not
allowed to travel in the front seat of a car unless the
child concerned is an infant of 9 months old or
younger and weighs less than 9 kg; however, the infant
must be placed in an appropriate baby seat (rear
facing) unless there is an air bag fitted, in which case
the infant is not allowed on the front seat at all. In the
rear seats, a ll children under 10 years must be in the
right restraint; if the child's weight is between 9 and
15 kg then he or she must use a child seat and if over
15 kg then a booster type seat should be used.
• In Spain, children aged 3 or under must use a suitable
child seat, and child ren up to 12 should use one if
ava il able. If not they can use ad ult seats.
• In Italy, it is illegal for children under 3 years to sit in the
front or rear seats without the proper child seat. Children
aged betvveen 4 and 12 years cannot be a passenger
unl ess they are using a su itable safety restraint seat or an
adaptor for a seat belt. If no seat is available for children
aged between 4 and 12 years then they can use the ad ult
seat belts in the rear if accompanied, in the rear, by
another passenger aged over 16 years.
The design of cars and the restrain ts within them has
exercised engineers and car designers for severa l decades,
with the emp hasis on design ing safer vehicles. Data col­
lected from actual incidents have assisted in ensuring that
proposed designs are effective 4 ­ 6
INVESTIGATION OF A FATAL ROAD TRAFFIC
COLLISION
The inves tigation of a fatal road traffic collision has a vari­
ety of face ts, described as foll ows.
The Scene of the Collision
Data collection should always start at the actual scene of
the fate" road in cident. The deceased child may have been
a passenger in a veh icle, a pedestrian or a pedal cyclist;
more rarely, the deceased child may have been a pillion
rider on a motorcycl e, a passenger on a pedal bicycle or the
occupant of a sidecar of a motorcycle. Establishing this
information should be the point of departure of the inves­
tigation of the particular death.
AJthough it may not be fe as ible for a forensic pathologist
to attend at evelY such scene of a fata l road traffic collision ,
if the accident appears to be likely to resu lt in an eventual
criminal prosecution then it is imp ortant in such cases for
the pathologist to actually attend the scene. When the acci­
dent is a 'hit and run inc ident' it is a lmost imperative that
the pathologist attends, together wit h t he rest of th e inves­
tigative team. In each case, evelY effort sho uld be made to
Fatal road traffic co llis ions I 387
docum ent the scene fully by sti ll photographs and on video
to ensure that the findin gs at the site of the incident are
availabl e for futu re reference, briefing and reconstmction.
It is often the case that the deceased has not been pho­
tographed in situ , as the attendance of ambulance person­
nel and paramedi cs usually results in the child being taken
away to hospital for at least atte mpted resuscitation. Pub­
lic access to the locus and ready visibility of the scene in
the interest of decency may also require the early removal
of the body from the scene of the collision .
Specialists in road traffic collision investigations are
nowadays available7 in most police forces. The scope of their
activity is to document as much as possible at the scene, tak­
ing measurements of such items as skid and tyre marks,8 oil
and other fluid leaks and any other data that can be g leaned.
They will log the location of di scarded items of clothing and
shoes on the roadway, bicycles, broken portions and debris
fro m a vehicl e (for example paint fra gments,9 side mirrors,
lamps,1O glass fragme nts and pools of blood) in an attem pt
at a full reconstmction of the incident. They wi ll eventua lly
produce a sketch of the incident and, if the incident warrants
it, a computerized reconstmction may be attempted. Valu­
able information about such ma tters as the condition of the
road, street lighting and prevailin g weather conditions will
be collected and later included in a detailed report.
An important part of the eventua l investigation, is the
determination of the speed at which the vehicle{s) inlJollJed
in the incident was trauel/ing. lI It has to be kept in mind
that the injuries on injured or deceased individuals invo lved
in t he collision are of little or no help in determining this .
One of the main pointers to impact speed are the markings
left on the roadway by a vehicle attem pting emergency
braking and the movement and transport of objects a long
the road. Any contact with walls, street furnitu re (such as
lamp- posts or signposts) and the defo rmation caused to
such items is more useful in the estimation of imp act speed.
Similarly, any deformation in the actua l framework of vehi­
cles involved can , after an assessment by a mechanical
engineer and appropriate calculations, assist in answering
t his importa nt question. The pathologist or clinician should
be wary about being led into giving definitive answers to
the question of speed estimates, i.e. responses that cannot
be later substa ntiated scientifica lly on cross-examination.
These investigators would also collect any trace evidence
that many be useful to identify the vehicle involved, particu­
larly in 'hit and mn' accidents. Even a small fragment of pain t
from the offending vehicle can be analysed to enable such
detailed information as to the make and year of manufacture
of the vehicle, any recent res prays of the bodywork, etc. 12
Eyewitness Accounts
The second source of data are the accounts given by eye­
witnesses to the elJent. Th ey will be interviewed by police
officers as early as possible after the incident and while
386 I Road accidents and ch il dren
constitute the most common accidental cause of child
fata li ties in the UK.
The speed at which the vehicle is travelling has an essen­
tial bearing on the number of pedestrian fatal ities and speed
reduction measures that have been used on roads (e.g. speed
bumps across the road and limiting the speed limit) have all
been shown to benefit from a diminution of child pedestrian
fatalities. An impact speed below 20 mph results in signifi­
cantly fewer fatalities. The faster the traffic, the greater the
11sk of death and serious injury. When children are hit by cars
travelling at 20mph, 5 out of 100 are killed; most suffer only
minor injuries and about 30 out of 100 suffer no injury at all.
At 30 mph nearly half of all children are kjJled and many are
seriously injured. At 40 mph 85 out of 100 children are killed.
Children make unpredictable pedestrians. Traffic-coping
skill s are complex and children do not have the ability to
judge speed, distance and danger accurately until around
the age of I I years. Even children above this age are easily
distracted and may not always behave as drivers expect.
Young children are also much smaller than adults and so
are difficult for drivers to see.
Children within vehicles who are sitting in the correct
child seat for their size and one that had been fitted properly
usually suffer only minor injuries in a car crash, but arou nd
one-half of all child seats are still not properly fitted. Rear­
facing infa nt seats reduce the risk of fatal injury in a crash
by more than 70 per cent, fon-vard-facing toddler seats by
more than 50 per cent and safety belts by 45 per cent. At one
time in the USA, only 10 per cent of children under the age
of five travelled unrestrained, but they accounted fo r more
than one-half of child deaths in cars. It is important that
child restraints are used properly if they are to save young
lives. A study by the Automobile Association (AA) and
Devon County Council found that 24 per cent of child car
seats were too loose, 21 per cent of the harn esses were too
loose and 21 per cent of the buckles were not fitted properly.
The law app lying to child car seats changed in September
2006 in the UK, meaning that all children under 14 years
should have some form of extra restraint as appropriate to
their weight and age while travelling in a car. Every year
around 30 children aged up to I I years are killed during car
journeys, 400 are seriously injured and over 7000 more are
injured; many of these deaths and iqjuries could have been
prevented if the children had been strapped in correctly.
Children who are not secured with the appropriate child
restraint, such as a chi ld seat, may be seriously injured in a
car accident. A ch ild should never be carried on an adult's
lap. Also, an adult sho uld never try to put an ad ult seat belt
around both him- or herself and a child; in the event of an
accident the child would be crushed by the adulfs weight.
As from 18 September 2006 it became law that a child
under the age of 3 years must have an appropri ate child
restraint when travelling in the front or rear seat. The only
exception to this is travel in a taxi - by law children aged
under 3 years may travel unrestrained in a taxi because the
ordinary ('adult') seat belt in the car is not suitable for them.
A chi ld aged from 3 to 12 years or under 135 cm (about
4 feet 5 inches tall) must use the appropriate child restraint
in the front seat and, if there is one availab le, in the rear.
Children over 3 years old may wear an ordinary ca r seat belt
in a taxi . If there are no seat belts in the rear of the car, by
law a child over 3 years old can travel unrestrained. Chil­
dren who are aged 12-13 years old, or younger children
who are over 135 cm tall, must use the ordinary car seat
belt, if available. Children aged 14 and older a re classed as
adult passengers. They must use the ordinary car seat belt
and it is their own responsibility to do so.
Car seats a re so ld in stages appropriate to the chi ld 's
age, and they range from rear-facing carders for bab ies up
to the weight of 13 kg to forward-facing seats that are suit­
able from 9 months to 4 years. After this age, booster seats
can be used until the child reaches the height of 135 cm.
The new law states that it is illegal to install a rear-facing
child seat in the front of any car when the air bag feature
has not been disabled. Instead, it is recommended that
instead any rear-facing infant seat should be insta lled
behind the driver on the rea r seat.
Frontal ai r bags and side air bags are fitted as sta ndard
practice in new cars within the UK and are designed to dou­
ble up on the effectiveness of the traditiona l three-point seat
belts, whic h have been comp ulsory in new cars since 1967.
Containing 35 L of propellant, air bags inflate to full capac­
ity within 25 milJiseconds, at a speed of 160 mph, which
goes some way to explain why it is imperative to deactivate
any air bag if the child is to be carried in the front. In a crash,
the air bag infl ates very quickly. It could hit anything close
to the dashboard with enough force to cause severe injuries
or even death. Because the back of a rear-facing child seat
sits very close to the dashboard , the seat could be struck with
enough force to cause serious, or even fatal , injuries to a
baby. Even older children (who have outgrown child seats)
are at risk from a deploying air bag if they are not properly
restrained with a lap/shou ld er belt. Reports of eight deaths of
child passengers in crashes involving air bag deployment are
of special concern because they involved low-speed crashes
that the children otl1ervvise might have survived. 3
Up to the age of 3 years it should always be observed
that no child should travel unrestrained unless in the
exception of traveLling in a taxi or licensed car hire vehicle
in which no restraint is availab le.
The use of the co rrect car seat for the child's weight is
essential, although it is a lso velY important to make sure
that yo ur child 's seat is in sta ll ed properly and is the right
design for the particular car. It is important to note, how­
ever, that the law varies in other co untries in rela t ion to
child restraints:
• In Genl1any, for example, children under 12 years of age
and under 1.5 m in height must use an appropriate child
restraint (e.g. a child seat or booster seat while in the
front or rear of a car). Rear-facing seats cannot be used
in front seats that are equipped with air bags. Children
may travel in the rear seats without restraints if all the
- .....
388 I Road accide nts and children
events are still fresh in their minds; they may be inter­
viewed again at a later stage to enable the police to go over
their initial statement with them once again, in more detail
and in the light of results of other investigations.
It has to be remembered than these incidents are always
the cause of major emotional trauma to the onlookers, and
thus it may often be useful to wait until they have had a time
to compose themselves before such statements are taken; at
the scene all that needs to be done is to obtain names and
addresses. Great sensitivity has to be exercised by the inves­
tigating police officers in the timing of taking statements,
particularly from next-of-kin eyewitnesses. On the other
hand, delayed attempted rationalization of the perhaps jum­
bled recollections of the rapid incident may result in prob­
lems in distinguishing fact from rationalization.
In this aspect of the data collection , it is always useful
to interview the paramedics , police officers and sometim es
also fire officers who are called to the scene. The informa­
tion that they may have picked up in the course of their
scene atte ndance, i.e. their involvemen t afte r the collision,
the appearances of the locus of the collision and the posi­
tion of the body when found, is particul arly useful. They
may have had to move vehicles or the body of the child or
to extricate the child 's body from within or under the vehi­
cle. It is then most important to know exa ctly the position
and appearances of the body prior to its movement.
In this respect, if any resuscitation has been carried out
then it is also essential to interview the ambulance crew,
paramedics and the members of the nursing and medical staff
who have been directly involved. This will enable them to
give information as to the methods of resuscitation used, any
drugs administered, and any other changes effected by them.
Clothing of the Decedent
The clothing of the injured parties may also have useful trace
evidence 13 on it, as well as features that indicate that certain
matters had taken place - for example, crushing under a
wheel, contact with the greasy undercarriage of the vehicle,
etc. If the child was taken to hospital, it is often the case that
clothing is cut off with scissors and sometimes has been
completely disposed of before the police have had a chance
to claim it. It is essential that such items of clothing be
retrieved and that those responsible for clothing removal are
instructed to be aware of its potential value as evidence.
Inspection of the Vehicles Involved
Another accessible source of data in fatal road incidents is the
vehicle (or veh icles) involved in the incident. If a mecha nically
propelled vehicl e is involved it is essential that exp erts care­
fully check its roadworthiness, including the brakes. The
vehicle is usually towed or taken away to a police-manned
garage where it is fully inspected and examined in terms of
_. - - - - - - - - ­
-------
problems with its components, seat belts, tyres, steering
mechanism and other features relevant to safe vehicle opera­
tion. A history of the vehicle may be usefully be sought
through maintenance garages.
The deformation and denting on the metallic body sur­
faces of a vehicle that have occurred in the course of the
fatal incident can be extremely usefu l in determining, with
some degree of accuracy, the speed at impact. Mechanical
engineers are able to examin e the vehicle and determine
by mathematical calculation, often assisted by computer­
aided programs , the speed at impact.
The vehicle will also bear in it dents and breakages at
sites of direct impact with the human body. If a child pedes­
trian has been knocked down, an inspection of the vehicle
may well indicate where the initial contact with the child's
body was; this co uld be the front bumper or bonnet of the
car. This inspection may be able to determine whether the
child has been elevated after the initial impact and pro­
pelled upwards to strike the car still moving in his direction
or wheth er the child had fallen beneath the car.
Impacts with pedal cyclists often result in the child being
thrown up in the air if the combined speed of impact is over
16-20 mph (25-32 km per hour). The body will often land
on the vehicle with second imp act: for example a primalY
impact may be with the windscree n and with the secondary
impact with the A-pillars or windscreen of the car. At such
sites of impact, there may have been transfer of hair, blood
and tissu e from the cyclist.
When the deceased child has been an occupant of a
vehicle, it is important to determine where he or she was
placed inside the vehicle, whether the child h ad been prop­
erly restrained and wheth er the restraints available in the
vehicle were appropriate, properly fitted and functional,
and whether there were any other occupants or other items
inside the vehicle that might have resulted in fuliher
impact and injury.
Information from Others Involved
FUliher useful data sources are the narratives from other
persons invoJved in the co llision, not least the driver of any
vehicle involved. In such instances, in addition to docu­
men ting any injuries which they may have sustained, it is
also impoliant to determine the driver's state of health and
whether or not they were incapacitated in any way, or
intoxicated, at the time th at the collision took place.
In all road traffic collisions the drivers of any vehicle
will invariably be breatha lysed for alcohol at the roadside
and a general observation made of them by police officers
present as to their state of orientation\ coordination, pupil­
Jary size and reaction. If the situation warrants it, statute
provides for their medical examination a nd the collection
of further samples for the estimation of alcohol and drugs
in body fluids. More detailed medical examinations and the
collections of specimens for drug testing wi II be carried out
-

at the police station rather than at the scene, or in hospital
if the driver has been injured.
The Decedent
The main source of evidence accessible to the pathologist is
the body of the deceased. If the child has not been unclothed,
it would be essential that the clothing is removed carefully
to avoid any further damage to it and to avoid any loss of
evidence from it as well as any cross-contamination.
The clothing, including the shoes, should be separately
bagged into brown paper bags and given to the police for
further examination as required.
The child's external injuries, no matter how trivial,
should be carefully measured and documented. It is useful
to photograph all aspects of the body. This may not always
be feasible and, in such instances, the importance of a care­
ful description of these injuries is crucial. Injuries with pat­
terns to them must be documented fully photographically.
This may have to be followed, depending on local legal
custom, by the formal identification of the decedent by
members of his family and by police officers who were actu­
ally present at the scene. Even if there is no legal require­
ment, it is essential than the parents are aJlowed early access
to their child after they have been fully briefed about the
appearances of their child by those who have been trained to
so do, for example police family liaison officers.
It is essential that the body of the child is presented to
them in an aesthetic way, perhaps with partial covering or
bandaging to obscure injuries or even following some clean­
ing and reconstruction of the body. The identification
process is fraught with emotional overlay for those who are
doing this in an official capacity, and indeed much more
poignantly so to the members of the family of the deceased.
The latter should be given enough time with the body
and not officiously rushed. They should not be discouraged
from touching the body and if, indeed, the family asks for
mementoes, such a locks of hair or handprints, then these
requests should be met, with the consent of the appropriate
legal personnel. Some families may wish to obtain a Polaroid
photograph of the face of the deceased child. Religious ritual
may also be requested as part of the acute bereavement pro­
cedure. All reasonable requests made by the family should be
heard and, if at all possible, agreed.
It is extremely useful to document bony injuries using
radiology before the autopsy commences. This ensures a
much fuller documentation of such injuries, particularly of
the ribs and vertebrae and in younger children in whom
ossification may be incomplete.
Another useful investigative adjunct, which is now avail­
able to most police forces, is the availability of a portable laser
source of light that enables the child's unclothed body to be
scanned in a dark room; this shows any marks on the body that
have been caused by tyre imprints or patterned injuries that
have resulted from contact with specific parts of the vehicle.
--- -. - ~ -
~ ~
Fatal road traffic collisions I 389
A full autopsy must be carried out and all internal
injuries documented fully being careful to examine closely
the vetiebral column, the rib cage in its entirety and the
scapulae - fractures at these sites are often missed.
If physical abnormalities are present then these should
be carefully recorded. It is also important to check that the
child that has not suffered from any previous significant
medical histolY by look.ing through the child's general
practitioner records; if the family doctor is not known then
this should be checked through the police. If such a history
exists, it is important to give due weight to any relevant
information. If the child suffered from learning difficulties,
epilepsy, visual impairment or hearing problems then these
may all be relevant; with regard to a history of hearing
problems the middle ears have to be opened and conditions
such as otitis media and 'glue ear' have to be excluded.
The autopsy should enable a full documentation of
all injuries, externally and internally. If any injuries show
any patterns in them (e.g. tyre mark imprints) then these
have to be documented fully by photographs and accurate
measurements.
Any paint or glass fragments or any other trace evi­
dence should be collected, particularly if there is some
dubiety as to the vehicle involved .
Even in autopsies on children, particularly in the late
teens, it is important to collect the following samples for
further analysis:
1. Blood for DNA profiling, particularly in 'hit and run'
accidents.
2. In appropriate cases, it may be useful to exclude acute
inhalation of solvents prior to the collision. In addition
to the collection of blood, this may require the
retention of a lung, with its main bronchus tied off
securely and placed inside a nylon fibre bag to avoid
loss of intrapulmonary air content. This enables assay
of the solvent content within the bronchial air.
3. In some cases, it may be necessary to test for alcohol
and other drugs. For these assays, appropriate
specimens of blood, urine, vitreous fluid and liver will
have to be retained on the off chance that these
become necessaty. In the older teens, it is always wise
to consider intoxication as a potential complicating
factor in road traffic collisions.
4. (Control) samples of plucked scalp hair may also come
in useful for matching with any hairs found at the
scene and to exclude any previous exposure to
controlled drugs.
Recent problems with organ retention from necropsies
may make it difficult for the brain to be retained for neuro­
pathological examination. In cases whe,n the death is very
acute, retention of the brain has only limited value in terms
of documenting injury. In delayed deaths, and in cases where
there is known to be a pre-existing cerebral problem (e.g.
epilepsy, hyperactivity syndromes, autism, etc.) it may be
necessary to suspend and fix the brain for an appropriate
 390 I Road accidents and children

period in form alin and to examine it as soon as possible
afterwards. Diffuse axonal injury does occur in babies and
children. If the child has surv ived long enough after the acci­
dent, this pathological damage is discernible with appropri­
ate special staining techniques an d microscopy.
In deaths in the intens ive therapy unit, bacteriological
samples may be required to exclude nosocomial infections
that may have contributed to the death. This may be of
importance if civil litigation is in progress in terms of a
novus actus interveniens and the apportioning of blame for
the death.
In some deaths, with the consent of the family and of the
legal authorities, organs have been harvested for tra nsplanta­
tion purposes. The current dearth of organs compared with
the demands of patients, combined with the current major
successes of transplanta tion, has made this phenomenon
more common than it once was. The contlibution of the
pathologist in such cases is more limited given that it is
assumed that all the organs removed had been intact, healthy
and functional. In such instances to complete the autopsy
examination, it is essential to speak with and to obtain a
report from tbe transplant coordinator as to by whom, wben
and wh ich organs were removed and whether these have
functioned adequately in their seco nd hosts. If it is considered
that some organs or tissues may be required for other pur­
poses it is essential that fully informed consent is obtained
from the family and that tbis matter is approached in as sen­
sitive and measured a manner as is possible.
VEHICULAR COLLISIONS
The data in Table 20.2 show that, by and large there has been
a major welcome decrease in fatal and serious road incidents
involving children. Unfortunately, in relation to less selious
non-fatal transportation injuries there is serious underreport­
ing of such collision-induced injuJies, particularly if the
injuJies are minor. One estimate placed this at 25 per cent of
all child pedestrian casualties l4 and another at 20 per cent
with injuJies in boys being less frequently reported than in
girJ. 1s However, such incidents still occur and they should
warrant the same amount of investigation that any other
sudden suspicious or accidental death is given in the medico­
legal context. Unfortunately, this is not always the case and
important information of potential preventative and safety
improvement importance may remain uncollected.
Car occupants are the largest category of road users
who sustain injulies if they becom e involved in collisions,
although fatalities are more common amongst pedestrians
(Table 20.3). This has been the reason why large amounts of
research, time, money and effort have been spent in ensur­
ing that these injuries are prevented by improvement in car
design, stre ngthening the fra mework of vehicles, the intro­
duction of internal safety devices - seat belts, air bags - and
through legislation improv ing road worth iness of cars and
general safety of cars.
Car collisions cannot be rationalized simplistically from
an isolated assessment of the forces invol ved; in what
appear to be entirely simil ar incidents, one car occupant may
survive with little injury, whereas the other may experience
fatality. The physical size, injury tolerance and the age of the
occupant are all important in the outcome of the collision.
Thus the different physical build of children is an importan t
consideration in predicting the injuries that they will sustain
in road traffic collisions (Table 20.4). The bead of a youn g
child forms a higher proportion of the overall body mass
than that of an older child or an adu lt. The neck of a child
cannot support the weigh t of the head as adequately as in an
adult. The lib cage is smaller and more flexible a nd pliable,
and leaves a larger unprotected area of the abdomen; the
child's pelvis is not as resilient as that of an adult and its cur­
vature is different. All of these factors will have a bearing on
the kinematics in the collision situation and the possible
types of injury that are sustained and their outcome.
Most collisions are frontal, with a tendency for the offside
fron t of the vehicle to become involved. In side impacts, t he
door of the vehicle on the side of the collision remains sup­
ported by the striking object dUling the impact phase and
acts as an unyi elding barrier that is pushed inwards into the
vehicl e's compartment and against which the occupant may

Table 20.2 Child fatalities and casualties in Scotland, England and Wales J997-2005

Scotland England and Wales

Child casualties Fatal Fatal and serious All severities Fatal Fatal and serious All severities

1997 26 745 379 8 229 5708 40751

1998 32 698 3536 174 5382 3991 4
1999 25 625 3196 196 5073 38872
2000 21 561 2999 170 4641 367 15
2001 20 543 2956 199 4447 35361
2002 14 527 2747
2003 17 431 247 8
2004 12 383 2394
2005 11 368 2166

- -- --- . _ -- - ­ --- --- -

-- -
 Supervening problems in collisions I 391

Table 20.3 Average number of road traffic fatalities per year for Obviously, the mass ratio between the two vehicles
1997-2001 in Scotland involved and the velocity change that occurs on impact are
further critical factors, and these are crucial elements that
Deaths Age (years) Total
have to be considered in assessing the force of the impact
0-4 5-11 12-15 and the injuries caused therein. If the struck object is station­
ary and unyielding then the velocity change in terms of
acute deceleration will take place in the impact vehicle only.
Pedestrians 2 (0.01) 8 (0.02) 6 (0.02) 16
In terms of injury, it is self-evident that the reduction and
Driver or rider 1 (0.00) 1 (0.00) 1 (0.00) 3
restriction of movement of the occupants of a vehicle would
Passenger vehicle! 1 (0.00) 3 (0.01) 3 (0.01) 7
tend to decrease or remove propulsion and thus secondary
pillion
impact against the interior of the vehicle. The rational e of a
Figures in parentheses indicate rates per 1000 of the population. seat belt, which is fitted properly and functioning well when
required , is to enable the wearer to 'ride down' the impact
when the rapid deceleration occurs; thereby the energy of
the impa ct is extended over a slightly longer timescale;
Table 20.4 Site affatal injuries in child car occupants by age the longer this time frame, the greater the dissipation of the
Site of injury Head Neck Chest Abdomen Other forces at the time of the impact. The webbing stretch of the
(no.) belt enables this deceleration rate and interval to be length­
ened over a somewhat longer period , thus allowing the
energy of the impact on transportation casualties to dissipate
0-9 months (4) 4 0 0 0 0 plior to contact of the body with an unyielding surface. An
10 months to 19 2 4 3 evaluation of the impact of front seat belt legislation in
6 years (29) Britain showed a marked reduction in injuries to children
7-14 years (33) 20 4 4 4 aged II to J 4 years who were restrained in the front seats of
Total (66) 43 6 8 2 7 vehicles when collisions took place. 20
'Most of the children were unrestrained and about one-third of them If the child has been restrained inside a vehicle, there may
were ejected from the ve hicle. be injuries on the body that indicate that this restraint has
Modified from Andersson and Jonansson 4 6 been in position and functi oned: friction-induced abrasions,
bruising and even fractures - of the clavicle, sternum and
ribs 21 - may follow the line along which a seat belt would
sustain injuries. Collisions with trees, lamp-posts and other have been applied with force at the time of impact. Internal
street furniture, and walls produce a more concentrated load, damage to the body such as intra-abdominal injuries 12,22,23
with the possibility of an increased intrusion into the (including ruptures of the liver and mesentery and major
passenger compartment and thus a higher risk of serious blood vessels 24 ) and intrathoracic injuri es (such as pul­
and even fatal injuries being sustained, particularly trunk and monary and cardiac injuries) may indeed be associated with
pelvic injuries in front-seat passengers as well as head and such external injuries; fatalities due to seat belt use have
neck injuries. In addition to the initial direction of the been recorded in children involved in transpoltation injuries.
principal force at the time of the collision, there will also be In telms of these considerations, given the relative mass of
post-impact vehicle rebound, depending of the type and pedestrian to a vehicle, particularly a high-sided vehicle, the
magnitude of the force of the collision and the part of the car relatively greater speed of the vehicle than of the pedestrian,
that has been struck. This also depends on the size, resilience the decreased resilience of the human body as contrasted
and safety features of the vehicle involved. These phenom­ with metal, and the Jack of protection in the pedestrian are all
ena will introduce a sideways and lateral component to the factors that render such collisions very traum atic, even at low
impact in the immediate post-impact phase, which, in tum, speed. 2s
will have a further direct bearing on both the configuration
of the impact and the contact between the body and other
parts of the inside of the vehicle as a seco ndary impact phe­ OTHER SUPERVENING PROBLEMS
nomenon. Such an understanding of the kinetics of a colli­ IN COLLISIONS
sion has led to improvement in the inner padding in cars and
the removal from the inside of the vehicle of any pointed Occasionally, the occupant of the car who had been poorly
surfaces and edges. Hyperextension of the neck, secondary restrained inside it is thrown out of the ,car and injuries are
to forehead and facial impacts, may result in serious neck sustained on external impact. If the car catches fire with the
and brainstem injuries that could be instantaneously occupants still inside then secondary fire-induced injuries
fatal; 16, 17 in severe intrathoracic injuries in particular there and inhalation of smoke will occur. These changes will result
may be traumatic rupture of the aorta 18 and compressive in a variety of other injuries to the child involved in such an
injuries to the heart. 19 incident.
392 I Road accidents and children

PEDESTRIAN INJURIES
child to fall into the path of the oncoming veh icle, and an
impact below the centre of gravity would cause the child to
If the child is a pedestrian then the primary impact is usually be lifted up into the air, then, as the vehicle proceeds, on to
by the front of the car with one or other side of the body that the vehicle, and after impact with the vehicle to fall behind
happens to be closer to the vehicle, particul arly if the child is the vehicle. If the child is on a bicycle then the centre of
in the process of walking or mnning in front of the vehicle. gravity at the time of the collision has been so displaced
The distribution of the casualties and fatalities in these inci­ further up (and this is so the larger the bicycle) that on
dents are examined further in Tables 20.5 and 20.6. impact the child will be thrown over the vehicle. From an
Depending on the child's height above the ground, dif­ examin ation of that body and the veh icle, the points of pri­
ferent varieties of impact are present. This will determine mary and secondary impact may be worked OUt. 26
whether the prim ary impact with the vehicle is above or The size of the vehicle involved , the age and build of the
below the centre of gravity of the child's body and in this child, projecting parts of the front part of the veh icle and
the kinetics of the child's body after the collision takes the speed at impact are all factors that have a bearing on
place, vary. By and large, the umbilicus is at about the level the severity of the injuries sustained by the child. Even
for the centre of gravity. A collision producing a primary minor impacts at low speed may produce severe or fatal
point of con ta ct above the centre of gravity will cause the head and neck injuries in a young childY
Most pedestrian childhood fa tali ties (about 80 per cent)
occur in built-up areas with in a I-km radius of the child's
Table 20.5 Casualties by speed limit at site of incident, mode of home ; a number of such fatalities occur on the way to or
transport and severity, 7997-2007 (Scotland) from school so that most deaths occur at the times of
school opening and closure during weekdays. Injuries are
Speed limit (mph) more prevalent during the summer months when children
are more likely to be outside their homes.
30 40 50 60 70 Other and
The psychological capabilities and developmental level of
unspecified
the child have a major bearing on their abili ty to cope with
road conditions; they have generally less well-developed
Fatal ski lls relating the perception of movement and velocity, the
Pedestrian 53 5 2 15 9 0 ability to judge distance and depth, the abi li ty to OIientate
Pedal cycle 4 3 2 0 themselves in relation to the source of sound and the ability
Motorcycle 6 0 0 30 2 0 to concentrate for a sufficient length of time, putting aside
Car 22 7 146 21 0 play activities and concentrating while croSSing or using
Bus/coach 0 0 0 0 0 1 roads. They are also not equipped with the ability to inte­
Other 2 0 12 3 0 grate all of this information swiftly and efficiently, and come
to the appropriate evasive conclusions.
Serious
Pedestrian 893 37 8 49 12 2
Peda I cycle 141 6 29 2
Motorcycle
CHILD CYCLISTS
114 10 7 190 2
Car 482 73 23 1187 175 1
These accoun t for about 17 per cent of deaths on the roads.
Bus/coach 54 2 0 1 0
Arguments have raged for the last decade or so as to the
Other 52 9 5 102 33 0
usefulness of cycle helmets for pedal cyclists and the design

Table 20.6 Child pedestrian (0-15 years) casualties in single-vehicle accident: mean per annum, 1997- 2007 (Scotland)

On pedestrian Within zigzag area Within SOm Crossi ng road Other Total
crossing of a pedestrian of a pedestrian in other areas
crossing crossing

CrOSSing road not concealed 68 12 58 804 84 1026

by vehic le
Crossing road concealed 9 22 360 ~1 423
by ve hicle
Sta nding/wa Iking 0 0 0 0 99 99
Other/unknown 0 1 7 46 54
Total 78 13 81 1171 260 1602

of such protective head gear. 28 There can be little doubt that
most pedal cyclists die because of head injuries 29 a nd that
helmets will afford some degree of protection, particularly
in collisions where little force has been expended.
In this vein, a number of individual states and countries
have suggested that legislation may be required on this
matter in concert with legislation concerning the use of seat
belts. In Victoria, Australia, following on from a massive
10-year-long educational campaign, legislation was intro­
duced ensuring that all cyclists wore helmets. 30 ,31 The
helmet-wearing rate tripled and a 70 per cent reduction in
head injury admissions occurred during the periods
1989-90 and 1991-92. There was also a corresponding
decrease in non-head injuries of 23 per cent in the first year
and 28 per cent over the first 2 years.
From a comparison of statistics between countries where
cycling is very popular (The Netherlands, Belgium) and those
where it is a less frequent activity and more of a leisure
pursuit, the incidence of injuries is lower in the former. This
has been extrapolated to the fact that in the Low Countlies
cyclists and motorists are segregated and cyclists have their
own cycle paths and do not mix with other vehicles on busy
roads, thereby decreasing the likelihood of collisions.
DIFFUSE AXONAL INJURY
Brain trauma produces a diverse spectrum of injuries in
which diffuse axonal injury (OAl) is a well-recognized
major contributing factor to long-term disability. Magnetic
resona nce imaging and computerized tomography have
long been thought to cause underestimation of the extent
of OAl because of their lack of sensitivity and therefore
newer and more sensitive imaging techniques have been
developed. Using these techniques with magnetic reso­
nance spectroscopic imaging, an increased rate of detec­
tion was shown in children in whom at one stage it was
thought that OAI did not OCCUr. 32 ,33
WHIPLASH INJURIES 34
Generally, very young people and elderly people are those
less at risk of sustaining whiplash injury from a car acci­
dent. Research suggests that 17 per cent of under-21-year­
olds and just 2 per cent of over-60-year-olds are less likely
to be involved in an accident resulting in whiplash than in
any other type of motoring accident. These results have
come to light despite the fact that younger people, espe­
cially small children, have far smaller and less well­
developed neck muscles. The action of a car coming to a
sudden halt usually causes the whiplash motion to occur.
The impact of a rear-end collision (when one vehicle hits
the rear of the vehicle in front) shunts the car forwards,
forcing the body forward with it while the h ead and neck
are thrown back. The head then tilts downwards towards
the steering wheel and the neck extends forwards. The
Other vehicular accidents I 393
extended head and neck are thrown backwards until they
hit the headrest with the consequence that the soft tissues
of the neck are stretched and torn. This is also known as
hyperextension followed by hyperflexion, and it occurs in
a frontal collision when the neck and head continue to
move forward while the body remains relatively still. 35
Rear-end collisions can result from any number of sce­
narios. For example, the driver in front may suddenly brake
to avoid a cyclist that pulls out in front of him or the car
behind may accelerate more than the leading car at a round­
about. The impact of hitting another car is the equivalent of
hitting a firm surface, such as a wall, at half the speed. Th at
is, rear-ending a stationary car while travelling at 30 mph
has the equivalent force of hitting a wall at 15 mph. Overall,
43 per cent of people who sustain whiplash injuries from car
accidents are involved in rear-end collisions. Side collisions
involve the side of one or more cars being impacted. These
are common in car parks and at T-junctions.
Personal injuries suffered as a result of side collisions
are determined by the part of the car that is struck, the
speed of the vehicles involved, the presence of safety fea­
tures such as side-impact air bags, and the weight and con­
struction of the object that strikes. In total, 35 per cent of
people who have been involved in car accidents and sus­
tained a whiplash injury were involved in side collisions.
A head-on collision involves the front ends of two cars
crashing into one another. This type of collision may result
from a dIiver coming round a bend on a narrow country lane
too quickly or a motOlist swelving into oncoming traffic on a
B-road to avoid an animal. Head-on collisions are often fatal
but only 32 per cent of people who receive whiplash injuries
from car accidents sustain them from head-on collisions.
INJURIES TO CHILDREN IN UTERO
Road incidents may involve pregnant mothers who are
passengers, drivers and, less frequently, pedestrians. 36 ,3 7
Injuries to the pelvis and abdomen may result in severe
damage to the uterus, rupture of the uterus and abruption
of the placental site with dire, often fatal, consequences to
the unborn child. Litigation may focus on whether or not
the road incident has directly caused the death of the fetus
or unborn infant. 38 ,39 The potential contribution of the seat
belt to these injuries may have to be considered 40 - 42 and
indeed pregnant mothers are legally exempt from wearing
a seat belt for this reason. This topic is considered in more
detail in Chapter 10.
OTHER VEHICULAR ACCIDENTS
Childhood accidents on the farm,4J,44 often involving vehicles
or machinery, are not uncommon, particularly in rural envir­
onments; they may require intervention from governmental
agencies involved in sentimental and work-reared health and
safety issues, as well as requiring public inquest.
394 I Road accidents and children

REFERENCES 24 Riches KR. James RA, Gilbert JD , Byard RW. Fatal childhood
vascular injuries associated with seat belt usage. Am)
Forensic Med Pat/lOi200]; 22: 193 - 5.
Basham DJ. TraJfic Account i\lIanagement. Springfield, IL:

25 Beattie YT, Ri chard s D, Belton N et al. Injuries in the

Charles C Thom as, J 979.

adolescent population in Scotland: patterns and types of

2
Clark WE. Traffic Management and Collision Investiga tio n.
injury sustained. Heath Bulletin 1999; 57: 165-74.
Englewood Cliffs, NJ: Prentice Hall, 1982.
26 Hitosuigi H, Takatsu A, Sigeta A. Injuries of motorcyc lists and
3 Centers for Disease Control and Prevention (CDC). Air- bag­
bicyclists examined at autop sy. Am ) Forensic Med Pathoi
assoc iated fatal injuries to infants and children riding in front
J999; 20:251 -5.
passenger seats: United States. MMWR J995; 44:845-7.
27 Wilson MH, Baxter SP, Tenet SP et al. Saving ChildTeI1. A Guide
4 Ashton SJ, Hayes HRM, Mackay GM. Child pedestrian
to Injury PreIJelltion. Oxford: Oxford University Press, J991.
injuries. Proceedings of the Internationa l Meeting on
28 Hillman M. Cycle Helmets: The Case For and Agail1st.
Biomechanics of Trauma to Children. Lyon: lRCOBl,
London: Policy Studies Institute, 1993.
September 1974, pp. 83-100 and 159-70.
29 Thomas S. Acton C. Nixon J et al. Effectiveness of bicycle
5 Roy AP, Mackay GM, GIOYllS PF. So me observations on the
helmets in preventing head injury in children - a case control
modelling of children in car collisions based on field accident
study. 8M) 1994; 308:J73-6.
investigations. Proceedings of the Fifth International Conference.
30 Cameron M, Vulcan A, Finch C et al. Mandatory bicycle helmet
Birmingham: tRCOBl, September J980, pp. 286-304.
uses following a decade of helmet promotion in Victoria.
6 Mackay GM. Children in cars and pick-up trucks. Saudi Med)
Australia: an evaluation. Accidel1t Anal Pre/} 1994; 26:325-37.
1987; 8:123-7. (Also: Seat Belt Loading. Proc. A.A.A.M. 36th
3 1 McDerm ott FT. Bicyclist head injury preventi on by helmets
Annual Conference, Portland , OR, October J992.)
and mandatory we aring leg islation in Victoria Australia.
7 MacKay GH. The role of the accident in ves tigator. ) Foren sic
Ann R Coli Surgeons 1995 ; 7 :38-44.
Sci 1970; 10:245- 54.
32 Tong KA. Ashwal S, Holshouser BA et aJ. Hemorrhagic
8 Hamm ED. Locating an area on a susp ect tire for compa rative
shearin g lesions in children a nd adolescents with
examination to a questioned track. Int ) Forensic
posttraumatic diffuse axonal injury: improved detection
Identification 1988; 38/40: 143-5 I.
and initial results. Radiology 2003; 227:3 32-9.
9 Dabdoub G. Dixon AC, Watson AD et al. The identifica tion of
33 Holshouser BA. Tong KA, Ashwa l S. Proton MR spectroscopic
domestic and foreign automobile manufacturers through body
imagin g depicts diffuse axonal inju ry in children wi th
primer characterisation. ) Forensic Sci 1989; 34: 1395-404.
traumatic brain injury. Am) Neuroradioi 2005; 26: 1276-85.
\0 Dolan DN. Vehicle lights and their use as evid ence. ) Forensic
34 Freem an MD. Croft AC, Rossignol AM. Whiplash-associated
Sci Soc 1971; 11 :69-82.
disord ers: redefining whiplash and its manage ment. Quebec
11 Ambourn RF. The calculation of motor car speeds froll1 curved
Task Fo rce. A critical eval uat ion. Spine 1998; 23: 1043-9.
tire mark s.) Forensic Sci 1989; 29:371-86.
35 Minton R. Galasko CSB. Murray PA, Pitcher M. Whiplash
12 Ryland SG, Kopec RJ. The evidential va lue of automobile
injuries, causati ve studies: final report 1998. Report of the
paint chips.) FOre11s ic Sc i 1979; 241(1):140-7.
Transport Research Laborato ries ISBN: 1-84608-256-0.
13 Monahan DL, Harding HWJ. Damage to clothin g: cuts and
36 Patterson RM. Trauma in pregnancy. Ciin Oustet GYl1eco/
tears. ] Forensic Sci 1990; 35:901-13.
1984; 27:32-8.
14 Tun bridge RJ, Everest JT. Wild BR. Johnstone RA. An In­
37 Corsi PRo Rassian S, Bechelli de Oliviera L et al. Trauma in
Depth Study of Road Accident Casualties and their Injury
pregnant women; analysis of maternal and fetal mortality.
Pattern s. Report RR. Crowthorne: Transport and Road
Trauma 1999 ; 30:239-43.
Research Labo ratory, 1988, pp. 136-7.
38 Stafford PS, Biddinger PW, Zunwaite RE. Lethal intrauterine
15 Ward H, Norrie J. Sang A et al. Urban Safety Project: the
fet al trauma . Am) Ousret GYlleco/ 1988; 159:484-9.
Sheffield Scheme. Contractors report CR J 34. Crowthorne:
39 Weinberg L. Wyatt JP, Busuttil A. Traum atic intrauterine fetal
Departm ent of Transport, Transport and Ro ad Research
spina l fracture following seat belt use : a case report.
Laborato ry, J 989.
] Trauma 2001; 51: 1195-6.
16 Kondo T, Sa ito K, Nishigami J. Ohshima T. Fata l injurers of
40 Mamhews CD. In correctly used seat belts associated with
the brain stem and/or upper cervical spinal co rd in traffic
uterine rupture following v ehicular collisio n. Am ] Oustet
cases; nine autopsy cases. Sci Justice 1995; 3 5:197-201.
GYl1eco /1975; 121:111 5-16.
17 Gunji H, Mitusaawa I, Hiraiwa K. The mecha nism underlying
41 Schoenfelkd A. Ziv E, Stein Let al. Seat belts in pregnancy
the occurrence of traumatic brainstem lesi ons in victims of
and the obstetrici an. Obstet Gynecol Surg 1987; 42: 275 -82.
traffi c accide nts. Legal Med 2002; 4:84-9.
42 Griffiths M. Hillm an G. Usherwood M. Seatbelt hazards in
18 Shkrum MJ, McCafferty KJ. Green RN et a l. Mechanisms of
pregna ncy (Case report). Br ] Obste/ Gynaecoi 1991;
aOltic injuri es in fatalities accming in moto r vehicle
98:320-1.
collisi ons. ) Forensic Sci 1999; 44:44-60.
43 Byard RW. Gilbert J, Lipsett J. James RA. Farm and tractor­
19 Feezko JD, Lynch L, Pl ess JE et aJ. An autopsy review of 142
related fatalities in children in South Australia. ) Paediatr
[non-pen etrating] blunt injuries to the aorta. ] Trauma 1992'
33:846-9. .
Child Health 1998; 34 :139-41.
44 Byard RW, Gilbert J, James RA, Lipsett J. Farm and tractor­
20 Lo wne R, Robelts A, Roy P et al. The effect of UK Seat Belt
related fatalit ies in children. ] Forensic Med Pachol 1999'
Legi s lation on Restraint Usage by Children. Technical Paper
20:73-7. .
Seri es 840526. New York: Society of Automotive Engineers
45 Road Accidents - Scotland 2006. Edinburgh: A Scottish
(SAI), 1984.
Executive National Statistics Publication. 2007.
21 And erso n PA, Rivara FP. Maier RV, Drake C. The epidemiology
46 Andersson A, J onasso n K. Proceedings oj the Si.xth
of seatbelt-associated injuries. ] Trauma 199J; 3:60-7.
Il1ternatiol1al COIlJerenceJor Accidellt and-TraJfic MedicinE.
22 May AK, Chan B, Da niel TM , Young JS. Ante rior lung
Melbourne. 1977 ; 279.
hern iat ion: another aspect of the seat-belt syndrome.
) Trauma 1995; 38:587-9.
23 Reid AB. Letts RM, Black GB. Pediatric chance fractures:
association with intra- abdo minal injuries a nd seat belt use.
) Trauma 1990; 30:384-91.

FORENSIC DNA PROFILING IN CASES
INVOLVING CHILDREN
Alex M Graham and David J Harrison
Introduction 395
Inheritance of genetic materia l 395
Forensic DNA analysis: history and techniques 398
Sample collection and processing 402
DNA evidence and child sexual offence 403
Y chromosome short tandem repeat typing 405
Mixed samples 406
Additional sample problems and solutions 407
INTRODUCTION
numbered 1-22 in decreasing order of size, and one sex
For centuries philosophers have deliberated which came
first, the chicken or the egg. In the context of forensic
paediatrics, the egg - or rather t he fertilized egg - bein g the
precursor and genesis of the livin g child, is the appropriate
place to start off the discussion on the role of genetics. It is,
after all, the fertili zed egg that develops into all 100 trillion
cells that make up the human being, and it is the genetic
information that comes together from mother and father at
the point of fertili za tion that provides the instructions for
the creation and development of the child.] While the result
is obviously of biological relevance, it is the analysis of the
individuality of the inh erited complement of genetic mate­
rial from person to person that has proved of significan t
value in a forensic context. In a forensic investigati on,
genetic evidence may be used to confirm or refute a known
individual as being the likely so urce of trace material or to
identify a body or body remains. This chapter will focus on
the above use of DNA in paternity testing. 2
INHERITANCE OF GENETIC MATERIAL
Humans inherit half of their genetic material from each bio­
logical parent in the fo nn of deoxyribonu cleic acid (DNA)
molecules neatly packaged up into chromosomes. The
mother's egg contrib utes 23 chromosomes, 22 autosomes
I CHAPTER 21 I
Mitochondrial DNA 407
Paternity testing 408
Identification of body remains and missing persons 411
Identification of the 'abandoned baby' or fetal materi al
and avenues for identifying the source of an
unknown profile 413
DNA databases 414
References 414
chromosome, in this case an X chromosome. The father's
spenn contributes another 23 autosomes, again numbered
1-22, plus either an X or a Y chro mosome, which determines
the sex of the offspring.] When an ovum fuses with a spenn
car!)'ing an X chromosom e, resulting in two X chromosomes
in the fertilized egg, the conceived child is female, denoted
XX. When an egg ce ll fuses with a sperm carrying a Y chro­
mosome the offspring is male, denoted XY.4 The chromoso­
mal complement of the fertilized egg and of the offspring is,
therefore, 46 (Fig. 21. 1), which makes up the individual's
genome. Figure 21. 1 shows these chromosomes as a human
male ka!)'otyp e in which the 22 pairs of autosomes are
neatly aligned in descending order of size, followed by the X
and smaller Y sex chromosomes. In a femal e ka!)'otype, the
Y chromosome is replaced by a second X chromosome.
Because it is the male that carries the Y chromosome in
humans, the presence of Y chromosome DNA in a genetic
sample indicates that a male has contributed to the sample,
and thus, this is a useful way of determining the sex of ori­
gin of a sample.
At the point of fertilization it is not possible to predict
whether the sperm cell that fuses with an egg cell carries an
X or a Y chromosome and, hence, which sex the offspring
will be; likewise, one cannot predict which one of the par­
ents' two copies of each of the other 22 chromosome pairs
will end up in the egg and sperm cells (also known as
gametes or germ cells) that combine. The actual combination
396 I Forensic DNA profiling
2 3
6 8
13 14 15
19 20
Figure 21.1 Karyotype of a normal human male showing chromosomes in a condensed state. Chromosomes were stained, photographed,
then cut out, paired with their partner and the pairs aligned in descending order of size. The sex chromosomes, the Xand smaller Y
chromosome, are positioned at the end of the pairs of autosomes. (Courtesy of RMurray, South East Scotland Cytogenetics Service.)
is the result of the parental chromosome pairs randomly and
independently segregating during gamete cell production by
the process of meiosis.4 It is this randomness, in part, that
gives rise to the diversity of genetic material from individual
to individual. Because which of the two possibl e copies of
each of the 23 chromosomes of the parent that enters the egg
or sperm cell is random, there are 223 possible combinations
of the mother's chromosomes in a human egg cell and, like­
wise, 223 possible combinations of the father's chromosom es
in a hum an sperm cell. Therefore, when a human egg cell
and sperm cell randomly fuse, the result is one of 223 x 223
possible different combinations of the parents' original chro­
mosomes, which equates to one of over 713 possible combi­
nations of the parental chromosomes in the resultant
offspring. This helps to explain why siblings, despite being
deJived from the same parents and inheJiting half of each of
their parents' genetic mateJial, have a different combi nation
of chromosomes and a different genetic make-up.
In actual fact there is a further level of genetic variation.
This arises because a process aki n to chromosomal shuf­
fling takes place between pairs of chromosomes, known as
homologous chromosomes, such that the members of a pair
swap segments of DNA to give rise to novel sequences
along the length of the chromosome compared with the
original chromosomes. In this way, one can appreciate the
biological mechanisms that contribute to the interindividual
genetic variation, the result of which we can see all around
4 5
9 10 11 12
16 17 18
21 22 x y
us in terms of the diversity of features and characteristics
that we all possess. It gives us an understanding of why we
feel justified in saying that, with the exception of identical
(monozygotic) twins, it is highly probable, thoug h not real­
istically feasible, to prove definitively that every individual
has a unique combination of chromosomes and a unique
genetic sequence.
Monozygotic twins are the exception, because they are
derived from a single fertilization event when the embryo
splits to form two individuals at an early stage of develop­
ment. Thus, monozygotic twins have the same chromoso­
mal and DNA complement. This precludes existing forensic
DNA analysis techniques from discriminating between their
DNA, and is a factor that should not be overlooked when
claims of the power of DNA evidence are made. Conven ­
tional forensic fingerprint evidence is increasingly coming
under attack for its alleged lack of scientific basis and cred­
ibility an d question ab le reliability, as exemplified by recent
high-profile cases of misidentification, such as the Madrid
bombing and the Shirley McKie case. 5 - 11 One feature of
conventional fingerprinting that is often cited in its favour
is its abi lity to differentiate between identical twins. How­
ever, such cases are equally, if not more, susceptible to the
assumptions and un certainties of the technique. 12
Nearly all body cells of a human possess the full comple­
ment of 46 chromosomes and, as such, are said to be in the
diploid state. The chromosomes are comp artmentalized into

an orga nelle of the cell known as the nucleus. Any cells with
a nucleus are, therefore, a potential source of this nuclear
DNA, which can be used in an investigation. Notabl e excep ­
tions include the egg and sperm germ cells previously men­
tioned, which only have half the number of chromosomes
and are said to exist in the haploid state, and mature red
blood cells (erythrocytes), which are anucleate and lack
nuclear DNA. However, an ucl eated cells are not totally
devoid of DNA because, li ke nucleated cells, they have a
much smaller DNA component which is stored in the cells'
mi tochondria and is known as mitochondrial DNA (mtDNA).
Mitochondrial DNA has its own characteristic inh eri­
ta nce patterns and features which make its analysis a valu­
able forensic tool in its own righ t in some situations. These
will be conside red later in this chapter.
Deoxyribonucleic Acid
While morphologically the members of a homologous
chromosome pair in any individu al look the same (see Fig.
21.1) and possess a high degree of similarity across their
length, they are not identical. This is due to the specific
make-up of the DNA that is transmitted by way of the
chromosomes. DNA lies at the heart of genetic analysis,
and in order to appreciate the techniques employed and the
evidence derived from forensic DNA analysis a basic level
of understanding of the molecule is required.
DNA is made up of three components: a sugar part
(deoxyrib ose), a phosphate part and a nitrogenous base
(Fig. 21.2). The sugar and phosphate components are joined
by phosphodiester bonds to form the sugar-p hosphate
backbone of the molecule, but it is the bases that link two
of these strands together, in what has been likened to rungs
of a lad der bridging two strands of linked sugar-phosphate
units, that are of interest in terms of conferring attributes
and individuality on living beings.
There are four bases, adenine, cytosin e, guanine and
thymine, commo nly referred to as A, C, G and T. These
bases hold the two strands of DNA together in a very spe­
cific way: A can pair only with T (and vice versa) and C can
pair only with G (and vice versa). When correctly paired ,
the two complementary strands can spi ral into the recog­
nizable double-helix form that is so familiar (Fig. 21.2).
It is the specific sequence of the bases along the length
of the chromosomes that may vary between members of a
homologous chromosome pair in an individual and between
chromosomes from different individuals. It is this sequence
variation that differentiates us at the genetic level and which
may be used as a means of tracing a DNA sample to its source.
Genes. Alleles and Genetic Variation
Some of these vari ations may confer phenotypic effects, i.e.
characteristics that can be observed. These lie in regions of
Inheritance of genetic material I 397
Vt=:::ILC:;;:;~_ Sugar
phosphate
backbone
~~~~~~'- Adenine
~=:TI:t=~~~- Cytosine
Thymine -*=::::::Jli=~~
Guanine --h~3:tt=;;>
~=TIt:::~~i\- Nitrogenous
base
L....l_ _ _ _ Base pair
Figure 21.2 Structure of deoxyribonucleic acid (DNA). This
figure represents the double-stranded he lica l structure of DNA.
The molecule comprises two strands of linked sugar and
phosphate molecules. The sugar-phosp hate backbone is held
together by nitrogenous bases forming pairs between adenine
and thymine or cytosine and thymine. These linked strands
then twist to form a helical spiral as shown. (Courtesy of
Talking Glossary of Genetic Terms, National Human Genome
Research Institute, National Institute of Health, USA:
www.genome.gov/10002096.)
the genome called genes, which provide the instructions to
make specific proteins that build and run our body, for
example the pigments that give hair its natural colour, the
enzymes that digest our food and metabolize ingested
drugs, the ion channels found in cardiac cells, or compo­
nents of the immune system. The last two examples may be
of paliicular interest in the investigation of sudden infant
death syndrome (SmS) as particular variants have been
implicated as putative contributors to a subset of SIDS
deaths.13
Di fferent variations of a paliicul ar gene's sequence,
known as alleles, may typically take the form of a change in
one or more base to an alternative base(s), the insertion of
extra bases or deletion of sections of DNA. These differences
mayor may not alter the fun ction or production of the pro­
tein product. For example, the different colours of eyes of
different people result from different alleles of the genes that
determine eye colo ur. 14 Whether a child suffers from the
genetically inherited condition cystic fibrosis depends on
which combination of allelic variants of the gene encoding
the cystic fibrosis transmembrane conductance regulator
398 I Forensic DNA profiling
(CFTR) protein are inherited. 15 Sufferers from this condition
inherit two faulty, or mutated, versions of the gene, which
affect the production or function of the gene product and
give rise to the observed clinical syndrome. 16
Different alleles of particular genes may be of impor­
tance in investigating the cause of death in a child or infant
if they affect the viability of the offspring, but in the major­
ity of conventional forensic work the focus is on regions of
the genome that Jie outside the genes because this is where
maximum interindividual variation in genetic sequences is
believed to exist and, hence, where the power of discrimina­
tion is greatest.
Interestingly, recent findings suggest that the level of
variation between individuals is much higher than previ­
ously thought. Large segments of DNA within and outside
genes are present in the genome of different individuals a
different number of times. This phenomenon, called copy
number variation (CNY), means that variation between indi­
viduals is 5- to lO-fold greater than previously believed. 17.18
That said, it is amazing how similar our genomes are. Prior
to this study it was generally accepted that over 99 per cent
of the genome is the same in every individual, not actually
much of a difference, especially when you also consider also
that there is only a couple of per cent difference between
humans and chimpanzees! 19,20
FORENSIC DNA ANALYSIS: HISTORY
AND TECHNIQUES
Restriction Fragment Length Polymorphisms
From a historical perspective, the advent of forensic DNA
analysis as a means to identify, visualize and trace the
interindividuality of genetic material, enabling a sample
to be matched to its source, was somewhat serendipitous.
Alec Jeffreys, dubbed the 'father of DNA analysis', was not
aiming to develop a forensic investigative tool when he
discovered a potentially groundbreaking way of visualiz­
ing individual genetic specificity in the 1980s; in fact, he
was looking for a human version of the myoglobin gene in
seal meat. 21
Jeffreys used a technique called restriction fragment
length polymorphism (RFLP) analysis in which genomic
DNA is cut or digested in a sequence-specific way by
restriction enzymes; this results in a mass of fragments of
DNA of differing lengths. These fragments can then be sep­
arated on the basis of their size and electric charge by a
process called gel electrophoresis; specific fragments that
share similarity in sequence are visualized by binding a
labelled DNA probe of complementary sequence to the
sought-after sequence, which is shared by a subset of the
fragments. 22 ,23 The result is somewhat akin to a 'barcode',
with DNA from different individuals forming different-sized
fragments that are represented by bands positioned at a dif­
ferent location along the length of the developed 'barcode'
-- - - - -
--
--
Figure 21.3 Deoxyribonucleic acid (DNA) samples analysed by
restriction fragment length polymorphism (RFLP) in a paternity
case. Each column represents a sample: M, mother; F, father;
C, child. It is clear that each individual is different. However,
closer examination of the children's samples reveals that whereas
every band in the profile of child 5 (C s) is seen in one of the
parents, as would be expected, the profile of child 7 (C 7) includes
bands that are not seen in either parent, indicating that this child
is not their biological offspring. (Courtesy of Cellmark Forensic
Services, Abingdon, Oxfordshire.)
(Fig. 21.3). Because of the analogy of comparing bands on
an electrophoresis DNA gel with points of comparison of
conventional fingerprints, Jeffreys named the technique
'DNA fingerprinting'. He claims it was coining this term that
attracted attention to the potential forensic application of
the technique. Using RFLP analysis, it is possible to compare
DNA extracted from biological samples from a crime scene
with those of a suspect; if the DNA patterns match, the sam­
ples most probably came from the same individual. If the
DNA patterns do not show the same pattern of bands, the
suspect can be eliminated as a likely source of the crime
scene sample. This was, in fact, the outcome of the first
application of the technique in a forensic case.
Two 15-year-old girls in Narborough, Leicestershire,
UK, were sexually assaulted and murdered. RFLP analysis
showed that DNA extracted from semen from both crime
scenes came from a common source, and thus a single man
was deemed to have been responsible for both crimes. How­
ever, the crime scene DNA did not match DNA from a blood
sample taken from the prime suspect. Subsequently, the first
mass DNA intelligence screen (DNA man hunt) was
launched. However, in an interesting twist in this investiga­
tion, the DNA screen failed to identify a match to the crime
scene profiles. The offender, Colin Pitchfork, was eventually
identified and convicted after it was overhead in a pub that
a colleague had been persuaded to give a DNA sample on

his behalf. Subsequently, Pitchfork's DNA was found to
share the same RFLP pattern as the crime scene samples.
Pitchfork was convicted of the murders and sentenced to
life imprisonment in 1988. 24 ,25 This case provides the first
example of a suspect being exonerated on the basis of DNA
evidence, an outcome of many subsequent analyses that
should not be overlooked. Although DNA evidence has
often been hailed as being responsible for Pitchfork's con­
viction, some argue that it was, in fact, 'an old-fashioned
tip-off from an informanf 26 that resulted in the case being
solved. However, there is little doubt that the probability
that two non-related individuals will show the same RFLP
pattern by chance is extremely low; this has led to many
convictions when samples from a crime scene and suspect
matched.
Short Tandem Repeat Profiling
PRINCIPLES
Although RFLP analysis is highly effective and shows a
very high level of discrimination, practical issues, namely
the requirement for large amounts of high-quality DNA
(which is often a luxury in forensic investigations), the need
for radioactive materials in RFLP analysis initially, the com­
plexity of analysing and interpreting the results (Fig. 21.3)
and the development of new techniques to analyse DNA
variants, mean that the majority of forensic DNA analysis
today focuses on a specific type of variation called micro­
satellite repeat polymorph isms, rather than the longer mini­
satellite repetitive DNA on which RFLP was based.
A polymorphism simply refers to the occurrence of more
than one allele or variant at a particular region or locus in
the genome within a population. Strictly speaking, to count
as a polymorphism the allele must be present in the popu­
lation at a frequency greater than 1 per cent; rarer variants
or de novo events giving rise to new sequences are termed
mutationsY The important point to be aware of is that dif­
ferences exist and, in terms of differentiating between indi­
viduals, any difference is of relevance.
Interspersed througho ut the genome are stretches of
microsatellite DNA in which short core sequences (4-6
bases) of DNA are found to be repeated in tandem, multiple
times, the exact number of repeats varying from chromo­
some to chromosome and from individual to individual. A
number of these so-called short tandem repeats (SIRs) have
been found to be highly polymorphic with lots of different
alleles, corresponding to the number of times the core
sequence is repeated. STR polymorphisms have proven to be
an ideal tool to use to compare DNA samples and look for
similarities and differences between individuals.
The greater the level of variation in repeat number at an
SIR locus in t he population, the greater the power of dis­
crimination. By analysing the number of times a particular
selection of STRs is repeated on each homologous chromo­
some in an individual, it is possible to build up a profile of the
Forensic DNA ana lysis I 399
individual's DNA. An individual who has an SIR, for exam­
ple THOI, which resides on chromosome II and has the core
sequence AATG, repeated six times on one copy of chromo­
some 11 and nine times on the other chromosome 11 would
be denoted as having a genotype 6, 9 for that SIR. Using esti­
mates of the frequency at which each known allele ex ists
in the population, one can determine the probability that the
particular allelic combination in question would arise in a
person chosen at random from that population. For example,
using data from the US Caucasian population, one can esti­
mate that this allelic combination at THOI would be found at
a frequency of 0.053 (or 5.3 per cent) in the population. 2s
If a crime scene sample sho ws the genotype 6, 9 at
THOI, it wiJl be appreciated that looking at this single locus
or region alone will not be particularly helpful in identify­
ing the individual who left the DNA because many people
in the population are likely to share this genotype. This is
similar to the situation that may arise using conventional
blood group typing as an investigative tool: many people
have the same blood group and thus the discriminatory
power of blood group typing is limited. 29 However, if one
also looks at other STR loci on different chromosomes, then
the more loci that are examined, the less chance there is
that any two unrelated individuals wi ll have exactly the
same combination of a ll eles at every STR examined.
TECHNIQUE
The result, in terms of STR repeat numbers, of analysing a
selection of SIR loci is known as an SIR profile. This can be
illustrated graphically, with the different alleles being repre­
sented as peaks at different positions along a horizontal axis,
an example of which can be seen in the electrophoretogram
shown in Fig. 21.4. These profiles are obtained by extracting
DNA from biological samples from a wide range of possible
sources, the proviso being that the sample contains cells that
contain a nucleus and that these have DNA of sufficient
quantity and quality. Although the erythrocytes in blood
lack the requisite nuclear DNA, the white blood cell compo­
nent makes blood a particularly good source of DNA for
forensic analysis. Blood is frequently analysed in a range
of investigations, for example, child murder or assault, or for
diagnostic purposes in relation to genetically inherited con­
ditions. Semen, saliva or fingernail scrapings may be exam­
ined in the case of alleged child sexual abuse; buccal swabs
may be taken from subjects in a paternity dispute or from
suspects apprehended by legal autho rities. Bones, skeletal
remains or teeth are particularly useful in cases in which
body remains are badly decomposed, have suffered fire dam­
age or have been exhumed, because of the increased resist­
ance of DNA from these sources to degradation. Muscle
tissue, skin, fetal material, placental tissue, urine or faecal
material and inanimate objects QJ1 which DNA traces may
have been deposited (such as cigarette butts, postage stamps,
chewing gum and toothbrushes) are also suitable for DNA
ana lysis (see Butleilo for further references). It is of note that

his behalf. Subsequently, Pitchfork's DNA was found to
share the same RFLP pattern as the crime scene samples.
Pitchfork was convicted of the murders and sentenced to
life imprisonment in 1988. 24 ,2 5 This case provides the first
example of a suspect being exonerated on the basis of DNA
evidence, an outcome of many subsequent analyses that
should not be overlooked. Although DNA evidence has
often been hailed as being responsible for Pitchfork's con­
viction, some argue that it was, in fact, 'an old-fashioned
tip-off from an informanf 26 that resulted in the case being
solved. However, there is little doubt that the probability
that two non-related individuals will show the same RFLP
pattern by chance is extremely low ; this has led to many
convictions when samples from a crime scene and suspect
matched.
Short Tandem Repeat Profiling
PRINCIPLES
Although RFLP analysis is highly effective and shows a
very high level of discrimination, practical issues, namely
the requirement for large amounts of high-quality DNA
(which is often a lux UlY in forensic investigations), the need
for radioactive materials in RFLP analysis initially, the com­
plexity of analysing and interpreting the results (Fig. 21.3)
and the development of new techniques to analyse DNA
variants, mean that the majority of forensic DNA ana lysis
today focuses on a specific type of variation called micro­
satellite repeat polymorphisms, rather than the longer mini­
satellite repetitive DNA on which RFLP was based.
A polymorphism simply refers to the occurrence of more
than one allele or variant at a particular region or locu s in
the genome within a population. Strictly speaking, to count
as a polymorphism the allele must be present in the popu­
lation at a frequency greater than 1 per cent; rarer variants
or de novo events giving rise to new sequences are termed
mutations. 27 The impoliant point to be aware of is that dif­
ferences exist and, in terms of differentiating between indi­
viduals, any difference is of relevance.
Interspersed throughout the genome are stretches of
microsatellite DNA in which short core sequences (4-6
bases) of DNA are found to be repeated in tandem, multiple
times, the exact number of repeats varying from chromo­
some to chromosome and from individual to individual. A
number of these so-called short tandem repeats (STRs) have
been found to be highly polymorphic with lots of different
alleles, corresponding to the number of times the core
sequence is repeated. STR polymorphisms have proven to be
an ideal tool to use to compare DNA samples and look for
similarities and differences between individuals.
The greater the level of variation in repeat number at an
STR locus in the population, the greater the power of dis­
crimination. By analysing the number of times a particular
selection of SIRs is repeated on each homologous chromo­
some in an individual, it is possible to build up a profile of the
Forensic DNA analysis I 399
individual's DNA. An individual who has an STR, for exam­
ple TH01, which resides on chromosome 11 and has the core
sequence AATG, repeated six times on one copy of chromo­
some 11 and nine times on the other chromosome 11 would
be denoted as having a genotype 6, 9 for that STR. Using esti­
mates of the frequency at which each known allele exists
in the population, one can determine the probability that the
particular allelic combination in question would arise in a
person chosen at random from that population. For example,
using data from the US Caucasian population, one can esti­
mate that this allelic combination at TH01 would be found at
a frequency of 0.053 (or 5.3 per cent) in the popuJation. 2o
If a crime scene sample shows the genotype 6, 9 at
THO]' it will be appreciated that looking at this single locus
or region alone will not be particularly helpful in identify­
ing the individual who left the DNA because many people
in the population are likely to share this genotype. This is
similar to the situation that may arise using conventional
blood group typing as an investigative tool: many people
have the same blood group and thus the discriminatory
power of blood group typ ing is limited. 29 However, if one
also looks at other STR loci on different chromosomes, then
the more loci that are examined, the less chance there is
that any two unrelated individuals will have exactly the
same combination of all eles at every STR examined.
TECHNIQUE
The result, in terms of STR repeat numbers, of analysing a
selection of STR loci is known as an STR profile. This can be
illustrated graphically, with the different alleles being repre­
sented as peaks at different positions along a horizontal axis,
an example of which can be seen in the electrophoretogram
shown in Fig. 21.4. These profiles are obtained by extracting
DNA from biological samples from a wide range of possible
sources, the proviso being that the sample contains cells that
contain a nucl eus and that these have DNA of sufficient
quantity and quality. Although the erythrocytes in blood
lack the requisite nuclear DNA, the white blood cell compo­
nent makes blood a paliicularly good source of DNA for
forensic analysis. Blood is freque ntly analysed in a range
of investigations, for example, child murder or assault, or for
diagnostic purposes in relation to genetically inherited con­
ditions. Semen, saliva or fingernail scrapings may be exam ­
ined in the case of alleged child sexual abuse; buccal swabs
may be taken from subjects in a paternity dispute or from
suspects apprehended by legal authorities. Bones, skeletal
remains or teeth are particularly useful in cases in which
body remains are badly decomposed, have suffered fire dam­
age or have been exhumed, because of the increased resist­
ance of DNA from these sources to degradation. Muscle
tissue, skin, fetal material, placental tissue, urine or faecal
material and inanimate objects 0)1 which DNA traces may
have been deposited (such as cigarette butts, postage stamps,
chewing gum and toothbrushes) are also suitable for DNA
analysis (see Butler.30 for fUliher references). It is of note tha t
400 I Forensic DNA profiling

03S1358 _ vWA _

• A t 1500

--~~ImUJ;---------------/}fE~----------~~----.L[ ~~~O

j -
,.------~
m
08S1179

j >----_
III
021S11

jm;---_~
018S51

~
m Ilil
2000
1500
1000
500

05S818 _ 0138317 07S820

----~
J m----------------~~m~------------'m m
A ~H~
Figure 21.4 A short tandem repeat (STR) profi le. An electrophoretogram show ing the deoxyribonucleic acid (DNA) profile of a sample
analysed at nine STR loci (three on each horizontal row) and a sex-determining region, ame logenin (labelled AMEL on the second row),
using the comm ercial Profiler Plus® kit (Applied Biosystems, Foster City, CAl. Each peak represents a different allele and the number in
green below the peaks denotes the number of times the particular STR sequence is repeated. Where two peaks are seen at a particular
STR locus, for example, 14 and 15 for D3S 1358 on chromosome 3, the individual has a diffe rent number of repeats on their homologous
chromosome 3s; where a singl e peak is seen at a particular STR locus, for examp le, 13 for D8S1179 on chromosome 8, the individual has
the same number of repeats on both copies of their homologous chromosome 8s. In this example a single peak representing the X
chromosome is seen at the amelogenin locus, indicating that the sample was from a female. (Courtesy of Dr JR Gilder, Forensic
Biomatics, Inc., Fai rborn, OH, USA.)

with the advent of lo w-copy-number (LeN) DNA analysis,
it is possible to obtain a DNA profile from as little starting
material as a single cell or the few epidermal cells left behind
in a fingerprint, and trace quantities of DNA that may sub­
sequently be analysed may be deposited simply by touching
articles or surfaces. 31-35 However, the evidential value should
be treated cautiously and the serious technical and interpre­
tative issues relating to such evidence should be taken into
account.
Once a DNA sample has been extracted, it is amplified
by a process called the polymerase chain reaction (peR) in
order to generate a sufficient quantity of DNA for detec­
tion. 36)7 peR is a technique that revolutionized molecular
biology w hen it was developed in the 1980s, and its intro­
duction enabled the shift from RFLP to STR profiling. It
provides a simple and readily automated way to replicate
specific preselected regions of the genome so that, from a
small number of initial copies of the specific DNA sequence
of interest, millions of copies of that sequence can be gen­
erated in a cyclical process. During each cycle of the reac­
tion, the number of copies of the targeted templ ate
sequence of interest is doubled, resulting in an exponential
in crease in the tota l amount obtained after multiple cycles.
In the case of STR profiling, the target sequences within
the genome are a selection of STRs distributed on different
chromosomes of the genome. With the use of commercial
kits, it is possible to amplify as many as 10 SIRs in a single
reaction tube . These kits app ly a fluorescent label or tag
to the different STRs to enable them to be detected and
differentiated from each other. The kits have been designed
and validated for specific conditions and quantities of tem­
plate DNA, conditions not met by the aforementioned LeN
analysis. In order to compensate for the trace levels of DNA
that would not otherwise be detected, LeN analysis increases
the number of rounds of peR amplification, but the cost is
an increase in potential artefacts. Spurious alleles can be
detected and genuine alleles may fail to show Up,34 casting
severe doubts on the acceptability of the results as being of
a sufficient evidential weight for a court of law.
Amplification of selected STRs of the sample DNA results
in a mixture of amplified STR fragments of different length
that need to be separated and visualized. This is done by a
process of gel electrophoresis or, more recently, capillary
electrophoresis, which separates the labelled fragments on
the basis of the electric charge and size of the fragments. 38
The fluorescent dye-l abelJed fragmen ts are passed through a
reten tive material (either a polyacryla mide gel or viscous
polymer), wh ich retards their migration through the material
under the action of a potential difference. Shorter fragments
pass through the material more quickly and are detected by
a laser detector and registered as a peak on the electrophore­
togram before the longer fragments. In this way profiles
from different samples can be generated and compared.
In addition to the SIR loci that are analysed, a gene on the
sex chromosomes called amelogenin is also typed in order to
identify the sex of the source of the sample. Although much
of the X an d Y chromosomes differ from each other and are
non-homologous, there are some stretches of the chromosomes


that show homology, and the amelogenin gene is one of
these. There is a difference in size between the X amelogenin
gene and the Y amelogenin gene, which can be differentiated
on the basis of the size of the PCR products when the amelo­
genin gene is amplified in the same way as the SIR loci
described above. When only a single peak corresponding to
the X version of the gene is detected, the sample is taken to
be from a female; when two peaks corresponding to the X
and the Y copies of the gene are detected the sample can be
considered to be from a male.
Considering any locus that is typed, we cannot be
absolutely confident in our interpretation in every case.
Genetics is no exception to the rule that biological systems
have their anomalies, and there are times when a sample
shows only an X amelogenin signal when the true source is
in fact male. As for all the STR loci, it is possible for DNA
to be present at too Iowa concentration or to be too badly
degraded to be detected, or for the reaction to fail; thus, a
negative result does not necessarily mean there is no sam­
ple present.
Potentially more interesting and problematic, in terms of
forensic investigations, are cases in which a mutation may
exist in an individual, resulting in anomalous results. For
example, the Y amelogenin gene has been found to be
deleted in a number of individuals, including specific Indian
populations ; conventional sex typing would suggest that
these males were female, with considerable consequences for
forensic casework. J9 - 42 Such events are not restricted to the
amelogenin locus; mutations giving rise to anomalous
results can occur at any locus.
INTERPRETATION
If two samples analysed by STR profiling give the same
result at all the loci examined, the samples are said to
'match' (Fig. 21.5). A match tells us that the two samples
could have originated from the same source but is not defin­
itive in identifying the source of the sample. Who knows if
one continued to compare additional regions of the genome
whether a difference between the 'matched' samples might
in fact be identified? One could tell this for certain only by
sequencing the entire genome of the two samples, which is
an unrealistic expectation given current technology. So it is
important to remember that failure to show a difference
should not be regarded as confirmation of identity - it is all
down to statistical probability, which will be discussed later.
On the other hand, if two samples do not match, i.e. they
do not share the same combination of repeat numbers ,at
every STR analysed, they would be deemed to have come
from different individuals because there has been a failure
to illustrate identity. Alternatively, the result may be incon­
clusive. This may be because the DNA is of poor quality or
present in too Iowa quantity to produce sufficient data, or
the data that are produced may be ambiguous or one may
be unable to interpret the results, for example, in the case
of mixed samples.
Forensic DNA analysis I 401
M,Xa.,.,:_ •••••••••••••••
-:1=-_____-::::u'--____.J. .P:~
BI:J\....___
IIil IIil m
m m
::JJ m
u u
~--~IIil------~mm~-------L
n~~
m
~____~J~------~u~------~ti~~
IIil IIil m
m m
Suspect 4
r 4000
__~___-.:!J
m mL---~m m.--------..l..t
!-----.J, 2000
III
Figure 21.5 Comparison of short tandem repeat (STR) profiles at
three STR loci (D3S 1358, vWA and FGA) from different sources. The
profiles of samples ta ken from four suspects were compared with
that of a blood stain found at a crime scene. The comparison shows
that only suspect 3 'matches' the blood sample at these loci, i.e.
shares every allele at each locus in common. The rema ining suspects
would be excluded as a source of the blood stain. (Courtesy of
Dr JR Gilder, Forensic Biomatics, Inc., Fairborn, OH, USA.)
It is worth noting that, in an analysis of DNA by STR pro­
filing, scientists are not able to tell if the sequence of As, Cs,
Gs and Ts of two matched samples identified as having the
same-sized SIR PCR product and concluded to share a par­
ticular number of repeats at the SIR locus is actually the
same over the whole of the SIR region. What is being looked
at are differences in lengths of sections of the genome rather
than the actual sequence across the entire length. It is possi­
ble for two amplified STR products to show the same length.
indicative of being the same, while actually differing in
sequence when they are examined in greater detail.
The investigative value of a match is typically presented
in terms of the 'random match probability' (RMP). which
expresses the likelihood that the match occurs by chance in
the population. The RMP is the result of a calculation using
statistics derived from population studies and assumptions
that originate from the study of population genetics, in
 402 I Forensic DNA profiling
particular the Hardy-Weinberg equil ibrium. 43 ,44 Analysis
has shown that in a stable, randomly mating popul ation,
devoid of selective pressures that would alter the frequency
at which existing alleles are found in the population , it is
possible to predict the frequency of the possible genotypic
combinations of these alleles in the population. 45 ,4G Using
statistics, one can get an idea of the likelihood th at the
given profile would arise by chance in the population from
which the frequency estimates of the different a lle les have
been derived and, hence, the weight of the evidence. One
should not lose sight of the fact that the data derived from
the population databases are based on predictions
established from relatively small samples of individuals
in the popUlation; the results genera ted are based on
probabilities.
The RMP is calculated by mult iplying together each of
the predicted fr equen cies of occurrence of the profiled
genotype for the STRs that have been typed in the sample
using t he product rule to determine the probability that the
given genotype would occur by chanceY The product rule
can be used because each STR typed is situated on a differ­
ent chromoso me and, as describ ed at the start of this chap­
ter, c hromosomes segregate to the ga metes randomly and
are inherited independently of each other.
Th e number of STR loci analysed depends on whichjuris­
diction one is in and, thus, the pow er of discrimin ation ofthe
resultant RMP. In the UK, 10 STR loci are routinely exam ined
at present, whereas in the USA 13 loci are looked at, the
latter number providing a greater power of discrimination.
Despite the fact that such analyses are typically prese nted in
terms of extremely sma ll RMP chance matches, in the order
of one in a billion in the UK COUlts 48 and smaller in the USA,
it is of note that Jeffreys has argued that we should be
analysing as many as 16 STR loci to increase the power of
discrimination in light of the increasing number of individ­
uals entered on the national DNA databases aga inst which
the crime scene profiles are compared. 49 This is because the
greater the number of profil es on the DNA databases with
which unknown samples are compared, in order to identify
a match, and thus a source of the sample, the greater t he
probabi lity of a chance match.
One only needs to look aro und at differe nt ethnic and
racial populati ons to see that interpopulat ion genetic dif­
fere nces exist. As such, it is important to generate popul a ­
tion databases for different race groups because certain
alleles will be found to ex ist at a higher frequency in some
populations than in others, for exa mple, the frequ ency
of the nine-repea t allele at the D1 3S317 STR lo cus is
4.5 times higher in the US Hispanic popul ation than in the
US African-American popul ation (Hisp anic frequen cy =
0.15357; African-American = 0.0 32955) (data taken fro m
Butler2B ) . If one were to calculate the RMP for a sample
based on data from a popul ation other th an the true source
of that sampl e, one may end up with an RMP that misrep­
rese nts the true likelihood of a match. Whether the differ­
e nces would be statistically significant w he n dealing w ith
---
--
.
---
-~~~~~~~~ - ---~~~~~~~~~~~~--
RMPs of the order of one in billions or less is debata ble, but
for a calculation based on an incomplete profile in which
not all the STR loci tested have generated a result, thus
resulting in hig her probabilities fo r the RMPs being calcu­
lated, discrepancies may be more significant. The issue is
ge nerally addressed by calcul ating an RMP based on the
population da tabase that generates the most co nservative
results, i.e. those most favourable to the defence in cases in
which a suspect's DNA is found to match a DNA sample
fro m a crime scene.
Further issues regarding population structure may be
take n in to account when deriving the predicted genotype
fr equencies for population databases when one is dealing
with populations with a hig h level of inbreed ing. It will be
apparent that the closer the relationship of two indiv idu als,
the more DNA they will have in commo n and, thus, the
more likely it is for individuals from inb red populations to
share a particular STR allele than would be the case for two
non-related individuals in a randomly ma tin g population ,
i.e. the conditions of random mating fo r Hardy-Weinb erg
equilibrium are not satisfied. If a susp ect a nd the true per­
petrator of a crime are blood relatives then their genotype
frequencies will not be independent and a correction fa ctor
or para meter measuring the popul ation differentia tion/
e
substructure, denoted as or FsT , is required. 47 ,48.50,51
An alterna tive to the RMP for the presentation of DNA
evi dence is likelihood ratios. A likelihood ratio expresses the
ratio of two alternative probabilities of the same evidence
under different hypotheses. In the context of DNA evidence,
when a suspect's DNA profile matches a crime scen e sample
profile, there are two possible hypotheses : either th e sus­
pect's profile matches because his biological sample was
found a t the Clime scene or the sample came from some
unrelated individual who has not been arrested and who
happens to match the suspect by cha nce. The like lihood
ratio is usually calculated by dividing the likelihood of the
prosecu to r's hypoth esis (the first above hypothesis) by the
likelihood of the defender'S hypothesis (the second above
hypothesis).47, 5! In its Simplest form, the likelihood ratio is
the inverse of the RMP. However, things become more com­
plicated when alternative scenarios for the evid ence that is
seen are presented, or if the DNA sample is a mix ture from
more th a n one contributor. To give a feel for the numbers, a
likel ihood ratio of 1000 or greater provides velY strong sup­
port for the prosecution's hypothesis, a likelihood ratio of
100-1000 provides strong supp ort and a likelihood ratio of
10-100 provides moderate supportY
SAMPLE COLLECTION AND PROCESSING
The ge nera l principles underly ing DNA evidence as applied
to a crim inal case involving a child, e.g. the identification
of a n unknown child's body or remains or other forensi c
investigations, are no different from those of an adult con­
text. Typically, a DNA sample believed to be central to the
­
--- -

crime or investigation - be it a blood sample, a sample of
saliva swabbed from a bite mark or, in the case of alleged
child sexual abuse, semen extracted from underwear, bed
clothing or an intimate sample - is analysed to produce an
STR profile, which is used to link a suspect to a crime scene.
The profile is compared with a reference profile taken from
a suspect, identified on the basis of additional evidence in
the case or, if no leads are originally present to identify a
suspect, with profiles lodged on a national DNA database. A
profile on its own is of no investigative value unless there is
a profile with which to compare it. Samples on such DNA
databases may be associated with a known and named indi­
vidual or they may be from anonymous samples collected
from previous unsolved crime scenes, which enables sus­
pects to be matched with crime scene profiles or crime
scenes to be associated via matches.
The legislation governing the power to collect DNA sam­
ples and retain profiles on the databases varies from nation
to nation. In the UK such powers are wide-ranging, and
DNA can be collected from anyone suspected of a record­
able offence and profiles retained indefinitely, regardless of
conviction or acquittal of the suspect (Police and Criminal
Evidence Act 1984 [amended]), but in other jurisdictions the
power of the criminal investigative authorities may be
much less.
From the point of identification of a crime scene or sub­
ject from which a sample for DNA analysis is collected to
the point at which a profile is generated and interpreted,
strict adherence to protocols, an unbroken chain of custody
of the evidence and careful sample handling must be
adhered to. 53-55
The risk of contamination of a sample and the need for
correct handling to avoid it cannot be overemphasized; pro­
fessional bodies have described procedures and guidelines to
help ensure quality control and quality assurance in these
processes, for example, the Federal Bureau of Investigation
(FBIl-appointed Technical Working Group on DNA Analysis
Methods (TWGDAM), now known as the Scientific Working
Group on DNA Analysis Methods (SWGDAM), and the DNA
Advisory Board (DAB).53,55,57 A case may be lost if doubt can
be cast on whether samples have been correctly handled ­
risk of contamination, compromise of a sample or poor
handling - the most noteworthy of which is the trial of
O.J. Simpson, accused of the murder of Nicole Brown Simpson
in 1994;30,58 false convictions may be secured as a result of
bad practice. 59
DNA is ubiquitous: we are all continuously shedding DNA
and leaving a trace of our movements in this way. In fact, we
need not even have been at a location ourselves in order for
our DNA to be found there. It may be deposited by secondary
or tertiary transfer, being carried and subsequently deposited
elsewhere by an animate or inanimate object. 5O - 55 Individuals
who pass through a crime scene after the event could unwit­
tingly leave their DNA, implicating themselves in the process.
Likewise, it is possible for DNA to be present at a scene prior
to the event and collected when samples were taken,55 or,
DNA evidence/child sexual offence I 403
depending on the nature of the sample, a perpetrator may be
able to suggest a plausible reason for the presence or his or
her DNA at a Clime scene, leading to termination of an inves­
tigation. 57 It is crucial, therefore, that there is evidence to link
samples directly to the event; that all persons entering the
crime scene and crime scene officers, scientists or other indi­
viduals collecting samples take precautionary steps to avoid
contaminating evidential samples with their own DNA, be
this from sneezing, coughing, shedding skin or the like on to
the sample; that clean equipment and gloves are used when
handling each sample; that protective clothing is worn to
limit contamination from the sampler; and that the samples
are securely sealed and appropriately transported and stored
to prevent degradation of the sample. 53 During processing of
the samples, it is imperative that case samples and reference
samples are handled separately and that samples that have
not been amplified by PCR are kept away from those that
have been amplified. It takes only a minute amount of ampli­
fied contaminant DNA to swamp any DNA present in a case
sample, resulting in only the contaminant amplified DNA
being detected, at the expense of the genuine sample DNA,
when the two are simultaneously subjected to PCR. Plas­
ticware and consumables used in the analytical steps are
another potential source of contamination as they may be
contaminated when they enter the laboratory,55,68 and the
problems of contamination are particularly apparent when
LCN analysis is conducted. 34 ,54,59 Forensic laboratories have
sbict standard operating procedures to which they adhere,
such as the DAB Quality Assurance Standards of recommen­
dations for forensic laboratories, in order to standardize and
validate procedures. 70
Despite all these efforts, errors will arise, be they the
result of human factors, the nature or quality of the sample
or technical or biochemical artefacts, and this should not be
overlooked. 71 - 73 There is debate as to whether the exceed­
ingly small RMP figures that are presented in court are
unfairly prejudicial when they are not accompanied by esti­
mates of these errors, particularly when the RMPs presented
are smaller than the probability of laboratory error. How­
ever, some argue that a meaningful statistic for an error in
a particular case cannot be provided. 74
DNA EVIDENCE AND CHILD SEXUAL OFFENCE
The major problem facing the forensic investigation in cases
of alleged child sexual abuse, and which is not unique to
DNA evidence per se, is that the majority of cases arise ret­
rospectively, which presents several difficulties in terms of
gathering forensic evidence (see Chapter 1). A number of
other issues that relate to DNA analysis in such cases are
also of relevance to DNA evidence in a more general sense,
for example mixed samples, DNA quantity considerations
and Y chromosome STR typing; discussion and introduction
of these in the context of sexual abuse investigation seems
appropriate.
404 I Forensic DNA profiling
In the typical case of sexual abuse that comes to attention
only months or years after the event, the opportunity to
obtain DNA samples has passed by the time investigation
begins. However, other indications, such as evidence of
pregnancy or sexually transmitted infections, wh ich may be
investigated using DNA techniques, can be highly probative
should the opportunity for collection of immediate post­
coital samples not present itself. The occurrence of sexually
transmitted infection in a child is unlikely in the event that
sexual contact had not occurred, though vertical transmis­
sion is a possibility, and one also needs to be aware that con ­
sensual sex with an infected partner may have taken place,
rather than abuse. In one study, genital human papillo ­
mavinls (HPV) infection, detected by the presence of HPV
DNA from perineum and vaginal samples, was found in
some of the sexually abused preadolescent girls examined
but not in children in whom sexual abuse was excluded, the
majority of the HPV DNA-positive cases being subclinical
infections.75 It is noteworthy, however, that the incidence of
sexually transmitted infections in victims of child abuse is
low, a reflection of the generally low incidence of infection
in perpetrators. In a recent study in Edinburgh, Scotland,
sexually transmitted infection was seen in less than 10 per
cent of referrals. 76
The appropJiateness of DNA testing for sexually trans­
mitted infectious agents using nucleic acid amplification test
(NAATs)77 in suspected cases of child abuse has been ques­
tioned because of t he potential for false-positive results, with
a possible consequence of false convictions. 78 .79 However, it
is paramount that the well-being of the child is prioritized in
any investigation and that the child's health is not compro­
mised at the expense of gathering forensic evidence. While
NAATs may not be as specific as traditional culture tests,
they show increased sensitivity. From a medical perspective,
the downside of a false- positive result is limited to the cost
and side-effects of superfluous antibio tic treatment com­
pared with the long-term potential serious health conse­
quences for the individual of untreated infection. 8o From a
legal perspective, given that culture tests are sti ll the only
test currently accepted by the UK courts 81 and some labora ­
tories no longer conduct culture tests, having implemented
NAATs alone,82 there are issues relating to obtaining suitable
forensic evidence. The recommended approach is to take tvvo
swabs so that a n initial NAAT-positive result can be con­
firmed by culture,81 though this involves additional distress
and discomfort to the child. If pregnancy follows sexual
abuse, it may be possible to gather evidence as to the likely
father by way of DNA paternity testing of offspring or
aborted fetal material ,83,84 a technique that is discussed later
in this chapter.
Although child sexual abuse includes a spectrum of
activities from intercourse to physically less intrusive sexual
ab use of a minor, whether samples taken for DNA analysis
are informative depends on many factors. These include the
nature of samples in relation to the offence; the time since
t he offence; whether a male perpetrator ejaculated during
the abuse, was azoosperrnic, oligos permic or used barrier
protection; and whether the victim subsequently washed,
changed his or her clothes, ate, rinsed his or her mouth, uri­
nated or defecated, all of which can diminish the quality or
quanti ty of detectable DNA, or even remove all traces of
detectable DNA. In the event that a child does present in an
acute case, conflicting opinions have been expressed about
the value of ta king intimate fore nsic samples up to 72 hours
after the event in prepuberta l victims 85 as recommended by
the American Academy of Pediatrics. 86 It is important to
weigh the distress and discomfort that may be inflicted on
the victim against the likelihood of evidence being found.
Christian et al 85 argue that, if one is seeking evidence of
semen or sperm in a vaginal swab sample to provide objec­
tive evidence of sexual contact with a child who does not
have a previous sexual histoty, data detived from adult cases
is inappropriate to prepuberta l situations. While the litera­
ture reveals marked differences in survival time of sperm
cells in different body orifices, the generally accepted maxi­
mum times for observin g sperm heads range from 24 hours
in the mouth to 7 days in the vaginal cavity, and in the
region of 2-3 days in the anus and rectum .87,88 This argu­
ment 55 may be va lid in the case of cytological observation of
sperm cells or the detection of semen by, for example, acid
phosphatase activity or the presence of phosphate-specific
antigen (PSA), which have a shorter period of detection BG ,89
However, g iven the sensitiv ity of DNA analysis and its abil­
ity to detect trace material refractoty to other detection
methods, particularly using LCN techniques,31-34 one should
not discount the possibility of obtaining DNA evidence
beyond the time periods suggested by Christian et al. 85
In one study, evidence of male DNA, detected by Y-STR
ana lysis, was found in over 90 per cent of vaginal or ana l
swabs taken from child sexual assault victims up to 72 hours
after the incident,9o supporti ng the fll1dings of others that
Y-STR DNA amplification can produce positive resu lts in
swabs cytologically negative for spelm cells. Such negative
results wo uld usu ally have led to tennination of biological
investigation,91 illustrating the va lue of conducting the DNA
analysis. The presence of Y-chromosome DNA in a high
vaginal swab from a fe male would be interpreted as tanta­
mount to evidence of intercourse, and in the case of a minor,
where consensual sex is not a defence aga inst statutoty rape,
is highly probative evidence of abuse.
There are practical considerations, in the case of a sma ll
child, in ensuring that internal samples genuinely represent
material exclusively of an internal nature. The anatomy of
a young girl does not lend itself easi ly to collection of such
samples, and care has to be taken to avoid contact with
external regions of the child 's body. There may be perfectly
acceptab le re asons for male DNA to be present on a child's
body, particul arly if male relatives have frequent contact
with a child, and sampling of t his material could lead to
inappropriate conclusions being drawn, incrimin ating an
innocent person. In this respect the use of the colposcope
with the magnification and illumination that it affo rds has
- .

assisted in the collection of non-contaminated intimate
samples.
Y CHROMOSOME SHORT TANDEM
REPEAT TYPING
As mentioned above, if one can detect the presence of
Y-chromosome DNA in a sample then, barring syndromes
with abnormal chromosome complement or structure, one
can conclude that male DNA is present. Such evidence is
particularly useful in the case of all eged sexual abuse by a
male against a female, although typing the Y chromosome of
a sample has add itional valuable applications, including
paternity testing and the identification of missing persons.
If an internal vaginal sample is taken from a girl under the
legal age for sexual intercourse and Y-chromosome DNA is
detected, this is convincing evidence of abuse.
In theory, one could test for male DNA using the amelo­
genin locus by conventional SIR typing, the X and Y amelo­
genin alleles being of different size. The problem with this
approach is the frequently encountered imbalance in the rel­
ative quantity of any male DNA that would be present in sam­
ples routinely taken from the female victim in an alleged case
of sexual abuse, i.e. vaginal, anal or oral swabs. Such samples
typically comprise a vast excess of female DNA, usually from
host epithelial or inflammatory cells, with only traces of male
DNA, typically in the form of sperm cells, although epithelial
and inflammatory cells may also be present, the last two
being morphologicaJly indistinguishabl e from the female
equivalents. Given the imbalance in the amount of male and
female template DNA added to the PCR, there would be pref­
erential amplification of the female DNA to the extent that
male DNA may go undetected. This is analogous to the situa­
tion described earlier of a DNA sample to be typed being con­
taminated with PCR-amplified DNA, due to poor laboratory
practice, and swa mping the sma ll amount of DNA present in
the sample, leading to preferential re-amplification and detec­
tion of the contaminant DNA. As ever, absence of detectable
DNA is not evidence of absence.
One way to get round this problem is to type a series of
SIR loci which are specific to the Y chromosome (Y-SIR
typing),92,93 so that female DNA is not amplified in parallel
and a male-specific profile or Y-SIR haplotype results. There
are hundreds of STRs distributed on the Y chromosome,94
and by typing a chosen set of these one can identify the par­
ticular Y haplotype of an individual male, i.e. the closely
linked alleles of that individual at the chosen loci. One gets
more information about the source of the samp le using
Y-SIR profiling than if one was just to type the amelogenin
locus to determine maleness because of the variation that
exists in the Y chromosome between non-related males and
between different ethnic, racial and geographic groups.
However, the power of discrimination of Y-STR typing in no
way approaches that of conventional autosomal STR typing.
This is because the Y chromosome is passed from father to
Y chromosome STR typing I 405
son, essentially unchanged in sequence, so all male offspring
through the generations descended from one particular male
will share the same Y chromosome and, hence, the same
Y-chromosome haplotype. Thus, given a Y-chromosome
haplotype alone, it would not be possible to conclude that
the man who was the origin of a DNA sample was its source,
or whether it came from a blood relative such as his brother,
son, father, grandfather, grandson, or uncle on his father's side
of the family.
A number of internet-accessible Y-SIR databases have
been established,92 which vary in the number of SIR loci for
which informatio n is avail ab le. For example, the frequ ently
used and largest Y-SIR haplotype reference database (YHRD)
was established in 2000 with the objectives of generating
' reliable Y-SIR haplotype frequen cy estimates for minimal
and extended Y-STR haplotypes to be used in the quantitative
assess ment of matches in forensi c and genealogical case­
work' and assessing the 'ma le population stratification
among world-wide populations as far as reflected by Y-SIR
haplotype frequency distributions'.95 The database can be
searched for frequency data and geographical distribution
data of a particular haplotype, both of which may aid in a
forensic investigation ; however, the haplotypes entered cur­
rently vary in the number of SIR loci for which data are
available. In 1997, a 'minimal haplotype' was recommended
that typed seven 10ci 96 for which data are ava ilable for all
entries on the YHRD SIR Database. Since then, European
labo rato ries have moved towards typing nine loci , and
SWGDAM recommends typing 11 loci. 97 Kits for these and
additional loci are commercially ava ilable, and current prac­
tice is moving towards 17 loci being analysed routinely,
further increasing the discriminatory power of the analysis.
It is likely that the foc us in the future will move towards
supplementing the information derived from Y-SIRs with
single-base sequence variations, called single nucleotide
polymorphisms (SNPs).
Various ways are used to try to enri ch a fema le/male
mixed sampl e for the male cell component in order to assist
in the analysis and maximize the chance of getting an inter­
pretable Y-SIR haplotype data. For example, differential
DNA extractio n techniques may be employed, which take
advantage of the increased resistance of sperm cells to par­
ticular reagents routinely used to digest cells during the DNA
extraction step of ana lysis. Epithelial or other non-sperm
cells are initially lysed and separated from residual sperm
ceJls, which can then be processed separately?O,98 Antibodies
specific to sperm cell antigens coupled to magnetic beads 30
or laser microdissection (LM) techniques 30 ,99.lOo may be used
to directly capture sperm cells and, more recently, fluores­
cence in situ hybridization in conjunction with LM has been
developed as an effective way to isolate not only sperm cells,
but also male epithelial and inflammatory cells. 101 , I02 The
latter technique presents a solution to the problem of
deposited semen samples that are devoid of sperm cells as a
result of a medical conditions or vasectomy, and offers a
way to include not only sperm cells, the subject of previous
406 I Forensic DNA profiling
analysis, but also additional male cells, which will boost the
quantity of extractable male DNA in small samples.
As with conventional STR typing, the result of Y-STR
typing can lead to an individual being excluded as the
source of a sample, if the two Y haplotypes do not match.
The results may be inconclusive, for the same reasons as dis­
cussed previously, or they may lead to an individual being
included as a possible source of the sample, when the sam­
ples show sufficient similarity that they could have origi­
nated from the same individual. In this case, in a court
setting, the findings are typically presented as being unable
to exclude the suspect (the individual found to match the
haplotype) as the donor of the crime sample, but also unable
to exclude all patrilineally related male relatives, as well as
an unknown number of unrelated males, as potential
donors.92 Courts often seek some kind of statistical meaning
to this conservative statement. The RMP calculation
described above is not appropriate because, unlike the auto­
somal STR loci that are typed for conventional STR profiling,
which are located on separate chromosomes and, hence,
inherited independently of each other, the Y-STR loci are
inherited together as a group of linked loci on all but rare
occasions. Instead, an indication of the probability of a coin­
cidental match may be evaluated using the counting method
or a Bayesian-approach haplotype surveying method. 92 • lo]
The counting method is the simpler and expresses the num­
ber of times the given haplotype is found in the database
that has been searched and upper-bound confidence limits
that can be placed on the haplotype's frequency. The signif­
icance of inferences made from these calculations is highly
dependent on the size of the database used. The 20th release
of the aforementioned YHRD STR Database included 46831
seven-loci 'minimal haplotypes'. Of these, 44974 were typed
for the nine European recommended loci and 17935 for 11
SWGDAM recommended loci. 95 The US National Center for
Forensic Science is currently in the process of creating a
National US Y-STR Haplotype Reference Database, which
will contain data on ll-loci haplotypes. The database is to be
created by compiling and consolidating existing govern­
ment, commercial and academic Y-STR databases and will
allow continuous updating of haplotypes entered on the
database by retyping existing samples as new STR loci are
introduced to create extended haplotypes.lOJ.l04 Making
allowances for advancing technology and new typing tech­
niques in this way shows foresight, but raises some of the
highly contentious issues that exist in relation to whether
only profiles of samples or actual DNA samples of DNA
database entries should be retained once an entry is lodged
on a DNA database.
In the context ofY-STR typing and forensic applications,
interesting data are coming forth suggesting a way in which
the common biological attribute of the Y chromosome
shared by male relatives might be exploited in conjunction
with another commonly shared feature of many male rela­
tives, namely their surname, in societies in which the custom
is for the family name to pass through the male lineage. 105.106
Sharing a surname in such societies significantly elevates
the probability of sharing a Y-chromosome haplotype,
leading to the proposal that the establishment of a large
surname-Y-chromosome haplotype database might be of
intelligence value in a case devoid of other leads and when
male DNA central to the case has been identified, for exam­
ple, in a rape case. By comparing the Y-chromosome haplo­
type of such a sample against the database, a potential
surname of the source might be identified and lead to iden­
tification of a pool of potential suspects. Because many indi­
viduals share a Y-chromosome haplotype, existing databases
do not attempt to attach identity to an entry. There are unde­
niably many barriers to the success of this proposed
approach, including non-paternity events, surname chang­
ing, adoption, the existence of more than one founder of a
particular surname, whether the theory holds true for com­
mon surnames and mutations. Civil liberty and privacy
issues in establishing and using such a database need to be
addressed. The method would not be able to conclusively
exclude or include any individual as a perpetrator, and more
research is required if the proposed approach is to become a
realistic option, but it is interesting to be aware of new
opportunities to exploit genetic information that may come
to fruition.
MIXED SAMPLES
Y chromosome-specific short tandem repeat typing is of
value when more than one male perpetrator may have con­
tributed to a sample because it reduces the number of possi­
ble allelic combinations that one is trying to analyse and is
useful in the event of a mixed male and female sample.
Where more than one individual contributes to a sample.
then multiple peaks representing multiple alleles may be
seen on the electrophoretogram of a conventional STR
analysis. If two individuals contribute to a sample then up to
four different alleles could be detected at anyone locus if
both individuals were heterozygous at the locus; in the case
of three contributors there could be up to six different alle­
les. Such scenarios introduce additional practical and inter­
pretative problems over conventional sample analysis. 107 - 11l
How does one know which of the possible allelic combina­
tions to attribute to the different contJibutors? One approach
is to attempt to allocate alleles on the basis of the strength
of the signal (peak height or peak area on the electrophore­
togram), working on the assumption that if one of the con­
tributors' DNA was in excess of the other at the start of the
PCR, then the alleles initially present in greater quantity
would produce larger peaks after amplification. This may not
hold true in every instance, however, and in cases where
near-equivalent quantities of DNA from more than one indi­
vidual are present, all signals detected being of similar inten­
sity, the approach will not be possible. If one considers the
scenario of five or more peaks, things get even more compli­
cated in trying to detem1ine how many different individuals
406 I Forensic DNA profiling
analysis, but also additional male cells, which will boost the
quantity of extractable male DNA in small samples.
As with conventional STR typing, the result of Y-STR
typing can lead to an individual being excluded as the
source of a sample, if the two Y haplotypes do not match.
The results may be inconclusive, for the same reasons as dis­
cussed previously, or they may lead to an individual being
included as a possible source of the sample, when the sam­
ples show sufficient similarity that they could have origi­
nated from the same individual. In this case, in a court
setting, the findings are typically presented as being unable
to exclude the suspect (the individual found to match the
haplotype) as the donor of the crime sample, but also unable
to exclude all patrilineally related male relatives, as well as
an unknown number of unrelated males, as potential
donors. 92 Courts often seek some kind of statistical meaning
to this conservative statement. The RlVIP calculation
described above is not appropriate because, unlike the auto­
somal STR loci that are typed for conventional STR profiling,
which are located on separate chromosomes and, hence,
inherited independently of each other, the Y-STR loci are
inherited together as a group of linked loci on all but rare
occasions. Instead, an indication of the probability of a coin­
cidental match may be evaluated using the counting method
or a Bayesian-approach haplotype surveying method. 92 , IOJ
The counting method is the simpler and expresses the num­
ber of times the given haplotype is found in the database
that has been searched and upper-bound confidence limits
that can be placed on the haplotype's frequency. TI1e signif­
icance of inferences made from these calculations is highly
dependent on the size of the database used. The 20th release
of the aforementioned YHRD STR Database included 46831
seven-loci 'minimal haplotypes'. Of these, 44974 were typed
for the nine European recommended loci and 17935 for 11
SWGDAM recommended loci. 9s The US National Center for
Forensic Science is currently in the process of creating a
National US Y-STR Haplotype Reference Database, which
will contain data on ll-Ioci haplotypes. The database is to be
created by compiling and consolidating existing govern­
ment, commercial and academic Y-STR databases and will
allow continuous updating of haplotypes entered on the
database by retyping existing samples as new STR loci are
introduced to create extended haplotypes. 103 ,104 Making
allowances for advancing technology and new typing tech­
niques in this way shows foresight, but raises some of the
highly contentious issues that exist in relation to whether
only profiles of samples or actual DNA samples of DNA
database entries should be retained once an entry is lodged
on a DNA database.
In the context of Y-SIR typing and forensic applications,
interesting data are coming forth suggesting a way in which
the common biological attribute of the Y chromosome
shared by male relatives might be exploited in conjunction
with another commonly shared feature of many male rela­
tives, namely their surname, in societies in which the custom
is for the family name to pass through the male lineage. lOS, 106
Sharing a surname in such societies significantly elevates
the probability of sharing a Y-chromosome haplotype,
leading to the proposal that the establishment of a large
surname-Y-chromosome haplotype database might be of
intelligence value in a case devoid of other leads and when
male DNA central to the case has been identified, for exam­
ple, in a rape case. By comparing the Y-chromosome haplo­
type of such a sample against the database, a potential
surname of the source might be identified and lead to iden­
tification of a pool of potential suspects. Because many indi­
viduals share a Y-chromosome haplotype, existing databases
do not attempt to attach identity to an entry. There are unde­
niably many barriers to the success of this proposed
approach, including non-paternity events, surname chang­
ing, adoption, the existence of more than one founder of a
particular surname, whether the theory holds true for com­
mon surnames and mutations. Civil liberty and privacy
issues in establishing and using such a database need to be
addressed. The method would not be able to conclusively
exclude or include any individual as a perpetrator, and more
research is required if the proposed approach is to become a
realistic option, but it is interesting to be aware of new
opportunities to exploit genetic information that may come
to fruition.
MIXED SAMPLES
Y chromosome-specific short tandem repeat typing is of
value when more than one male perpetrator may have con­
tributed to a sample because it reduces the number of possi­
ble allelic combinations that one is trying to analyse and is
useful in the event of a mixed male and female sample.
Where more than one individual contributes to a sample,
then multiple peaks representing multiple alleles may be
seen on the electrophoretogram of a conventional STR
analysis. If two individuals contribute to a sample then up to
four different alleles could be detected at anyone locus if
both individuals were heterozygous at the locus; in the case
of three contributors there could be up to six different alle­
les. Such scenarios introduce additional practical and inter­
pretative problems over conventional sample analysis.107-lll
How does one know which of the possible allelic combina­
tions to attribute to the different contlibutors? One approach
is to attempt to allocate alleles on the basis of the strength
of the signal (peak height or peak area on the electrophore­
togram), working on the assumption that if one of the con­
tributors' DNA was in excess of the other at the start of the
PCR, then the alleles initially present in greater quantity
would produce larger peaks after amplification. This may not
hold true in evelY instance, however, and in cases where
near-equivalent quantities of DNA from more than one indi­
vidual are present, all signals detected being of similar inten­
sity, the approach will not be possible. If one considers the
scenario of five or more peaks, things get even more compli­
cated in tlying to determine how many different individuals
--

might have contributed to the sample. There are numerous
scenarios depending on whether the contributors are het­
erozygous for alleles or homozygous for alleles and whether
individuals share alleles in common . Mixture analysis is
complex and should not be considered definitive. loa Sophis­
ticated software is available to assist in the interpretation of
these profiles,112 but one is justified in being wary of over­
reliance on algorithms, just as one is justified in being wary
of human error and misinterpretation or evaluation of such
complex scenarios.
ADDITIONAL SAMPLE PROBLEMS
AND SOLUTIONS
Clearly, there are many practical and technical hurdles to
be overcome when analysing DNA in the case of alleged
child sexual abuse. The problems of failing to obtain a
sample of suitable quality or quantity will determine
whether an interpretable profile or Y haplotype is obtained,
be it mixed or otherwise. However, if one spreads the net
wider and looks at the ge neric situation where the identity
of the source of a DNA sample is being sought (this could
be samples collected in connection with abuse, murder,
kidnap, the identification of the remains of a body follow­
ing a natural or man-made disaster, or any other investiga­
tion one may wish to consider), then other pitfalls may
hinder successful analysis, over and above the potential for
contamination of DNA irrelevant to the case in hand,109
which has been touched on ea rlier (see p. 403).
Amplification of a sample by PCR may be a complete or
partial failure due to partial or complete degradation of
the DNA or as a result of the presence of PCR inhibitors in the
sample that have not been removed during extraction of the
DNA. Degradation may be the result of enzymatic, chemical
or environmental factors,30, 113 including warm humid con­
ditions, contamination with soil bacteria or gastric contents,
or exposure to the air, fire or chemicals. Fai lure to amplify
DNA may be caused by exposure to ultraviolet (UV) light or
by a variety of contaminating substances depending on the
sample, e.g. humic compounds in soil, haem in blood, poly­
saccharides and bile salts in faeces, melanin in hair or
tiss ue, urea, certain lubricants used on proctoscopes or
specula when collecting sampl es or by leather and textile
dyes such as indigo, used to dye denim, to name a
few. 30,I09,114 Clearly, not only the collection, handling and
storage of the samples, but also the environment and treat­
ment experienced by the sampl e since its deposition have to ,
be considered.
Deliberate attempts may be made by a perpetrator to
eliminate incriminating evidence, for example, setting fire
to remains, which can adversely affect DNA analysis. STR
typing of cremated remains is considered unreliab le and by
some to be unsuitable for forensic purposes,115 although
others have been able to get successful STR profiling
results even from badly charred remains. 116,11 7
Mitochondrial DNA I 407
MITOCHONDRIAL DNA
In some cases there may be no way around the problems of
poor quality or quantity of a sample, but it is worth bearing
in mind that an alternative to nuclear DNA as an investiga­
tive tool can be found in the form of mitochondrial DNA
(mtDNA). STR and Y-chromosome typing use nuclear DNA
as a template and length of SIRs as the basis of discrimina­
tion and source attribution of a biological sample. However,
our cells also have a DNA component contained in their
mitochondria, the organelles that are the site of energy pro­
duction in the celJ. 118 The mitochondrial genome is much
smaller than the DNA content of the nuclei, being approxi­
mately 16.5 kilo bases compared with over 6.4 billion bases
in the DNA of the 46 nuclear chromosomes. 119 However,
although our cells contain only one nucleus, with one copy
of each maternally and paternally inherited allele that are
STR typed, there may be hundreds of mitochondria per cell,
with each mitochondrion having an average of four or five
copies of mtDNA. It has been estimated that on average there
are 500 mtDNA molecules per cell, 120, 121 which is highly
advantageous if one is faced with very sma ll biological sam­
ples with nuclear DNA at sub-SIR detection levels because
the higher copy number of mtDNA per cell increases the
likelihood of recovery of detectable DNA.
Another advantage of mtDNA is its increased resistance to
degradation, so that in samples in which nuclear DNA can no
longer be analysed it may be possible to get genetic informa­
tion by analysing the residual mtDNA component. For exam­
ple, mtDNA from human skeletal remains from the Vietnam
War which had been exposed to extreme environmental con­
clitions, including heat and humidity, for at least 17 years has
been successfu lly extracted and analysed despite unsuccess­
ful attempts to type nuclear DNA.122 Mitochondrial DNA has
also been sequenced from ancient remains including the Ice­
man, also known as Otzi, found in the Alps in 1991 , who
lived more than 5000 years ago,i23 as well as from Nean­
derthal fossilsl 24 and from the mammoth. 125 It is worth bear­
ing in mind, however, that the environment experienced by
these remains may have been conducive to survival of DNA,
for example, the frozen state of the Iceman. Successful
extraction of mtDNA, far less nuclear DNA, is in no way
guaranteed, particularly when the sample has been subjected
to adverse environmental conditions.
While mtDNA may have its advantages, there are down­
sides too, the main one being the low power of discrimina­
tion. In a somewhat analogous way to inheritance of the
Y chromosome, mtDNA has a characteristic inheritance
pattern, being inherited only through the maternal line. It
is only the mother of any offspring who passes on mtDNA
to her sons and daughters. Mitochondrial DNA that is pres­
ent in the sperm cell does not enter or survive in the fertil­
ized egg. 121 There have been reports suggesting rare
instances of paternally inhelited mtDNA, but others have
questioned the authenticity of these findings as being truly
paternally derived mtDNA sequences. 126 It is generally
408 I Forensic DNA profiling
accepted that, with the exception of mutation events, all
siblings and maternal relatives have the same mtDNA, so it
is not possible to identify the source of a particular mtDNA
sample to the degree of specificity with which nuclear DNA
may be attributed.
For forensic purposes mtDNA is analysed by sequencing
two specific sections of the mitochondrial genome that show
the most interindividual variation, namely hypervariable
regions I and II. The sequence results of an analysis are
reported in terms of differences relative to a reference
sequence l27 of the hypervariable regions. 121 When an
unknown sample and a known sample are compared and
they share the same sequence, it is not possible to exclude
the known sample as being the source of the unknown sam­
ple, but, analogous to the case ofY-STR typing, nor can one
exclude all individuals originating from the same maternal
lineage. If more than two differences in nucleotide sequence
are seen, the known sample is excluded as the source of the
unknown sample. If there is one difference in the sequences,
the result is said to be inconclusive to account for mutations
that have been seen to occur from mother to child. 128 Mito­
chondrial DNA has a higher mutation rate than nuclear
DNA, and tracking the conservation of a particular sequence
through generations and the emergence of mutations has
been a useful tool in genealogy studies, acting like a molec­
ular clock, and has provided support for the 'Out of Africa'
hypothesis of human genesis from a woman who lived
200000 years ago. 129. 130
A consequence of mtDNA's susceptibility to mutation can
give rise to a situation of heteroplasmy whereby an individ­
ual can possess mtDNA molecules with different sequences.
As there are multiple mitochondlia and multiple mtDNA
molecules, if a viable mutation alises in a subset of these
molecules they may be passed on as the cells divide, giving
rise to mixed populations of mtDNA molecules in an indi­
vidua1.121.131 Whether more than one mtDNA sequence is
detected on sequencing a sample from an individual exhibit­
ing heteroplasmy will depend on the relative proportions of
the different populations of mtDNA molecules in the sample
analysed, which may vary from cell to cell and from tissue
to tissue. Heteroplasmy appears to be a more common phe­
nomenon than previously thought 13l and the consequences
for forensic analysis should be taken in to account when
interpreting results. This appears to be particularly relevant
in relation to hair samples, which are often found at crime
scenes and used as evidence in a case. Because the root of
the hair is needed for nuclear DNA analysis and is lacking in
many hairs from crime scenes, mtDNA is more frequently
performed and heteroplasmy has been found to be more
common in hairs than in other tissues. 132 Probably the most
famous example in which DNA evidence was found to
exhibit heteroplasmy was in skeletal remains identified as
Tsar Nicholas II from exhumed remains believed to belong to
members of the Russian Royal family murdered at Ekaterin­
burg in 1918. The investigation concluded that both Tsar
Nicholas II and his brother exhibited heteroplasmy.1 33,134
As for Y haplotypes, a statistical estimate of the signifi­
cance of a match is needed when mtDNA sequencing fails to
exclude an individual as being the source of an unknown
sample. The process is similar to the Y haplotype evidence,
being a reflection of the frequency at which the given
mtDNA sequence occurs in established mtDNA databases. 121
PATERNITY TESTING
Inheritance of mtDNA and the Y chromosome provides a
means of tracking blood relationship through the genera­
tions, but often one wishes to have a more definitive means
of detelmining the biological parentage of a child. Proof
of paternity or maternity may be needed to settle inheritance
disputes, child maintenance responsibilities or immigration
applications, not to mention the psychological and relation­
ship issues and problems that go along with doubts in trust
and fidelity of relationships. The principles behind paternity
testing may also be employed to assist in the identification
of the remains of an unknown when no reference sample of
the individual is available, if there is evidence to support
parentage by individuals who can provide samples.
DNA paternity testing uses STR typing and the princi­
ples behind the manner by which DNA is inherited from
our parents. It will be appreciated that, because any indi­
vidual inherits half of his or her genetic material from each
biological parent (one copy of each of the chromosome
pairs), the presence of genetic material that is not consis­
tent with this mode of transmission, i.e. of SIR alleles not
present in the alleged parents, is strong evidence that the
alleged parents are not in fact the biological parents.
In most cases, the biological mother is known and one is
seeking to determine whether an alleged father could be the
biological father, though it may be the father who is known
and it is maternity that is being investigated. Alternatively,
one may be faced with the more difficult situation that only
a parent and alleged child are available for analysis 135 or it
is necessary to determine, in the case that the mother is
known, whether two children have the same or a different
father. 136 In the first of these scenarios in which the mother
is known, if one SIR types the mother, the alleged father and
the child, and compares the mother's alleles with those of the
child at each SIR locus, it may be possible to determine the
allele that the father must have passed to the child, the 'obli­
gate allele'. For example, if the mother has alleles 15, 16 at a
particular SIR locus and the child has alleles 13, 15 then the
child must have inherited allele 15 from the mother and,
hence, must have inherited allele 13 from the father. If the
father has alleles 13, 14 then he could not be ruled out as the
father on the basis of DNA evidence, but nor could any other
man carrying allele 13 at this locus. If, on the other hand, the
mother was 13 , 15 and the child was 13, 15, then it would
not be possible to nalTOW the father'S contribution to a sin­
gle allele because the mother could have passed on either
allele 13 or allele 15. A man with either a 13 or a 15 allele

could have been the father, i.e. there would be more possible
men who cou ld have fathered the child. Obviously, one looks
to more than a singl e STR locus when conducting this analy­
sis to determine if the alleged father has alleles consistent
with fatherhood at each locus typed; the greater the number
of loci examined, the greater the evidential weight.
Tf a n alleged father cannot be excluded as the father,
on the bas is of his genotype being inconsistent with those
required to fi t the genealogy, then one needs to have a sta­
tistical measure of the weight of this evidence as indicative
of paternity. This is most commonly presented in terms of
the combined paternity index (CPl), though the probability
of paternity, which is based on a Bayesian approach, is also
w idely used. 121, IJ7 - 139
The paterni ty index is a likelihood ratio w here the proba­
bility of obtaining the observed genotype in the child at a
particular locus, given that the alleged fa ther is t he biologi­
cal father, is di vided by the probability of obtaining the
observed genotype, g iven that a random man is the father. Tn
order to determine the probability of obtaining the observed
genotype, given that a random man is the father, one uses
the allele frequencies of the obligate allele from population
databases of the race of the alleged father. Tn order to obtain
the probability of obtaining the observed genotype, given
that the alleged father is the biological father, one looks at
his genotype. If he is homozygous for the obligate allele the
probability would be 1 because there is no alternative allele
that he could have passed on to a child; ifhe is heterozygous
for the allele the probability would be 0.5, because there is
an equal chance that he passed on either of his altern ative
alleles. 121, IJ8 Having calculated the paternity index for each
locus, a CPI is calculated by multiplying the paternity
indexes for each locus analysed. A cpr of 100 is accepted to
establish 99 per cent probability of paternity and a CPI of
greater than 1000 indicates that the probability of paternity
is greater than 99.9 per cent. A CPI of 100 is the genera lly
accepted minimum standard for inclusion,140 although some
laboratories use a hi ghe r cpr as their minimum.141
In the context of paediatric cases, paternity testing is
not just applied to the situation where one wishes to iden­
tify the father of the subj ect of the case; t he subject may be
the mother, for instance in the case of statutory rape. By
DNA profiling feta l material from a borted pregnancies and
from the child (mot her) one can determine the likelihood
that a given suspect was t he father an d, hence, guil ty of
statutory rape. 8J
It is important to appreciate that no CUlTent DNA test is
definitive and there a re a numbe r of issues relating to pater­
nity testing that need to be considered. As stated earlier,
relatives share a highe r proportion of their genome than
random men, so problems may arise if a relative of the real
father, for example, father, brother or son, is being consid­
ered as the biological father. One study found that, in over
31 per cent of cases, an un cle could not be excluded as the
father, with five cases showing no mismatches. These five
cases would have resul ted in paternity by the uncle being
Paternity testing I 409
identified had he been the only putative father tested. 142
Other studies also highlight the problem of relatives being
mistaken as a biological father; 14J a way to determine more
definitively if a given allele is inherited from a given indi­
vidual would be of great value in preventing potentially mis­
taken conclusions.
Determining the parental origin of alleles would also be
of help in cases in which the mother is not available for test­
ing and DNA is availabl e only from the child and alleged
father, in the case that t he mot her and child share the same
heterozygous genotype and when multiple male relatives
are suspected of incest. A novel test has been proposed that
migh t be app licable to these cases. It would look at genomic
imprinting that resul ts fro m epigenetic modifications, which
change the chemical structure of the DNA without altering
the underlying DNA sequence.144-146 Certain genes have
been shown to exhibit this phenomenon, commonly by way
of DNA methy lation, with the changes being parent specific.
It is possible to detect these changes; thus, there may be a
way of determining the parental origin of a particular allele.
The approach has been proposed for forensic investigations
and paternity testing, 147, 148 and it will be interesting to see if
it becomes a routine part of forensic DNA analysis, add ing
strength to support the RMP, where not only the same alle­
les are shown to match a suspect but a common geno mic
imprinti ng pa ttern is observed. On a related vein, analys is of
epigenetic differences has also been proposed as a potential
way to differentiate between monozygotic tw ins. As dis­
cussed above, monozygotic twins inherit the same ge ne tic
sequence, and it has been shown that they are epigenetically
indistinguishable at birth and in the early years of life ; how­
ever, older twins exhibit detectable differences in epigenetic
modifications of particular sequences t hat were acqu ired
during life. This may be one way to differentiate between
identical twins.149
While the increase in shared alleles between relatives
may lead to problems in identifying the real father in a
paternity case, it is worth mentioning that, because relat ives
have more alleles in common, the effect of th is in incestu­
ous relationships gives rise to an increased level of homo­
zygosity in the offsprin g, which may be used as an ind icator
of incest. 150
CPT or alternative statistics may be reported when an
alleged man is not excl uded as the biological father. But on
the basis of ge nea logy, how different do the alleles need to
be from the alleles we would expect the father to have in
order to lead to an exclusion? DNA paternity testing works
on the same basis as blood group testing, proposed in the
19205, 151 and late r refined, as a means of investigating
paterni ty disputes, but may add a greater degree of discrim­
ination. Rather than looking at the phenotypic effect, i.e. the
observed effect of our genetic material, namely what blood
group an individual is, 29 DNA typ ing looks at the genetic
material itself, which provides more information. Despite
frequently being overl ooked today, blood group testing was,
and still is, an effective way of, relatively quickly and
 410 I Forensic DNA profiling
cheaply, excluding an individual as being the source of a
samp le or the biological father of a child. If the blood groups
did not match a suspect when trying to identify the source of
a sample, then a suspect could be excluded or, in the case of
a paternity dispute, if the offspring's blood group was not
consistent with the possible combinations that could have
arisen from the mother and alleged father, then the individ­
ual in question is excluded .
Despite this appa rentl y clear-cut log ic, things may not
be quite as simple as far as DNA profiling is co ncerned, for
reasons to be explained below, and, from histo rical cases,
court decisions have not always followed the scientific
theory and evidence for blood group testing. For instance,
in the famou s paternity dispute in 1946 involving the actor
and director Charlie Chaplin, the apparent clear impossibil­
ity of fatherhood, on the basis of blood group testing, was
not viewed as bei ng conclusive, and Chaplain was ordered
to pay child support to the child of Ms Joan Berry.152.153
In 1983, Shaw l52 stated 'Motherhood has always been a
biological certainty; now fatherhood will be as well. We will
have come one step closer to equality of sexes.' Little did
he anticipate the consequences of errors that have been, and
no doubt will continue to be, made in conducting in vitro
fertili zation (NFl. A growing number of cases have come to
light in which embryos have been transferred to the wrong
. woman, so that she has been impl anted with another
woma n's eggs; the wrong spelID has been used to inseminate
eggs; or both errors have occurred. 154 ,155 It has been admit­
ted that these mix-ups are a regular occurrence in fertility
clinics in the UK;156 similar cases have arisen in other coun­
tries. 157 ,150 Th e majority of cases t hat have come to light
involve white parents having a black child, or vice versa.
This is the most readily apparent way of detecting any mis­
takes. How many other cases go undetected is open to ques­
tion. Paternity testing is one way that ca n bring the errors to
the fore . The ramifications of such events for the parties
involved cannot be underestim ated, encompassing psycho­
logical trauma, legal parenthood and custody issues as well
as compensation considerations. 155 ,158-1 60
Mutation and Anomalous Genotypes
While the decision in the Charlie Chaplin paternity case has
been cliticized, given what appeared to be an impossibility of
blood group inheritance, using DNA analysis, failure to
exclude an individual as a parent, even when alleles of the
alleged father do not appear t o fit those expected, do es have
scientific reasoning. The fact is that SIR loci, like any region
of the genome, are susceptible to mutation. 161 In fact, these
regions of the genome are particularly susceptible to muta­
tion and show a high rate of mutation during gamete forma­
tion, when the cells undergo meiosis, which can lead to allel ic
change from one generation to the next. 162-165 This can result
in an allele in the offspring being different from the parental
allele from which it was derived, and without further testing
might lead to a conclusion of non-paternity being
dra wn. 166- 169 Mutation rates vary across the SIR loci and
some may be as high as 0.3 per cent of meiotic events. 165
Evidence of two mutations occuning between generations in
several cases has led to a minimum of three independent
allel ic inconsistencies being required for a putative father
to be excluded as the biological father. 167 ,168 When there is
doubt whether inconsistent alleles are due to mutatio n or a
genuine excl usion of the alleged father, typing the Y chromo­
some SIRs of the child, if male, may provide further evidence
to show a genuine exclusion or not, as the case may beY6,1 70
The mutations discussed above relate to ge rmline events,
which are passed from parent to child and affect all the cells
of the offspring, though somatic mutations may also arise
in cells other than the parental gametes. If a mutation arises
at an STR immediat ely post fertilization and that STR is
subject to forensic ana lys is, then the resulting cells of the
offspling will have a profile that differs from that expected
on the basis of genetic inheritance. All the child 's cells will
carry the mutated allele and not the parental one. This has
consequences for pat ernity investigations.
It is not just in paternity investigations that mutations
can affect the outcome of t he investigation. Somatic cell
mutations in SIRs can lead to anomalous profiles with more
than two alleles being seen per locus 171, 172 and discrepan­
cies in results of comparison of profiles, even if the profiled
sa mples originated fro m t he same individual. If a mutation
a rises in a subset of cells that are profiled , and the mutated
cells are at a level above the threshold of detection, a com­
parison between pro files of cells of a different tiss ue, or a
sample of cells lacking a sufficient proporti on of mutated
cells for detection, can lead to the possibility of a common
source of the samples, and thus a match, being missed. This
may be of relevance if, for example, blood cells sampled at
a crime scene have a different profile from those of buccal
cells taken as a reference sample from a suspect.
Individuals who possess cells with different genotypes
that have arisen from a single zygote by way of mutation
are known as mosaics. 173, 174 An individual may also possess
cells of differen t genotypes because of ch imerism resulting
from natural anomalous events in which more t han one
zygote fuses to create a single individual; 175-1 78 when cells
derived from a different zygote are present in a single indi­
vidual (e.g. if dizygotic twi ns exchange blood across a pla­
centa l79 ,lool or as a result of medical intervention, be it bone
marrow or organ donation.181-184 In the case of bone mar­
row transplantation, for example, it is possible for false
identification to ar ise when the donor may, in fact, be the
guilty party, but because the recipient harbours cells shown
to match the donor's cells left at a crime scene the recipient
may be falsely accused or convicted. 185
It will be apparent that chimerism and mosaicism can lead
to inappropliate exclusion of the true source of a sample,
when looking for a match, and to the exclusion of a biologi­
cal parent as being so if the possibility of the phenomenon
is not taken into account during an investigation. 175 ,176, 186

The problem of the perception of the infallibility of DNA evi­
dence was highlighted in the media-grabbing case where a
mother was accused of committing benefit fraud because it
was alleged that the she could not have mothered the children
in question and risked them being taken in to care. It was only
after her third child's birth was witnessed, and DNA evidence
also showed that baby appeared not have been her child, that
it was discovered that she was a chimera. The cells giving rise
to her offspring were of different genotype to those typed
during the maternity investigation. 177
The Use of Mitochondrial DNA and the
Y Chromosome in Paternity Cases
As mentioned above, the Y chromosome may be used to pro­
vide additional information in cases where there may be
uncertainty over whether a mutation has arisen from father
to child. There may be occasions when paternity is in dispute
and samples of the putative parent are not available, which
causes practical problems. Sometimes, material may have
been taken for medical investigations and, if it has been
stored, it may be accessed and used as a source of DNA.166
Obviously, this is not always possible, and one approach that
might be adopted in the case of a male offspring is analysing
the Y chromosome haplotype of male relatives of the puta­
tive father, given that the Y chromosome is passed from
father to so n and, barring mutation, aU these family relatives
would share a common Y chromosome. This was put in to
practice in the attempt to address the long-standing histori­
cal controversy over allegations that the US President
Thomas Jefferson fathered the sons of one of his slaves,
Sally Hemmings. The first source that on e would look to in
order to type a deceased's Y chromosome would be known
male offspring of the alleged father, but Jefferson had no
surviving sons. Instead data were gathered from male-line
descendants of two sons of the president's paternal uncle,
male-line descendants of the two men Jefferson was alleged
to have fathered (Thomas Woodson and Eston Hemmings)
and male line descendants of the sons of Jefferson's sister,
also suspected of fathering one of Hemming's children.
Analysis of the Y chromosome haplotypes excluded Jeffer­
son of fathering Thomas Woodson, but the younger son,
Eston Hemmings, sha red the same Y chromosome haplotype
as Jefferson 's relatives, providing support that he may have
been Jefferson's illegitimate son. 187 However, the very fact
that Jefferso n's male relatives had a common Y haplotype
also means that any of Jefferson's male relatives alive and
capable of fatherhood at the time co uld not be ruled out as
the father and the president was only one of several candi­
dates, despite the incriminating publicity the results of th e
study attracted. 188 The lack of conclusive evidence has
meant the descendants of Jefferson's daughters have not
accepted the relationsh ip,1 89 highlighting the fact that DNA
evidence cannot conclusively identify an individual as being
the source of a sample or an individual as being a biological
Identification: body remains/missing persons I 411
parent. The best it can do is provide evidence in terms of sta­
tistical probability and it is up to a jury or those concerned
to determine if it fulfils their criteria of proof.
IDENTIFICATION OF BODY REMAINS
AND MISSING PERSONS
The loss of a child is devastating for any parent, and one can­
not overemphasize the importance of closure for families in
the event of such tragedies by way of positive identification
that their child has actually died. If the remains of a deceased
person can be identified, this can put an end to uncertainties
or hopes that a child may have sUlvived and can enable
return of the body or remains to the next of kin for appropri­
ate burial, events that may be of deep psychological and reli­
gious significance and comfort at a time of distress.
DNA analysis is one way that can assist in the identifi­
cation of human remains, but the approach is not without
problems. Not only does one have to address the state in
which the remains are found , which is often far from ideal
because of degradation or decomposition, but one needs to
acquire a reference profile with which to compare that of
the remains. There is also the issue that following some
incidents, for example, explosions, body parts may be frag­
mented and widely dispersed, adding to the difficulty of
reunify in g an identified whole body.
If the remains show few signs of decomposition, it will
generally be possible to obtain DNA from blood samples or
internal soft tissue which has been largely unaffected. As the
degree of decomposition increases, blood becomes less of an
option and superficial soft tissue may exhibit signs of putre­
faction, in which case tissue or muscle deep er within the
body and/or bone marrow may yield DNA suitable for STR
profiling. When remains are at an advanced state of decom­
position, bone marrow may still be able to provide material
for a DNA profile but, as decomposition progresses, the
extraction process increasingly faces problems of cell debris,
decomposition products and contaminants that can inhibit
the PCR reaction, and thus need to be removed. If bone mar­
row does not yield a profile then hair may be STR typed, if the
root rema ins, or mtDNA from the shaft may be sequenced.
Alternatively, DNA could be extracted from skeletal struc­
tures, which is also the option when remains are fully skele­
tonized. The resistance of structures such as bone and teeth
offers protection to DNA contai.ned in the bone matrix or
tooth pulp; however, this, in turn, requires special proce­
dures to extract and purify it. The level of nuclear DNA that
- can yield an STR profile from these sources is low; sequenc­
ing the more resistant mtDNA may be the only option avail­
able. On rare occasions one may be faced with naturally
mummified or desiccated remains, which present their own
challenges, particularly if there is a degree of decomposition.
Extracting DNA from soft tissue of such remains has given
little, or marginal, and inconsistent success in STR profiling
(Jason Eshl eman, personal communication; Carlos Morales,
412 I Forensic DNA profiling
personal communication), though extracting and typi ng
DNA from bones and teeth of such remains does not appear
to offer significant problems over other sources of bone and
teeth (Tim Clayton, personal communication; Jason Eshle­
man, personal commu nication) . DNA has been extracted and
sequenced from dried superfici al parts of Egyptian mummi­
fied remains, suggesti ng that it is possible to get genetic
information from such material, but the DNA in this case
was in a fra gmented state, making its suitability for STR
analysis questionable. It was only one of many samples
studied that indicated that recoverable DNA might be
obtained. 190 It is possible tha t natural and artificial mummi­
fication may affect the success of extraction.
Despite these issues and potential problems, DNA analy­
sis has been of Sign ificant value to investigating teams in
major incidents or mass disasters over the years. Although
techniques such as visual identifi cation, odontology, anthro­
pology, fingerprint analysis, radiology and facial reconstluc­
tio n can go a long way in helping to identify remains, they
may not always be appropriate or possible. DNA analysis has
succeeded in many investigatio ns, in conjunction with these
tech niques, or in cases in which other techniques have been
unsuccessful.
For example, in 1998 a tragic fire in Manila resulted in
the death of 23 children between the age of six months and
eight years. The bodies were initially buried unidentified,
but three mon ths later the burial site was exhumed in an
attempt to identify the remains. 191 Exhumation recovered
22 bodies, and a combination of autosomal and Y chromo­
so me typin g was carried out on 21 of the bodies that had
not been otherwise identified. The identification process
was assisted because th e identity of the children present at
the tragedy was known, though this may not always be the
case in such investigations. It was also possible to deter­
mine the approx imate age range of the re mains, which was
of further benefit. Ten of th e exhumed bodies were believed
to be sibling pairs and 11 were unrel ated to each oth er. 191
Despite the bodies hav ing been burnt, buried an d
exhumed, DNA analysis succeeded in matching 18 of the
remains with a known ch ild. In two of the cases mothers
were able to provide reference mate ria l in the form of stored
umbilical tissue from their child, which had been kept in
accordance with Philippine custom an d which provided a
positive identification . 192
Iden tificatio ns of the other children, for whom ante­
mortem reference sa mples were not ava ilable, was achieved
by patern ity-type ana lysis using autosomal STR profiling,
as well as by comparison ofY haplotypes, by analysing DNA
reference samples from 10 mothers, three fathers and a ­
paternal grandfather of two children for whom a paternal
sample was unava ilab le. Y haplotypes were valuable in iden­
tifying not only father/grandfather to son/grandson rela­
tionships but also those of sibling brothers. 191
In the Manila fire tragedy, the remains had been subject
to extreme conditions so it was no t su rprising that a full
profile was not obtained for evelY sample typed. Where
alleles cannot be identified at evelY 10CllS, th e evidential
weight is lower but va luable information can still be
obtained, as illustrated by this case.
Having an ante-mortem reference sample of a mlss1l1g
individual with which to compare DNA profiles of remains
goes a long way to aiding the process of identification, and
over the years such samp les have been taken from tooth­
brushes, hairbrushes, clothes, towels and leftover food with
bite marks.30.193 It is important to realize, however, that
sllch items are mobile and may have been used by someone
other than the owner and, hence, might cany DNA of
someone other than the individual in ques tion. Archived
medica l samples are a very useful source to turn to for ref­
erence material if they are available.
The paternity-type approach, using pedigree analys is to
identify remains when reference material is ava ilable from
the pa rents of a ch ild suspected to have di ed, is commonly
used. Knowing each parent's alleles, one can predict the
range of possible genotypes that their child could have and
eliminate remai ns that do not cOITespond to these. If a match
is identified, the li kel ihood that the remains are the biologi­
cal child of the couple rather than the child of an unrelated
co uple can be determined statistically. Admittedly there can
be a velY large number of possi ble combinations (for a 10­
loclls profile this cou ld be as high as 1048576), but the
approach has been used successfully in many mass disasters.
For example, after the Waco, Texas, incident of 1993 - when
the remains of 61 individua ls were recovered in varying
states of preservation, some in a n advanced sta te of putre­
faction and others badly charred - 26 positive identifications
were made in this way.194 It is also possible to gather suf­
ficient inform ation when a single paren t is available as a
reference sample, 195 though the statistical weight of the evi­
dence will be less, and comparison with a sibling's genotype
may also be helpful. In aeroplane traged ies, families are
often present together on the same flight and the common­
ality of alleles across the generation can be used in identify­
ing family members.
The 9/11 disaster of 2001, when 2792 people were killed
in a terrorist attack on the World Trade Center, must surely
have been one of the most challenging investigations, and
those w ith the task of recoveling and identifying the remains
faced lo gistical, technical, practical, scien tific and medical
difficulties. Although direct matches with referen ce samples
were used where possible to identify the remains, other DNA
techniques, including the paternity approach, were also
employed. 196
It will be appreciated that in paternity analysis there
is always the chance of a coincidental match. A particular
all ele at a locus is shared not only by a parent and child,
but also by approximately 1 in 1000 random pairs of indi­
viduals, so an attempt was made to minimize the number
of false-positive matches obtained during the identification
process. 197 The number of false-positive results is related to
the size of the reference list and the number of victims, and
therefore is a major consideration in investigations such as

the 9111 tragedy, but it should also not be overl ooked in
investigations invol ving smaller numbers of victims.
As men tioned previ ous ly, the state of the remains can
affect the amount of information that can be obta ined from
STR analysis. Poor-qu ality DNA may not provide results for
the alleles present at all loci typed, or it may resu lt in iden­
tifica tion of only a sin gle allele in a hete rozygous individual
when there really should be two alleles; one always needs to
be aware of the possibility of this so- ca lled allelic dropout. In
many cases, the reason for poor SIR resu lts is that the STR
fragment in the sa mple is no t intact along its whole length,
as it should be. As DNA degrades it fragments into smaller
pieces a nd breaks may occur within the SIR; thus, the true
alleles are not detected. It may be possible to design PCRs for
STR profiling that look at smaller sections of DNA, 198 and
commercial kits adopting this approac h are no w available ;
the sma ller t he targeted DNA being identified, the greater the
chance it remains intact. Where there are deficiencies in SIR
typing, mtDNA may be sequenced, as discussed ab ove, or the
in format ion gathered from partial STR profiles may be sup­
plemented by, or rep laced, with SNP data. 199 SNPs are single
base sequence variations between indiv iduals at a particular
positio n in the gen ome. These sites are ab undant (estimated
10 million) throughout the genome, but because they are
mostly biallelic they lack the polymorphic variation across
the population that SIRs exhibit. 2oo However, if one geno­
types a DNA sample at many SNPs, one can approach the
discriminatory power of SIR profiling. Estimates vary as to
how many are needed, but it is likely that between 50 and
100 may be required,199 consistent with the 70 used in rela­
tion to the 9/11 investigation. 196
IDENTIFICATION OF THE 'ABANDONED BABY'
OR FETAL MATERIAL AND AVENUES
FOR IDENTIFYING THE SOURCE
OF AN UNKNOWN PROFILE
The approac hes described above, which are used to identify
the remains of a ch ild , ha ve been geared towards identify­
ing a previously know n livi ng individual, either by directly
matc hing a profil e with a reference sam ple from the indi­
v idual or through pedigree analys is. However, one may be
faced with the scenario of an aba ndoned newborn or fetus
requirin g investigation. In such a scenario t here will be an
endeavour to iden tify the mother. Whethe r maternal DNA
can be recovered from the scene to ass ist will depend on
the post-partum circum stan ces and the conditi on in which
the newborn is found.
Ma ternal DNA has successfully been recovered and pro­
filed from placental material found abandoned with a
wrapped dead newborn, as well as from blood found on the
wrappin g material. Using the profile, it was possible to
determine the mothe r of the baby in this case. 83 Obviously,
this outcome requires that a reference sample is available
fro m the mother fo r comparison. For exam ple, entries on a
--. ~;
Identification of an unknown profile I 413
nationa l DNA database can be searc hed agai nst the profile
in question or from a woman suspected of being the
mo ther during the investigation.
Placental material is an ideal source of maternal DNA in
such circumstances, but it may not always be available. In
another repolied case, a live newborn was fou nd abandoned
in a box but no pl ace ntal materia l was found with the baby.
It was, however, possible to obtain a profile of the mother
by extracting maternal DNA from vernix caseosa on the ha ir
and body of the newborn, which conta ined maternal
blood. 83 One should be aware of the potential for such mate­
rial being relevant and useful in similar investigations.
A more problematic scenario is the case of the abandoned
baby fou nd long after deli vely. The rem ains may have
decomposed, with the concomitant difficulties of obtaining
DNA profiles discussed previ ously, or there may be circum­
stances in which the deceased has been concealed or the
environment has been conduci ve to mummification. As men­
tioned earlier, DNA profiling of desicca ted material is prob­
lematic. [n the case of a mummified baby, there is the
addition al issue that bone and teeth, a commonly used source
of DNA in the case of remains, will be unavailable if the
neonate's bone is cartilagi nous and teeth have not erup ted.
Despite the difficulty in obtaining suitable genetic informa­
tion from desiccated tissue, profiles have been obtain ed from
mummified tissue from babies, though in these cases there
was no accompa nying decomposition of the bodies, which
may have contri buted to the successful results.
[f no maternal DNA is found with the fet al remains but a
DNA profile of the baby can be obtained, genetic analysis
resorts to the pedigree type of analysis: comparing the baby's
profile with those of suspected parents in a paternity testing
approach to look for consistent transmission of alleles. As
stated above, the statistical weight of such an analysis is less
than if information is available from at least one parent. 195
The problem is obtaining DNA from a suspected mother
when there may be no lead to the parents. However, this is
true of any investigation centring on DNA. An SIR profile
alone provides no meaningful information other tha n the
sex of the individual. What fo llows desc ribes possible means
of identifying a candid ate so urce when there is no match
with existing profiles on national DNA datab ases or with
suspects in the case, these being the first two options to
consider.
Typ ing the less specifi c mtDNA or Y chromosome might
be of assistance in spread ing the net wider: identifying a
match might go so me way in identifying a possible pedigree
that may lead to a parent, but this type of investigation may
also lead to nothing. A similar technique, so-called 'familial
searching', is also increasin gly being used. In this case, rather
than searching existing databases for a profile that matches
all alleles at all loci, a less stringen t search is employed, allow­
in g profiles with some allelic mismatches to come to light.
The logic beh ind this approach is that such profiles might
belong to relatives of the true source, who would be expected
to share more all eles in common than a random individual,
414 I Foren sic DNA profiling
and it may be possible to home in on the true source of the
sample through a family member. The techn ique has been
used successfully20' but raises many ethical issues and
remains highly controversiaI.202-204 As with many genetic­
based investigations, there is the potential to uncover other­
wise unknown non-blood relatio nshi ps in individuals
believed to be true relatives, with ramifications for family
relationships that shoul d not be considered lightly.
A number of studies have pointed to the potential to
obtain an indication about the physical charactelistics of a
source of DNA in a forens ic investigation in which no
matching reference sample is found, based on the fact that
the physically recognizabl e characteristics of an individual
are, at least in part, genetically determined.14.2os-207 Acquir­
ing information about the race of the source of the sample
has also been proposed, based on inter-racial genetic differ­
ences. However, moving from looking at physically incon­
sequential as pects of our genetics in the form of rou tinely
analysed STRs, currently believed to have no influence on
our characteristics or health, to more personal and sensitive
information is something that wiJi have social implications
that need careful consideration, particularly when such
information is unlikely to be conclusive and may be of no
investigative va lue, given the com plex play of genetic and
environmental facto rs t hat make us who we are.20B-2IQ
In the specific case of the abandoned baby or fetus, one
may come across rare cases in wh ich there are indications in
the deceased of a genetically inherited condition, for exam­
ple the charactelistic features of Down's syndrome. 173 In
such cases prenatal genetic testing of the fetus may have
been carri ed out, and it may be possible to co mpare the pro­
file of the baby with stored clinical samples. A match wi th
a clinical sample could lead to the parents through medical
records. With increased use of electronic medical records,
th is approach could become a more realistic prospect in the
future. Admittedly these are tenuous possibilities, requiring
significant investigative power and resources, and one has
to accept that DNA evidence may not provide constructive
evidence in an investigation.
There may be times when an investigator wants to know
the age of the source of a sample, for example if a newborn
is illegally removed from hospital, or to determine if fet al
blood is present in putative products of conception in crim­
inal abortion investigations. Physical chara cteristics can be
of some help in some cases, but inte resting research has
identified a method that can differentiate between blood
fro m a newborn less than a day old , blood from an infant
less than four months old and blood from infants over four
months of age by assaying the expression levels of partic- ­
ular haemoglob in genes, which exhibit a highly age­
specific expression pattern .211 Implementation of this
ap pro ach in forensic investigations may provide va luable
inform ation for future investigations. Rath er than looking
at the DNA itself, one is looki ng at the translated product
of DNA, ribonucleic aci d (RNA), which acts as a intermedi­
ary in translating genes into an end product.
DNA DATABASES
In this chapter, mention has frequent ly been made of DNA
databases. Currently, national DNA databases are the norm,
com ing under national legislation and management, wi th the
nations' authorities having varying powers to collect and
analyse samples from individuals depending on the severity
of the suspected crime. In the UK, authorities have the power
to take and analyse a DNA sample from anyone suspected of
having committed a recordable offence. These samples are
entered on th e UK National DNA Database, the largest data­
base in the worl d, where they are cross-checked against exist­
ing entries on the database to look for a match. Profiles
may be from individuals convicted of a crime, fro m previous
arrestees or from samples taken from unsolved crime scenes;
matches connecting crimes with named known individuals or
linking crimes with other crimes can be made. Under English
law samples taken from a suspect remain on the UK National
DNA Database indefinitely, regardless of whethe r the suspect
is co nvic ted, acquitted or never brought to trial. Authorities
in other countries or states have lesser powers to take or
retain samples, and there is significant valiation throughout
the world. The question is whether full harmonization of laws
and intercountry procedural standardization can ever be
established. The Combined DNA Index System (CODIS) oper­
ates in the USA, and it enables laboratories to exchange and
compare DNA profiles at a state and national level. The exis­
tence of an Interpol DNA database and DNA Gateway enables
the sharing of data across nations, so there are moves
towards global cooperation. One has to wait to see whether a
truly international database is esta blished containing every­
one's profile with the a im of assisting in solving Clime at an
internationa l level. This is something aspired to by some, but,
unsurprisingly, a view not shared by many, who feel that
such a move constitutes a serious violation of civil liberties.
DNA has had a major impact on the field of forensic sci­
ence and investigations. Without DNA we would not be here
as living beings; without DNA a nalysis, undeniably, many
investigations would have gone unsolved, but we should be
aware that the techniques have their limits. DNA provides
only one piece of evidence in any case, and any resul t and
statistical likelihood of a chance match needs to be cons id­
ered in the context of the case as a whole.
REFERENCES
Albe rts B, Johnson A, Lewis J et al. Development of
multicellular organisms. In Molecular Biology of the Celi,
4th edn. New York: Garland Scien ce Taylor and Francis
Group, 2002.
2 Jobling MA, Gill P. Encoded evide nce: DNA in forensic
analysis. Nat Rev Gen 2004; 5:739-51.
3 Hartl DL, Jones EW. The chromosomal basis of heredity. In
Essential Genetics: a Genomics Perspective, 3rd edn .
Sudbury, MA : Jones and Bal11 ett Publishers, 2002.
4 Alberts B, Johnson A, Lewis J et al. Germ cell s and
fertilization. In Molecular Biology of the Cell, 4th edn.

New York: Garland Science Taylor and Francis Group,
29 Spalding RP. Id entificati on and characterization of blood
2002 .
and bloodstains. In Jam es SH, Nordby JJ (ed s) Forensic
5
Rafferty S. Fingerprint evidence doubts clear policewoman
of perjury. Scotsman, Saturday 15 May 1999, p. 4.
6 McKie IAJ. There's nane ever fear'd that the truth shoul d be
heard but they whom the truth would indite. Sci Ju sriu
2003; 43: 161-5.
7 Cole SA. More than zero: accounting for error in latent
fingerprint identificati on . J Crim Law Criminol 2005;
95:985 -107 8.
8 Webb J. The myth of fingerprints. New Scienrisl 2005;
Proceed in gs from the Second European Symposium of
2517:3 .
Hum an Identifi catio n, 1988, pp. 131-7.
9
Saks MJ, Koehler JJ. The coming paradigm sh ift in forensic
id ent ification sc ien ce. Sciellce 2005; 309:892- 5.
10 Broeders APA. Of earprints, fingerprints, scent dogs, cot
deat hs and cog nitive contaminati on - a brief look at the
present state of play in the forensic arena. Forensic Sci 1m
2006; 159: 148-57.
II Dror IE, Charlton D, peron AE. Contextual information
renders experts vulnerable to making erroneous
id entificati ons. Forensic Sci Inr 2006; 156:7 4-8 .
12 Jain AK, Prabhakar S, Pankanti S. On the similarity of
id entical twin fingerprints. Pattem Recognition 2002;
35:2653-63.
13 Opdal SH, Rognum MD. The sudden infant death syndrome
gen e: does it ex ist? Pediatrics 2004; 114:506 - 12.
14 Duffy DL. Montgomery GW, Chen Wet al. A three-single­
nucleotide polymorphism ha plo type in intron I o f OCA2
explains mo st human eye-co lor variation. Am J Hum Gener
2007; 80:241-52.
15 SudbelY P. Single-gene disorders. In Huma n Molecular
Generics, 2nd edn. Harl ow, UK: Pearson Edu cation , 2002.
16 Kum ar V, Abb a AK, Fausto N. Robbin (lnd Corran's
Pathologic Basis of Disease. Philadelphia: Elsev ier Saunders,
2005, pp. 489-95.
17 Red on R, Ishik awa S, Fitch KR et a!. Global variation in co py
number in the human genom e. Narure 2006; 444:444-54.
18 Shianna KV, Willard HF. In sea rch of no rmality. Naru re
200 6; 444:4 28 - 9.
19 Li W-H, Saunders MA. The chimp anzee and us. Nature 2005;
437:50-1.
20 The Chimpanzee Seq uen cing and Analysis Consortium.
Initial se quence of the chimpanzee gen ome and compariso n
wi th the hum an genome. Naru re 200 5; 437:69-87.
21 Newton G. Discovering DNA fingerprint ing. Wellcol11e Trusr:
The Human Geno me Web sire. 2004 (accessed April 2007).
http://geno me.wellcome.ac,uk/d oc_wtd020877.h tm I
22 Jeffreys AJ, Wilson V, Th ein SL. Hypervariabl e 'millisatellite'
region s in humon DNA Narure 1985; 314:67- 73.
23 Rudin N, Inman K. Forensic DNA Allalysis. Boca Raton,
USA: CRC Press, 2002, pp. 71-6.
24 Forensic Science Service Casefi les: Colin Pitch fork - first
murd er co nvicti on on DNA ev idence also clears the prim e
suspect. Th e Forensic Science SeflJice Website. www.forensic.
gov.ukl fo rensic_ t/i nsidel news/! isCcasefi les.p hp?case = I
(accessed April 2007).
25 Butler JM. Overview and histoJY of DNA typing. In Forensic
DNA Typillg, 2nd ed n. Amsterdam: Elsev ier, 2005, p. 3.
26 Aronson JD. DNA fingerprinting on trial : the dramatic early
history of a new forensic technique. Elldeauour 200 5;
29 : 126-31.
27 McKonkey E. Lecture 3 - genet ic diseases: the conseq uences
of mutat ion. In How the Human Genom e Work s. SudbuIY,
1\rJA: Jones and Bartlett Publishers, 2004, p. 43.
28 Butler JM . Appe ndi x II: U.S. population data - STR allele
frequencies. In Forensic DNA TVpiHg, 2nd edn. Amsterdam:
Elsevi er, 2005.
Referen ces I 415
Science: all In troduction ro Sciellrijic Investigative
Techniques. Boca Raton, FL: CRC Press, 2003.
30 Butler JlVl. Sam ple collection, DNA extracti on an d DN A
quantita tion. In Forensic DNA Typin g, 2nd edn. Amsterdam:
El sev ier, 2005.
31 Van Hoofstat DEO, Deforce DLD, Brochez V et al. DN A
typing of fing erp rints and skin debris: sensitivity of capillary
electrophoresis in forensic applicati ons using multiplex PCR.
32 Findlay I. Tayl or A, Quirke P et al. DNA fingerprinting from
single cells. Nature 1997; 389: 555-6.
33 Vall Oorschot RAH, Jones MK. DNA fing erprin ts from
fing erprints. Na ture 1997 ; 38 7:767.
34 Gill P, Whitaker J, Flaxman C et al. An investigation of the
rigor of interpretation nIles for STRs derived from less than
100 pg of DNA Forensic Sci lilt 2000; 112:17- 40.
35 Wickenheiser RA Trace DNA: a review, discussion of
theory, an d application of the transfer of trace quantities
of DNA through skin contact. J Foren sic Sci 2002;
47:442-50.
36 Hartl DL, Jones EW The chemical structu re, repli cation, and
manipulation of DNA Essential Gen erics: A Gellomics
Perspectiue, 3rd edn. SudbulY, lVlA: Jon es and Baltlett
Pu blishers, 2002.
37 Butler JM. The polymerase chain rea ction (DNA
amplification). In Fore nsic DNA Typin g, 2nd edn.
Am sterdam: Elsevier, 2005.
38 Butler JM. DNA separat ion methods : slab ge l and capill ary
electroph oresis. In Forensic DNA Tvping, 2nd edn.
Amsterd am: Elsevi er, 2005.
39 Thangaraj K, Reddy AG, Singh L. Is the amelogenin gene
reliab le for ge nd er identification in forensic casework and
prenatal diagnosis? Inr J Leg !VIed 2002; 116:121- 3.
40 Chang YI\rJ, Burgoyne LA, Both K. Hi gher failures of
amelogeni n sex test in an Indi an population group.
J Forellsic Sci 2003: 48:1309-1 3.
41 Mitchell RJ, Kreskas M, Baxter E et al. Amelogen in Y
negative males: multiple origins. Illr Congress Ser 2006;
1288 :274-6.
42 Jobling lVlA, Lo IC, Turner DJ et al. Structura l variation on
the short arm of the hum an Y chrom osome: recurrent
multigene deletions encompassing amelogenin Y. Hum Mol
Gen 2007; 16:307-16.
43 Sudbe ry P. Hu man population geneti cs and evolution. In
Humal/ Molecular Generics, 2nd ed n. Amsterdam: Pearson
Ed ucation, 200 2.
44 Butler JM. STR popul ation database analysis. In Forensic
DNA Typing, 2nd edn. Amsterdam: Elsevier, 2005.
45 Hartl DL, Jones EW Molecular mechanisms of gene
regulation. In Essential Gel/etics: A Genomics Perspective,
3rd edn. Sudbury, MA: Jones and Bartlett Publishers,
2002.
46 Winter PC, Hickey Gl, Fletcher HL. Section D popul at ion
genetics and evolution. In Ham es BD (ed.) Genetics: Insrant
Notes Series, 2nd edn. Ox ford, UK: BIOS Scien tifi c Publishers
Ltd, 2002.
47 Butler JfVl. Profile frequency estimates, likelihood ratios, and
so urce ottributioll. In Forensic DNA Typing, 2nd edn.
Amste rdam : Elsevier, 2005 .
48 Foreman LA, Evett IW Statistical analysi s to support
fore nsic interpreta ti on for a new ten-locus STR profiling
syste m. Inr J Leg Med 200 1; 114 : 147-55.
49 Jha A DNA fingerprinting 'no longe r foolproo f. Guardian,
Thursday 9 September 2004.
416 I Forensic DNA profiling
50 Lu cy D. DNA I. In Introduction to Sra listics for Forensic
Scientists. Chichester: John Wiley and Sons, 2005.
51 Lucy D. Eva luation of evidence in practice. In Introduction fO
Statistics fo r Forens ic Sci entists. Chichester: John Wiley and
So ns, 2005.
52 Evert IW. Towards a uniform fram ework fo r reporting
opini ons in forensic science casework. Sci Justice 1998;
38: 198-202.
53 Forensic Science Service, Association of Police Surgeo ns and
Association of Chief Police. Officers' Guidelines for the
Collection of Specimens. London: Forensic Science Service,
2000.
54 Inman K, Rudin N. Good field practice - processing a crime
scene. In Principles and Practice of Criminalistics: the
Profession of Forensic Science. Boca Ra ton, FL: CRC Press
LLC, 2001.
55 National Criminal Justice .Reference Servi ce. DNA evidence:
what law en forcement officers should know. NCJ 200908.
Nat In st Justice J 2003; 249: 10- 15. also Available online:
www.ncjrs.gov/pdffilesl/jr000249c.pdf (accessed April 2007).
56 Kaye DH, Sensabaugh Jr GF. Reference guide on DNA
evidence. In Reference Manual on Scienfific Evidence, 2nd edn.
Washington, DC: Federal Judicial Center, 2000, pp. 485-57 6.
57 Fa igman DL, Kaye DH , Saks MJ , Sanders J. Modem
Sc ientific Evidence: Th e Law and Science of Expert
Testimony. 25 -2.5 DNA Typing; The scientific statu s af DNA
typing: Laboratory performance. St Paul , MN: West
Publishing Co., 2002, pp. 257-63.
58 Lynch M. The discursiv e production of uncertainty: the
OJ Simpso n 'Drea m Team ' and the sociology of knowledge
machin e, Soc Stud Sci 1998; 28 :829-68.
59 Scientifi c Testimo ny, an onlin e journal: DNA tes ting
problems. Avail abl e online: www.scientific.org/
DNAProblems/DNA-Problems.htm (acc essed April 2007) .
60 Toledano T, Quarino L, Leung S et al. An assessment of DNA
contamination risks in New York City medical ex amin er
facilities. J Forensic Sci 1997 ; 42:721-4.
61 Rutty GN. Human DNA contamination of mortuaries: does it
matter? J Palh al 2000; 190:410-1 I.
62 Rutty GN, Watson S, Davison 1. DNA con tamin atio n of
mOltuary instruments and work surfaces: a significant problem
in for ensic practice? In l J Leg Med 2000; 114:56-60.
63 Lowe A, Murray C, Whitaker J et at. The propensity of
individuals to deposit DNA and second ary transfer of low-level
DNA from individuals to inert surfaces. Forensic Sci lilt 2002;
129:25- 34.
64 Rutty GN. An investi gati on into the transference and
survi vabili ty of human DNA follo win g s imul ated manual
strangulation with consideration of the problem of third
party con taminat ion. lnt J Leg Med 2002; 1J 6:1 70-3.
65 Sullivan K, Johnson P, Rowlands D, Allen H. New
devel op ments and challenges in the use of the UK DNA
database: addressing the issue of contaminated consumables.
Forensic Sci Int 2004; 146:5175-6.
66 Peel C, Gill P. Attribution of DNA profiles to body fluid
stains. lnl Congress Ser 2004; 1261 :53-5.
67 Bond JW. Value of DNA evid ence in detectin g crime.
J Forensic Sci 2007; 52:128-36.
68 Gill P, Kirkham A. Development of a simulation mod el to
ass ess the im pa ct of co ntamination in casework using STRs.
J Forensic Sci 2004; 49:485-91.
69 Rutty GN, Hopwood A. Tucker V The effectiveness of protective
clothing in the reduction of potential DNA contaminati on of
the sce ne of Clime. ln t J Leg Med 2003; II 7: 170-4.
70 Butler JM. Appendix N: DNA Advisory Board Quality
Assuran ce Standards. In Forensic DNA Typing, 2nd edn.
Ams(erdam: Elsevier. 2005.
71 Thom so n WC, Ford S, Doom T et aJ. Evaluating foren sic
DNA evidence, Part 2: Essential elements of a co mpe tent
defense review. Champ ion Magazi ne, 24-28 May 2003,
www.nacdl.org/_852 570 IC00653 9F8.nsf/0/9 C5F 7C6D724A4
7518 52 56E540074C156?Open (accessed April 2007).
72 Teichroeb R. .Rare look inside state crime labs reveals
recurring DNA test problems. Seattle PosHntelligencer,
Thursday 22 July 2004, http://seattlepi.n wsource.com/
local/183007 _crimel ab22 .html (accessed April 2007).
73 Pompa non F, Bonin A, Bellemain E, Taberlet P. Gen otyp ing
errors: causes, consequences and so lution s. Na t ReIJ Genet
200 5; 6:847- 59.
74 Faigman DL, Kaye DH, Saks MJ, Sand ers J. Modem
Scientific E,Jidence: The Law and Science of Expert
Testimony. 25-1.2 DNA Typing; Th e legal rele/Jance af DNA
tests: Objections ro DNA eIJidence. St Paul , MN: West
Pu blishi ng Co., 2002 , pp. 210-36.
75 Stevens -Simon C, Nelligan D, Breese P et al. The prevalence
of geni tal human papillomavims infections in abused and
non ab used preadolescent girls. Pediatrics 2000 ; 106:645-9.
76 Thompson C The GUiltl/Paediarric il/ le /joc e. Presentation at
22nd lusn Europe Meeting. Versaill es, Abstract 0.048, 2006.
77 Kl aus ner JD. The NAAT is out of th e bag. Clin Infect Dis
2004 ; 38:820-1.
78 Hammerschlag MR. Use of nucleic acid amplification test in
inves tigating child sexual ab use. Se.r Tran sll1 Infect 2001;
77: 153-7.
79 Hammerschlag MR. Appropriate use of nonculture tests for
the detection of sexually transmitted disea ses in children
and adolescents. Semin Pediatr Infect Dis 2003; 14:54-9.
80 Kellogg ND, Baill argeon J, Lukefahr JL et al. Comparison
of nucleic acid amplification tests a nd culture techniques
in the detect ion of Neisseria gonorriloea e and Chlamydia
trachoma tis in victims of suspected child sexual abuse.
J Pediatr Ado/esc Gynecol2004; 17 :33 1-9.
81 Tho mas A, Forster G, Robinson A, .Rogstad K. National
Guidelines on th e Management of Suspected Sexually
Transmitted Infectio ns in Children and Young people.
Clinical Effectiven ess Group (Assoc iation of Genitourinary
Medicine and th e Medical Society for the Study of Ven ereal
Diseases), 2002. Availabl e on lin e:
www.bashh.org/guidelin es/2002/
ado lescencfi nal_0903.pdf (accessed April 2007).
82 2001 Nati onal Guidelines on the Management of Adult
Victims of Sexual Assault. Clinical Effectiveness Group
(Ass ociation of Genitourinary Medicine and (he Medica l
Society for the Study of Venereal Diseases). Avail able online:
www.bas hh .o rg/gu idelines/2002/sexassauIC0601.pd f
(accessed April 2007).
83 Csete K, Beer Z, Varga T. Prenatal and newborn paternity
testing with DNA analysis. Forensic Sci Int 2005 ;
147:S57-60.
84 Robino C, Barilaro MR, Gino Set al. Inces tuous paternity
detected by STR-typing of chorionic villi iso lated from
archival formalin-fixed paraffin-embedd ed abortion
mate rial using laser microdissection. J Forensic Sci 2006;
51 :90-2.
85 Christian CW, Lavelle JM, De Jong AR et a!. Forensic
evidence findin gs in prepubertal victims of sexua l assault.
Pediatrics 2000; 106: 100-4.
86 American Academy of Pediatrics (Committee on Child Abuse
and Neglect). Guidelines for the eva lu ation of sex ual ab use
of children: subj ect revi ew. Pediatrics 1999; 103:186-91.
87 Allard JE. The collecti on of data from findings in cases
of sexual assau lt and the sig nificance of spermatozoa
on va gina l, an al and oral swabs. Sci Ju sti ce 1997;
37:99 - 108.
---
418 I Forensic DNA profiling
131 Tully LA, Parso ns TJ, Steighner RJ et aJ. A sensitive denaturing
gradi ent-gel electro phoresis assay reveals a high frequ ency of
heteroplasmy in hyp erva liabl e region I of the human mtDNA
co ntrol region. Am j Hum Gel1et 2000; 67:432-43.
132 Paneto GG, Martins JA, Lon go LVG et a!. Heteroplasmy in
hair: differences am ong hair and bl oo d From the same
individuals are still a matter of debate. Forel1s ic Sci lilt 2007;
173: lJ 7-21.
133 Gill P, Ivanov PL, Kimpton C et aJ. Id en tifi cation of th e
remains of the Romanov family by DNA analysis. Nat Gel1et
1994; 6:130- 5.
134 Ivan ov PL, Wa dham s MJ, Roby RK et al. Mitochondrial DNA
sequence hetero plasmy in the Grand Duke of Russia Georgij
Romanov establishes the authenticity of the remains of Tsar
Nich olas II. Nat Gellet 1996; 12:417-20.
135 C6li ca MY, Rodriguez Card ozo MB , Abovi ch MA et aJ.
Paterni ty cases when the alleged father is missing.
Int Co ngress Ser 2004; 1261 :520- 2.
13 6 Mayor LR, Bald ing DJ. Discrim in ation of halF-sib lings when
maternal genotypes are known. Forel1sic Sci lilt 2006;
159: 141-7.
137 Tra cey M. Sho l1 tandem repeat-based identification of
indi vid ual s an d parents. Croatian Med j 2001; 42 :23 3-8.
138 Linacre A, Goo dwin W. Role of and techni ques in fo rensic
hem atogentics - paterni ty testing. In Payne-James J, Busutti l
A, Smock W (eds) Fo rensic Medicille: Clinical al1d
Pathological Aspects. London: Greenwich Med ica l Media Ltd,
2003, pp. 677-88.
139 Fun g WK, Hu Y-Q , Ch un g Y-K. On statistica l analysis of
forens ic DNA: th eory, methods and co mputer programs.
Forel1s ic Sci lilt 2006 ; 16 2:17-23.
140 Coleman H, Swenso n E. DNA ill the Courtroom: a Tmil
Watcher's Gu ide. Washin gto n, DC: GeneLex Press Seattle,
2000. Avai lab le on line: www.genelex.com/
pate rnitytesting/pate rnityboo k.ht ml (accessed Ap ri l 2007).
141 Relationship Testin g Program Unit 2004. Annual Report
Summary for Testi ng in 2004. Ava il abl e online:
www.aabb. 0 rglD ocu men ts/ Accred ita tio n/Pa ren tage_Testin g_
Accreditation]rogram/rtannrpt04.pdf (accessed April 2007).
14 2 vo n Wurmb- Schwa rk N, M<ilyusz V, Simeoni E et aJ. Possible
pitfalls in motherless paternity ana lysis with related pu ta tive
fa thers. Forensic Sci Int 2006; 159:92-7.
143 Beren t J, Jacewicz R, Szram S. Effect on paterni ty index of
substituting the biological fath er by hi s brother ­
preliminary repol1. Int COl1gl'ess SrI' 2004; 1261 :476-8.
144 Hartl DL, Jones EW. Population ge neti cs and evo lution. In
Essential Genetics: a Genomics Perspecrive, 3rd edn.
SudbUly, MA: Jones and Bartlett Publishers, 2002 .
145 McKo nkey E. Lecture 3 - Gen etic diseases : th e co nsequ ences
of mutation. In How fiJ e Human Genome Works. Sud bury,
MA: Jones and Bartlett Pu blishers, 2004, pp. 49-50.
146 Sasaki H, lshino F. Gmomic Imprintil1g. Basel: Karger, 2006;
11J[ 1-4).
147 Naito E, Dewa K, Fukuda M et al. Novel paternity test ing
by distin gui shing parema l alleles at a VNTR locus in the
differentially methylated region upstrea m of tile hum an
H 19 gene. ] Forensic Sci 2003; 48: 1275-9.
148 Zhao G, Yang Q, Huan g D et al. Study on the app li cation of
parent-of-origin specific DNA methylation markers to
forensic geneti cs. Forensic Sci In t 2005; 154: 122-7.
149 Fraga MF, Ballesta r E, Paz MF et al. Ep igenetic differe nces
arise durin g th e lifetime of monozygotic twins. Proc Natl
Acad Sci 2005; 102:1 0604-9.
150 Co rach D, Pen ac ino G, Marino Met al. Incestuo us offspring
detection inference by VNTR homozygosity in crement. lilt
Co ngress Sel' 2003; 12 39:861-4.
151 Ottenberg R. Medicoleg al application of human bl oo d
grouping.] Am Med Assoc 192 1; 77:682-3.
\ 52 Shaw MW. Paterni ty determination 1921 to 1983 and
bey ond . jAm Med Assoc 1983; 250:2 53 6-7.
153 Albrecht K, Sch ulth eiss D. Proof of paternity: historical
refl ect ions on an andrological-forensic ch al lenge.
Andrologia 2004 ; 36:3 1-7.
154 BBe. Embryo mix-up at IVF hospital. Monday 28 October
2002. http://ne ws. bbc. co.u k/ l/hi/h ea lth/2 36770 5.st m
(accessed Ap ril 2007).
155 BBe. $1111 pay-outJol' IVF embryo mix-up. Thursday 5
August 2004. Available onl ine: ht tp://news.bbc.co.uk/
l/hi/he alth /353 7950.stm (accessed Apri l 2007).
156 BBe. NF mix-ups 'occur regularly'. Wedn esday 24 July
2002. Ava il ab le online:
http://news.bbc.co.uk/ l/hi/health/21484 23 .
still (accessed April 2007).
157 Dy er O. Black twins are born to white parents after infertility
treat ment. BM] 2002; 325:64.
158 Spriggs M. lVF mixu p: white couple have bla ck ba bi es, ]
Med Ethics 2003; 29:65.
159 BBe. ORA: Legal minefi eld of IVF 'm istake '. Monday 8 July
2002 . Avai lable on line: http://news.bbc.co.uk/ I/hi /health/
2116589.stl11 (accessed April 2007).
J 60 BBe. Court rules on fath er of IVF mix- up twins. Wedn esday
26 Febru ary 2003. Avail ab le onlin e:
http:// news.bbc.co. uk/ I/hi/
hea lth /2797J69.stlll (a ccessed Apri l 2007 ).
161 McKonkey E. Lectu re 2 - the mechanisms of mutation: how
the hum an ge nome cha nges. In How tiJe Hum an Genome
Works. Sud bury, MA: Jones and Bartl ett Publishers, 2004,
pp. 19-33.
162 Brin km ann B, KJintscher M, Neuhuber F et al. Mutati on rate
in human microsatellites: influence of the structure and
length of the tand em repea t. Am j Hum Gel1 1998;
62:1408-15.
163 Saja ntil a A. Lukka M, Syvanen A- e. Experimentally
observed germ line mutatio ns at hum an micro- and
minisateJlite Jo cl. Eur] Hum Genet 1999 ; 7:2 63 -6.
164 Kayser M, Roewe r L, Hed man M et al. Characteristic and
frequency of germline mutations at microsatellite lo ci from
the human Y chromosome, as revealed by direct observation
in father/son pairs . Alii j Hu m Genet 2000; 66:1580- 8.
165 STR DNA Interne t Data base Mutation Rates for Common
Loci. Ava ilabl e onlin e: www.cstl.n ist.gov/biotec h/strbase/
mutati on.htm (accessed Apri l 2007).
166 Yamamoto T, Tam aki K, Hu ang XL et al. The application of
minisatellite variant repea t mapping of PCR (iVIVR-PCR) in a
paternity case showin g fals e exclu sion du e to STR mutation.
] Forel1sic Sci 2001; 46:374-8.
167 Brandt- Casad eva ll C, Gene M, Piqu e E et al. Presence of two
mutations between fathe r and child in two cases of paternity
tes ting. 1111 Congress Ser 2003; 12 39:S 15-1 8.
168 Nutini Ai, Mariottini A, Giunti L et al. Double
incompa tibili ty at human alpha fibrinogen and penta E loc i
in pa te rnity testing. Croatian iVIed] 2003; 44:34 2- 6.
169 Thangaraj K, Reddy AG, Si ngh L. Mu ta tion in th e STR locus
D2 1S 11 of father caus ing all ele mi smatch in the child.
] Forel1s ic Sci 2004; 49:99-103.
170 Jun ge A, Brinkmann B, Fimmers R, Madea B. Mutations or
exclusion: an unusu al case in paterni ty testing. Int] Leg
Med 2006; 120:360-3.
171 Crouse CA, Rogers S, Ami ott E et a1. Analysis and
interpretation of sh ort tand em rep eat microvariants and
three-ba nded allele patterns usin g multiple allele detection
systems. j Fore nsic Sc i 19 99 ; 44 :87 -94.

172 Clayto n TM , Guest JL, Urq uhart AJ, Gill PD. A ge netic bas is
for ano malous band patterns enco untered during DNA STR
profiling. J Forensic Sci 2004; 49: 1207-14.
173 Sudbery P. Hu man genetic disease. In Hum lln Molecu lllr
Genetics, 2nd edn. Harlo w, UK: Pea rso n Ed ucat ion. 2002 .
174 Dauber EM, Wenda S, Glock Bet aJ. Mosa icism as a poss ibl e
reason for poor amp lification of ame logenin-Y in three
hum an male indiviciua ls. fl1l Congress SCI' 2004:
1261 :508-10.
175 Farber CM. Georges M, De Bock G et aJ. Dem ons tration of
spo ntaneous XX/XY chimeri sm by DN A fin ge rprinting.
Hum Genet 1989; 82:19 7-8.
176 Yu N, Krus kall MS, Yun is JJ et al. Disp uted maternity
leading to identificatio n of tetra ga metic chimerism. N Eilgi J
Med 2002; 346:1545.
177 Mayne E, Borland S. The moth er with three chi ldren who
don' t share her DNA. Ma il all SUlidav. 5 March 2006.
17 8 Mild e A, KJ1I1-Burm eiste r R. Ritz-Timme S. Ka atsch HJ. DN A
typin g in cases of blood chimeri sm. flit J Leg Med 1999;
11 2:333 - 5.
179 Rub ock i RJ, McCue BJ, Kelly MT et aJ. Natural DNA
mixtu res ge nerated in fraternal twin s in utero. J ForellSic Sci
200 1; 46: 120-5.
180 Dau ber EM, Fat' i, Stad lbache r S et al. STR typ in g in a pair
of chimeric tw ins. In( Congress Ser 2003: 1239 :569 - 71.
181 Dau ber EM, Dorner G, Mitte rbau er Met al. Discrepant results
of samples taken from diffe rent ti ss ues of a singl e ind ividual.
jilt COllgress Ser 2004; 1261 :48- 9.
182 Dauber EM, Muller CJ. Schii niger-H ekele M et al. Artificial
blood chimerism and grafl-versus-host disease after liver
transpl anta tion. ll1 i Congress Su 2006; 1288:840- 2.
183 Pop e S, Chap man I-I, Lam bert J. The effec t of bon e marrow
transp lants on DN A profi les : a case exam ple. Sci Justice
2007; 46:23 1-7.
184 von Wurl1lb-Schwark N. Bos inski H. Ritz-Timme S. What do
the X and Y chromosomes tell us about sex and gende r in
foren sic case analysis? J ForeJIsic Legal Mer! 2007:
14 :27-30.
185 Aldhous P. Bone marrow donors risk DNA id enti ty mix-up.
New Scientist 2005; 2523: 11.
186 Rodri guez Cardozo MB , Co lica MV, Abov ich fVIA et aJ.
Maternity testing in a ch imerical child . jJIt COll(Jress SCI'
2004; 126 1 :502-4.
18 7 Foster EA. Job ling MA. Tay lor PG et aJ. Jefferson fathered
slave's child. Nature 1998; 39 6:27-8.
188 Marshall E. Wh ich Jefferso n was th e father? ScieJIce 1999;
283: 153- 5.
189 Check E. Jefferso n's desce nd ants continue to deny slave link.
Nature 2002; 41 7: 213.
190 Pa abo S. Molecular cl on in g of anci ent Egyptian mumm y
DNA. Naiure 1985; 314 :644- 5.
19 1 Ca laca l G, De lfin F, T'111 M et aJ. Id entificatio n of ex hum ed
rema in s of fire tragedy victims using conve ntio na l method s
and autosomallY-chromosomal short tandem rep eat DNA
pro filin g. Am J Forensi c ivied Patlw12 005; 26:285- 9 1.
192 Ca lacal G, De Ungria MC, Delfin F et al. Iden titlca tion of two
fire vict im s by co mp arat ive nuclear DNA ty ping of skeleta l
remains and stored umbili ca l tissues. AI1I J Forensic Med
Patilol 2003 ; 24 : 148- 52.
References I 419
193 Meye r HJ. The Kaprun cab le ca r fire disaster - aspecrs of
forensic organ isation t'ollowing a ma ss fata lity with 155
victims. Fore nsic Sci fJI t 2003 ; 138: 1- 7.
194 Clayton TM. Whitaker JP, Maguire CN. Id en tificati on of
bodies from the sce ne of a ma ss disaster lI sin g DN A
amplification of sholt ta ndem repeat (STR) loci. Forel1sic Sci
fl1t J995: 76:7 -15
195 Hsu CM, Hu ang NE, Tsai LC et al. id entificatio n of victims of
the 1998 Taoyuan Airbus cras h accid ent using DN A analys is.
JJIt J Leglll Mer! 1999; 11 3:43 - 6.
J96 Pres iden t's DNA initiative. Lessons learn ed from 9/ll: DNA
identification in mass fatality inciden ts, 200 6. Avail able
online: www.ncj rs.gov/pdftil es I/110/214 78 1.pd f (acc essed
Ap ril 2007).
197 Bre nner CH. We ir BS. Issues a nd strateg ies in the DN A
identification of World Trade Ce nter victims. Theor Popul
Bio i 2003; 63:173 -8.
198 Coble 1\110, Butler J1\I1. Ch'lracterization of new miniSTR loci
to aid ana lysis of deg raded DNA. J Foreusic Sei 2005;
50:43 -53.
199 Butler J iVI. Single nucleotid e polymorphis\11s ami oth er
bi -allelic mark ers. In Forellsic DNA T)'pin g, 2nd edn.
Amste rdam: Elsevier, 2005.
200 Sob rin o B, Brion lVI, Carracedo A. SNPs ill fore nsic genetics:
a review on SNP typ ing methodologies. Forens ic Sci fni
2005 : 154:1 8[ - 94.
201 Fore nsic Science Se rvice Casefiles : Craig Harman - Family
DNA link offers crime breakthrough. Tile Forensic Science
Sen'ice Website, www. forensic.gov. uk / forensic_t/inside/
news/lis c casefiles.php?case=24 (accessed April 2007).
202 Beiber FR. Guilt by associa tion. Ne/U Scientist 2004;
2470:20.
203 Gree ly HT, Riordan DP. Garrison NA , Mou ntain JL. Family
ties: the use of offen der da ta bases to catch offen der's kill.
J Law M ed Erliics 2006; 34: 248- 62 .
204 Haimes E. So cia l and ethical issues in the USe of familial
searchi ng in forensic investigations: insi ght from famil y and
kin ship stu dies. J Law Med Ethics 2006; 34: 263 - 76.
205 Grimes EA, Noake PJ, Dixon L. UrlJ uart A. Sequ~nce
polymorphism in the human l11 elanocoltin I rece ptor geJl e
as an indicator of the red hair phenotype. Forensic Sci flit
200 1; 122: 124- 9.
206 Rees JL. Geneti cs of hair and sk in co lor. AIIII Re/J Genet
2003; 37 :67 -90.
207 Uil11aso n RL, MohicJeen MPK, Mes t JR et al. SLC24A5, a
pu tative cation excha nger, affects pigmentation in zebra fi sh
and huma ns. Sciellce 2005: 310: 17 82-6.
208 Cho MK. Sa nk a r P. Fore nsic genet ics and ethica l legal and
social implicati ons beyon d th e clinic. Nat Genet 2004;
36:S8 - J 2.
209 Ossorio PN. About face : fo ren sic genetic test ing for race and
v isible trai ts. J Law Med Ethics 2006; 34:277-92.
210 Will iams R. Johnson P. Inclusi veness, eITectiveness and
intrusiveness: issues in the develop ing uses of DNA profilin g
in supp ort of cri minal investiga tions. J Law /Iiled Eth ics
2006; 34:234-47.
211 Alva rez M, Ballantyne J. Th e ide ntification of newbo rn s
usin g messenger RNA profiling analys is. Anlll Biochem
2006; 357:21-34.
I CHAPTER 22 I
THE DENTIST'S ROLE IN CHILD ABUSE
AND NEGLECT
David Whittaker
Introduction 420
Dental neglect 422
Facial and oral pathology 422
INTRODUCTION
it is not the responsibility of general dental practitioners
Damage to a child, other than by accidental means, has been
recognized for centuries I but app ears to have firs t been
described in a formal manner in 1962. 2 The authors used
the term 'batte red child syndrome' but this forms part of
what is more commonly called 'child abuse'. Ch ild abuse
may include non-accidental injury (NAIl but also emotional
ab use, sexual abuse and neglect. It has been increasingly
recognized that the denta l surgeon may be the first profes­
sional to suspect physical abuse, especially in relation to
injuries involving the orofacial structures. 3 A child is consid­
ered to have been abused if he or she is treated in a way that
is unacceptable in a given culture at a given time. 4 For the
purposes of this review, only physical abuse in the form of
NAI and neglect will be considered . Neglect occurs when an
adult knowingly allows a child to endure pain or suffering or
fails to provide the basic requisites for proper maturation.
Since the syndrome was specifically named it has
become possible to add fu rther injuries to the pattern of
injuries described. These further injuries include bite
marks,s which may be found anywhere on a child's b;dy.
Bite marks in children rep resent child ab use until proven
otherwise. They are rarely acci denta l and are good indica­
t ors of genuine child abuse. 6 There have been ap proxi­
mately 50 articles in the international dental li terature
relating to child abuse and bite marks. The dentist's rol e in
diagnosis and management was described in 1986. 7 How­
ever, the problem is still not well appreciated by the profes­
sion at large. Apart from the specific injury of bite marks,
The dentist accused of child abuse 425
Bite marks 425
References 432
(GDPsl to diagnose child abuse. B Dental surgeons are, how­
ever, in a good position to diagnose NAI to the face and
oral structures and they should be aware of the possibility
of abuse and have some knowledge of the key factors
involved in its diagnOSis. The diagnosis of NAI is difficu lt
to make with certainty. and va ri ous pathological or post­
mortem changes have been mistakenly identified as child
abuse.9 Conversely, child abuse or neglect has gone unde­
tected in some unexp ected deaths in children 3 or in re po rts
of facial or intraoral injuries. 3 The incidence of NAJ to the
head and neck region is high, v arying between 11 and 50
per cent of cases. 10,11 Therefore, the dental profession com­
prises an important part of health professionals involved
in identifying and reporting child abuse or neglect. 12,13 A
number of key indicators of NAt have been identified . 14
The more of these that are present the more likely it
becomes a defini tive dia gn osis:
• delay in seeking help;
• a vague history or a history that v aries between
interviewees;
• a history incompatible with the injury seen;
• abnorma l behaviour of the parents;
• the child's app earance, e.g . withdrawn, frightened or
sad ;
• other signs of abuse, such as neglect or deprivation;
• dis closure of an incident by the child.
The diagnosis of NAI is discussed in greater detail in
Chapter I.

The prevalence of child abuse is reported differently in
different countries, presumably reflecting social and cultura l
differences. 15 The true extent in a given population is diffi­
cult to determine, but in 1967 in London it was believed that
at least 300 cases occurred per year. 16 In terms of mortality,
in the 1980s at least 4 children in Britain 17 and 80 children
in the USA 18 died weekly as a result of abuse or neglect. It
was reported in 1988 in Britain that at least one child in
every thousand under 4 years of age suffers severe physical
abuse each year. This might include fractures, brain haemor­
rhages, severe internal injUli es or mutilation. 19 In 1991, an
average of 4.2 children out of every thousand in the UK were
known to have been physically abused , the highest rate
being seen in children less than 1 year of age. 20 It is thought
that one in 10 000 children in Britain will die as a result of
abuse, although many workers consider mortality rates to be
considerably higher. It is generally believed that proven
cases of child abuse may be four or fi ve times as
frequent as t hey were some 10 years ago. 4 In 1974, 97 loca l
authorities in the UK were approached by questionnaire in a
survey covering 90 per cent of the population. Some 5700
cases came to light in the last 9 months of that year with a
mortality of 0.7 per cent and a significant risk of re-injury.21
In the USA, it is mandatory for health-care workers to
report suspected cases of child abuse. The second National
Incidence a nd Prevalence Study of Child Abuse and
Neglect22 reported that 4.95 children in every thousand were
physically abused. Girls were more li kely to be sexually
abused than boys, although there was no gender difference
for physical abuse. Low family incom e is significant, and
children of Afro-Caribbean origins were more likely to be
physically abused than other racial groups. Sexual abuse is
more common in Caucasian children. Whether or not the
true occurrence is increasi ng, it is clear that increased aware­
ness results in many more cases being reported and, on the
basis of current statistics, an individu al dental surgeon may
be expected to encounter a case at least once every 3 years.
In spite of this, British dentists have not formally been
requested to develop an awareness of NAls to children, and
many health authority child abuse procedure documents
make no mention of the possible role of the dentist. 23
Since the Children Act (1989) and Working Together
(1991), child welfare services in the UK are required to have
procedural guidelines to deal with suspected child abuse
cases. These app ly to aU persons working with children,
and general denta l practitioners are includ ed in this group.
National guidelines do not exist for GDPS,8 and the document
Working Together (1991) gives very little information about
the role of the GDP. The British Dental Association in its
booklet Ethics in Dentistl)! does produce guidelines for child
protection 24 but it suggests that each individual GDP should
meet the guidelines laid down through their local district
health authority. It is important that GDPs keep adequate and
accurate up-to-d ate dental records of all their patients; in the
case of child abuse, these may need to be referred to in case
conferences or even court procedures. The guidelines point
Introduction I 421
out that the relationship between parent and child, the child's
reaction to other people, the child's reaction to a dental or
medical examination or the general demeanour of the child
may have a bearing on possible child abuse.
It is important that the dental practi tioner without direct
extensive experience in referral techniques refer cases to
the local consultant in paediatric dentistry, the local con­
sultant in dental public health or an appropriate consultant
paediatrician. It may be necessary and wise to discuss any
potential referral with the general medical practitioner of
the child concerned. Under the 1989 Children Act and
Working Together (1991) the local authority has a legal
duty 'to make or cause to be made' such enquiries (by
another agency, such as the National Society for the Pre­
vention of Cruelty to Children [NSPCC]) as t hey consider
necessary, to enable them to decide whether t hey should
take any further action to safeguard the child 's welfare.25 ,26
There are a number of reasons why general dental practi­
tioners or their staff may fail to report potential child abuse.
Therefore, each local area should have a simple system,
preferably referrin g t he case through the normal referral sys­
tem that dentists might use, such as to their local paediatric
dentist or paediatric consultant. Failure to refer a child may,
of course, have serious consequences. Referral rates by GDPs
are not known, but it is known that healt h professionals in
general refer 17 per cent of suspected abuse cases, whereas
teachers and school nurses refer 23 per cent of cases,u
Reporting cases of child abuse is mandatory in the USA,
where 10 subtypes of child abuse and neglec t have been
classified 28 The most frequent of these is physical abuse,
with an incidence of a lmost 32 per cent, and this is closely
followed by educational neglect (28 per cent) and byemo­
tional a buse (26 per cent).
Suggestions as to how t he dentist should examine a child
thought to be suffeling from child abuse or child neglect have
been made. 29 These include the relationships between the
child and his or her parents; the cleanliness, stature and
nutritional status of the child; the quality of his or her cloth­
ing; and the presence of physical problems, such as a limp or
difficulty in climbing into the dental chair. It is recommended
that the body be systematically examined for traumatic
injury, bruises, previous scars, burn marks or injury caused by
foreign bodies. If the dentist's suspicions are aroused, the par­
ents should be informed that a n injury has been observed; the
dentist must then contact the approp riate child abuse or child
neglect authority. In the UK, it has been said that it is not the
responsibility of GDPs to make the diagnosis of child ab use,
nor should they ever do SO.8 However, the same authors state
t hat no GDP should feel gui lty abou t referring children with
suspected child abuse. This does not mean that they are
accusing either parent or carer but they are simply seeking a
second opinion on a condit ion that is outside their expertise.
Failure to refer a child in need may result in more severe
injuri es in t he future. The failure to follow up a suspicion of
child abuse by a professional health-care worker may be
construed as a form of professional negligence. Accidentally
422 I Dentist's role in child abuse and neglect
acquired bruises or pseudobntises must be differentiated from
inflicted injuries. It is self-evident that infa nts who ca nnot
crawl cannot cause a self-inflicted accident. It follows that
severe bruising or fractures in a child below th e ages of 6-9
months are almost always inflicted non- accidentally by a
second party.
Alth ough not directly involved, general dental practition­
ers should be aware of the problem of child sexua l abuse,
although this requires specialist and expert intervention
from those health-care workers trained specifIca lly in this
area. 30 There is certainly in adequ ate in for mation both in
training and in th e dental literature to assist the paediatri c
dentist in examining these patients. Any physical examina­
ti on of a child should normally be confmed to those parts
of the body considered to be within the expertise of the
dentist. Suspicion of abuse lying in the purely medical fi eld
should be referred to a consultant paediatrician for complete
examination.
DENTAL NEGLECT
Neglect of a child 's dental treatment needs could be con­
strued as a form of abuse. In this insta nce it is best described
as an act of omission rather than an act of commission.
Gross dental caries, exacerbated by poor diet, use of car­
cinogenic drinks and poo r ora l hygiene procedures, clearly
lie within the resp onsibility of the child's parents. However,
the rate of progression and extent of dental ca ries may also
be genetically influenced and may dep end to som e ex tent
on the 'q ua li ty' of the tooth ena mel, on the microbiological
flora of the oral cavity an d on the flow and buffering cap ac­
ity of sa li va. These are clearly not the responsibility of the
parents. Th e issu e is therefore co mp lex an d sho uld be
app ro ached with care.
In the USA, state laws have defined chi ld neglect as
negligent tre atm ent, including a failure to provide ade­
quate care, su pport, nutrition or med ica l or surg ica l care so
as to threaten a child's health or welfare. In Ca lifornia, den­
tal neglect is listed as a component of physical neglect. 28
In 1998, 15000 denta l surgeons in California were ques­
tioned concerning their knowledge, attitudes and practices
regarding child ab use and neglect. 31 Seventy-one per cent
had never seen a case of child abuse or neglect. Twen ty-eight
per cent received no forlllal training in the recogn ition of
child abuse or neglect whilst at dental schoo l. However, 84
per cent had read postgraduate literature on the subject s~nce
qualifYing. Sixty-four per ce nt were aware that they were
required by la w to report ab use and neglect, but alm ost the
sa me percentage did not kn ow whether failure to report was
a felony. Approximately the same percentage were aware that
fa ilure of parents or gu ardians to foll ow through in dental
treatment once they had been infoillled about a chi ld 's ram­
pant caries could be considered as child neglect. AJ.most hal f
of those questioned were aware that there is a strong con'ela­
tion between dental neglect and the presence of physical
neglect. 32 Betvveen 1991 and 1992, 2274 paediatric dentists
and orthodon tists were surveyed in relation to reporting child
abuse. Forty-two per cent of pa ediatri c dentists reported hav­
ing seen a case of suspected child abuse in the previous year,
whilst only 15 per cent of the orthodontists did. More signif­
icantly, only 11 per cent of the olthodontists actually reported
their suspected cases, whereas 48 per cent of the paediatric
dentists did. Among those paediatric dentists, denta l neglect
was a frequent reason for suspicion?1 Dental neglect, a com­
mon form of child maltreatment, should be suspected if ram­
pant caries and oral infection, bleeding and trauma persist
despite the elimination of financial and transportation obsta­
cles. Questioning the child and parents separately may help
un cover an obvious discrepancy ben,yeen the clinical findings
and the histo ry33 There are no published repolis of child
abuse or criminal neglect in th e UK relating to advice or con­
trol of gross caries. Guidelines from the USA 34 indicate that
dentists an d dental personnel have an obligation to repoli
occurrences of dental neglect. Dental neglect is 'wilful failure
of parent or guardian to seek and follow through with treat­
ment necess ary to ensure a level of oral health essential for
adequate function and freedom from pain and infection'.
Dentists and dental staff should be trained in recognizing and
reporting this type of ab use and neglect. There is evidence
that dental neglect is misunderstood an d under-reported.
Certain fami lies appear to be more at risk of such neglect than
others 35 Dentists, particularly paediatric dentists, are in a
unique position to eva luate not only the dental needs of chil­
dren but also famil y situations and circumstances. Greater
collaboration between paediatricians and dentists in the eval­
uation and management of ab use an d neglect is in the best
interests of the child 36 It is clear that further research is
required to develop effective methods, both for eva lu ation
and for the education of at-risk individu als.
FACIAL AND ORAL PATHOLOGY
The facial region is said to be a common site of NAls in chi l­
dren, and it has been claimed that almost 50 per cent of chil ­
dren suffer injUly to this part of the body.5 Of 260 victims of
ch ild ab use surveyed in Boston, 49 per cent had sustained
orofa cia l trauma 3 ? and a to ta l of 65 per cent sustained head
and facial tra um a. In fact, head or facial trauma was the prin­
cipal reason for ad mission to the hospital in 45 per cent of all
cases. Despite this, few of the 537 dentists surveyed at that
time knew of their legal responsibility to repOli these injuries.
Eleven per cent of them had seen orofacia l trauma cases of a
suspicio us nature. Ho wever, only 22 confirmed cases of child
abuse had been noted by th e den tist ; of these, only four were
repolted to the social agencies. Other in vestigato rs have
reported an incidence of 60 per cent of cases involving injury
to the head and neck region 38 The incidence in the UK of
fa ci al injuries related to child abuse has been cited as 43 per
ce nt. 39 Whilst it may be open to discussion as to whether
dentists should become involved in the diagnosis of child

abuse, 15 it seems clear that he or she would have an imp or­
tant role to play in the recognition, diagnosis and treatment
of injuries to the face and oral cavity. The largest survey
ava il able appears to be one from 1992 which investigated the
denta l aspects of 1248 cases of child abuse in a major Amer­
ican hospital over a 4-year period. 12 There were an equa l
number of girls and boys, except where sexual abuse was
involved. Fifty-three per cent of children were in the 0- to 4­
year age group, and when all cases were considered together
38 per cent presented with injuries to the head, face and
mouth. In a smaller group surveyed the same yea r, 56 per
cent of children had abnOlmalities within the head and neck
region. In a larger studlOover a 5-yea r period, 4340 patients
were studied. Sixty-eight per cent of these were victims of
sex ual abuse and 32 per cent of physical abuse. Of these
physically abused children, 49 per cent had evidence of
injury to the head and neck region, whereas in the sexually
a bused children only one per cent had injuries in this area.
There were 11 deaths in this particular series.
Extraoral Injuries
The head and facial region is a common site for NAI not
only in t he West, 12 but also in t he Indian continent. 41 It is
not known why the face often seems to be the focus ofvio­
lent attack in children, but it may be because it represents
the persona of the individu al.
The types of injury have been classified J4 and include
contusions and ecchymoses (66 per cent), abrasions and lac­
erations (28 per cent), burns and bites (four per cent) and
fractures (two per cent). Bruises, lace rations, burns, cuts,
scars and black eyes will suggest NAI.4 2 Extensive bruising
of the face with a history of limite d trau ma, especially if both
old a nd n ew bruises are present , is almo st diagnostic. High­
velocity slaps or lashes may cause an unbnlised, negative
image of the object that caused the injury. 2o The margin of
the injury may often be demarcated by petechial haemor­
rhage from capiliaries. 4J This type of injury is sometimes
known as the tattoo bruise. Bruising to the soft tissues of the
fac e a nd head may occur in many different ways, but acci­
dental falls rarely produce bruises in the areas of the cheek;
they are more likely to involve skin overlying bony promi­
nences, such as the forehead or zygomatic bone. Bruises
change colour as they age,44 but the rate of change from red
to blue, to purple, to green, to yellow cannot be used as a
reliabl e indicator of the age of the injury.
Many injuri es to the head and neck are caused by the
adult human hand . This can leave press ure bruises in the
form of grab marks by fingers and thumb, and these are
velY commonly seen follo w ing pinching or pulling of the
ear. In these circumstances, there wiH usually be a match­
ing bruise on the posterior surface of the ear. They are seen
more commonly in babies and infants and may be indica­
tors of potenti ally more serious injuries, such as subdura l
haemorrhage from shaking, and should always be taken
Facial and oral pathology I 423
seriously. Grip marks on the cheeks or chin can arise dur­
ing restraint during feeding. Hand marks may also be in the
form of a slap or a punch. Slaps may leave an imprint of
the abuser's hand, whereas punches to th e forehea d, nose
and eye region may result in unilateral or bilateral peri­
orbital ecchymosis. Other reasons should be eliminated ,
such as infective, ren a l or haematological conditions but, if
the dentist has any doubts concerning unexplained or sus­
picious lesions on the face, the child should be referred to
a consultant in paedi at ric dentistry or in paediatrics.2,27
Approximately 10 per cent of NAIs to the child involve
thermal injUly 4 5 These may be burns (73 per cent) or sca lds
(27 per cent). Thirteen per cent of non-accidental thermal
injuries occur on the face , head and neck;4Gthe rol e of the
dental surgeon is clear in recognizing, diagnosing and
reporting such injuries. Between 2 and 12 per cent of all
inpatient burns are NAls infl icted with the purpose of pun­
ishing or controlling the child's behaviour. Burns to t he
face have been found to be the most frequent injury in
so me studies,J9 and those from hot, solid objects a re the
easiest to diagnose, and are usually of second degree with­
out blister formation. The shape of the burn may resemble
the hot object producing it. For example, cigarette burns
produce a circular, punched-out lesion of uniform size, but
these are relatively rarely seen on the face. However, they
are pathognomonic for child abuse. The use of hot hair
dryers to inflict facial injury has been reported. 47 Friction
burns have been described , and these present with broken
blisters over bony prominences, including those of the
face, and can be caused by being dragged across a carpet.
Scalds can be on any part of the body but, in so far as the
face and head are concerned, the most common injUly is
due to hot liquid being thrown at a child. 2o
Non-accidentally produced fractures of the facial bones
are uncommon in children and feature in only two per cent of
cases. 42 Several patterns of skull fracture that are indicative of
NAI have been described. 48 These are multiple fractures with
complex configurations - fractures that a re depressed, wide
or more than 5 cm long ; involvem ent of more than one cra­
nia l bone and non-parietal fractures ; or fractures associated
with intrac ranial haemorrhage. Multiple injury and inade­
quate explanation in the history are the best indicators of a
skull fracture being of non-accidental origin. Most fractures
in physically abused children occur up to the age of three. 49
Accid ental fractures occur more commonly in older children
of schoo l age. The most common facial fracture is nasal (45
per cent), mandibular (32 per cent), and zygomatic and orbit
complex (20 per cent).50 Specialized treatment is required for
facial fractures and their definitive management has been dis­
cussed in the oral surgery literature. J7
Intraoral Injuries
A review of 1248 cases of child abuse 12 desc ribes five tooth
fractures , three tongue and tongue frenulum lacerations,
424 I Dentist's role in child abuse and neglect
two lip frenulum lacerations, 11 injuries to the ora l mucosa ,
eight palatal lesions, loose a nd missing teeth a nd ev idence
of den ta l neglect. It has also been suggested 51 that sexu ally
trans mitted diseases in children's mou t hs are a strong indi­
cation of sexual abuse. There are claims that child re n
exposed to oral sexual activities over a co nside ra ble perio d
of tim e may develop higher ca ries rates a nd erosion on the
palatal surface of the maxill ary teeth. Den tal neglect in
these patients may be pal1 of gen eral neglect.
Although some reports in the literature have included
the term 'neglect' in t he title, only a small number of these
have discussed dental neglect as distinct from gen eral neg­
lect of the chi ld. An extraordinary case was reporied in
the UK in 1990, whi ch escaped detection for several years.
Three children in a family of six were found to ha ve miss ­
ing permanent incisors. It tra nsp ired that their parents were
extracting teeth as a punishment for bad beh aviour. 52
A 6-year-old Hi spa nic girl was reporied to have avulsed
a tooth during physi cal abu se and the sw allowed tooth
caused an oesophageal perforation resulting in a retropha ­
ryn gea l abscess a nd a medi ast inal abscess co nta inin g the
toot h. This ap pears to be a unique ma nifestation of dental
neg lect. 5)
The mou th is a frequent site of NAl in children, possi bly
because of its psych ological signifi cance; it is the orga n of
phonation , a nd v iolence may be used aga inst it in order to
silence the child. 54 Contusi ons and bruises, which occurred
in 43 per cent of a series of 14 cases reviewed , are the most
common injuries to the mouth;42 lace ra tions and trauma to
the t eeth eac h account for 28 per cent of injuries.
Examination of the mouth should commence with the
lips; the lips usually heal well follo wing min or injuries a nd
scarring should alert the dentist to previous severe trauma.
Bruised and swo llen lips or ab rasions at th e corners of the
mouth are suggestive of bl ows from a fis t; NAls are espe­
cially likely if the bruises or injuri es are at varying stages
of healing.
The teeth themselves are vulnerable to blows to the
mouth, from either a fist, a hand slap or a weapon. It must,
of course, be remembered t hat children frequ ently avulse or
fracture teeth in the course of normal play activities, but an
unsa tisfacto ry history from the child or parent, or repeated
atte ndance for treatment of injuries, should arous e the sus­
picion of the dent ist. The roo ts of the anterior teeth may be
incompletely formed in children and, as a result, blows to th e
face often result in complete avulsion and loss of the tooth
rather than in fra ctu re of the roots. When teeth are missing,
full radiographs of the jaw should be taken to rule out some
abnormality of development or conditions of partial ano­
dontia; retained frac tured roots should be sought as these
will require surgical removal. Mult iple healed micro frac tures
of tooth roots may sometimes be v isible in intraora l radio­
graphs - these are al most always indicative of trauma by
repeated abuse. Slight movements of the teeth may rupture
delicate blood vessels pass ing through the apical fo ramina
and, as a result, anterior teeth that survive trauma of this
-~~~ -
nature may become devitalized, especially if the root apices
have completed their development. The crowns of the teeth
are likely to become discoloured in the months fo llowing
such trauma. A grey hu e usually develops ow ing to the pres­
ence of blood breakdown produ cts in the dentinal tubules,
a nd the presence of discoloured anterior teeth for which a
sa tisfactory ex pl anation is not forthcoming may suggest
t hat NAI has occurred.
The report of the au topsy findings in 29 fatal cases seen
durin g a 2-year period at the London Hospital Medical Col­
lege included o ne of the first assessments of soft-tissue
intraoral injury fo llowing child ab use. 5 Fifty per cent of vic­
tims had injuries to the head and neck, and 45 per cent of
these had suffered damage to the upper lip region. This con­
sisted of laceratio ns to the mucosa of the inner aspect of the
upper lip, close to the labi al frenulum ; these were so metim es
so severe that th e reflected mucosa of the lip was completely
tom away from that attached to the g ing iva. Tea ring of the
labial frenulum is not uncommon in very small children,
who may fall accidentally onto the mout h whilst learning to
walk, but this lesion should arouse suspicion as to a possible
non- accidental cause when it is found in children less tha n
about 1 year old and more than ab out 2 years old. These
injuri es can be produced by agg ressive bottle feeding or by
forcing a spo on into the mouth, as is sometimes done by a
frustrated parent whose child is a slow eater. The labial
frenulum may also be damaged by blows to the mouth or by
tearin g the upper lip away fro m the gingival attachment
(Fig. 22. 1) (see a lso Fig. 8.6, p. 152). Following this report, the
tom labial frenulum has been regarded as pa thognomonic of
child abuse, but it must be noted t hat no later reports of intra­
oral da mage in cases of child abuse have indicated such a
high incidence. Many paediatric dentists of considera ble
experience worki ng in the UK have never seen a case of labial
fren al damage that t hey could unequivocally relate to NAI.
When it does occur, the condit ion is often associated with
subluxation or fra cture of the upper centra l inciso rs, damage
to the teeth being caused by the same blow.
Figure 22.1 Dam ag e to the labia l frenum may be caused by
forcible feeding or, as in this ca se, by a blow to the mouth.
Teeth may be displa ced.

Penetrating injuries to the palate, the vestibule and the
floor of the mouth resulting from forced feeding or forcible
introduction of spoons and other implements into the
mouth have been described. Burns of the oral mucosa fol­
lowing enforced ingestion of hot or caustic liquids are seen
quite commonly,27 as are intraoral cigarette burns and tear­
ing of the oral mucosa by a parent's finger placed in the
mouth. Dentists who are unfamiliar with the range of oral
pathologies of a non-accidental nature that may occur in
the mouth should refer patients to a specialist if they are in
doubt as to the cause of a particular lesion.
THE DENTIST ACCUSED OF CHILD ABUSE
It is known that child abuse or NAIs to children are fre­
quently caused by an adult in close relationship to the child.
It has been pointed out 55 that this may include a ch ild 's pro­
fessional adviser, such as a dentist. In the state of Tennessee
in the USA malpractice insurance rates have increased by
1500 per cent between 1981 and 1987, and cases of child
ab use/criminal assault have been reported against several
dentists performing procedures on patients. Dental practi­
tioners have been accused of child abuse by using excessive
restraint, for example by placing a hand over the mouth in
order to deal with a difficult or noisy patient. This exercise
was singled out by the Virginia Board of Dentistry as lead­
ing to reports of child abuse against dentists.
Another risk is the non-consensual touching of a
patient, even if this is aimed at improving a child's health.
Questionnaires in the USA are currently being designed to
discover how many dentists have been accused of battery
or non-consensua l touching. The information is not easy to
acquire, and until 1987 only one repol1 was available cov­
ering three cases. 55 These three incidents occurred during
the course of normal t reatment and all three resulted in an
arrest. As child abuse is a criminal offence, if it occurs in
the dental surgery it is not covered by malpractice insur­
ance. 56 There is a danger of dentists becoming overdefen ­
sive in these matters, and this may eventually be to the
detriment of child patients.
BITE MARKS
Bite marks have come to mean any injury or mark produced
in flesh, foodstuffs or other material by the teeth and the
surrounding soft tissues. 57 In non-living materials and in
foodstuffs, the mark left is usually produced entirely by the
teeth, but in bites left on human skin, the injury may be
contributed to by pressure and suction and/or pressure from
the tongue as well as from the teeth themselves. Recogni­
tion of bite marks as part of the multiple injuries sometimes
seen in NAIs to children is a relatively new development.
Attention appears to have been first drawn to the incidence
of bite marks in NAls in a report of 13 forensic dental cases,
Bite marks I 425
two of which involved teeth marks in battered babies. Trag­
ically, both children died. 58 Shortly afterwards, a case of a
bite on a 17-month-old child was reported ;59 bite marks
and a torn labial frenum, associated with battered child
syndrome, were described . 5 The incidence of bites in victims
was given as 12 per 100000 in the population of New York
in 1977. 60
However, there are few studies on the frequency of bit­
ing associated with child abuse. In one study 61 nurses work­
ing in facilities that admitted child abuse patients were
trained by forensic dentists to recognize bite marks on their
patients. Bites were divided into incised bites without bruis­
ing, those incised with bruising or discolouration, marks
with sucking bruises and marks in which the teeth had been
d ragged over the tissue. Of 1100 children examined, 17 had
evidence of bite mark abuse - an incidence of 1545 per
100000 sheltered chi ldren. Most of the bites had been
inflicted on children between the ages of II and 15 years,
but two were in ch ildren less than 3 years of age and four
in chi ldren between 16 and 18. Most of the male victims
were in the 4- to 10-year age group, whilst most females
were in the 11- to 15-year-old age group. The most common
site of bites was the head and neck region (43 per cent). The
limbs and trunk were also common sites of attack. In an
unpublished studyG2 of 99 cases of bitten children there
were 169 bites. Thirty-three per cent were in children aged
between 6 and J 8 months, 23 per cent in children aged
between 18 months and 3 years but there were 17.2 per cent
of bites in children under 6 months of age. There were no
statistical differences in this study between the girls and the
boys. Few studies have been carried out on the distribution
of bite marks speCifically in chi ld abuse cases, but the
anatomical distribution has been investigated in a series of
67 victims of varying ages. 63 Among this group, 13 were
aged below 15 years. Female victims were most commonly
bitten on the breasts, arms and legs, while the arms and
shoulders were the most common sites in males. These
results differ somewhat from those reported in the UK,
where bite marks on the breasts of females were found to be
more common than in the American sample. Bite marks on
the arms were seen less frequently.64
Other studies on the anatomical distribution of bites
include th ose on 122 injuries observed in a New York hos­
pital 65 and on 114 cases reported from Kansas.66 In both of
these studies, bites were most often seen on the hands and
fingers. It appears that predilection for the different parts
of the body has changed in more recent times. In NAIs in
children, the entire surface of the child's body may be
attacked, resulting in bite marks on the cheeks, shoulders,
chest, abdomen, arms, legs and buttocks. 67 Multiple bites at
the same site have been described. In the Cardiff series,52
37 per cent were on the shoulder or arm, 31 per cent on the
leg or buttock, 18 per cent on the head or neck region (of
which 15 per cent were on the face) and 12 per cent were
on the ab domen or back. Less than three per cent were on
the chest region. There were no significant differences in
 426 I Dentist's role in child abuse and neglect
Figure 22.2 Multiple bite marks are often seen. This child had
16 separa te injuries.
location between the males and females. Eighty per cent of
th e children had more than one bi te, compared with only
20 per cent in a comparable adult group. Of these, the boys
were twice as likely to have multiple bites as the girl s and
one of the boys had 16 bites (Fig. 22.2). Bites on the hands
and fingers may be ca used when th e child attempts to pro­
tect him- or herself. Some foren sic odon tologists believ e
that bi te marks found on infants tend to be in locations di f­
feren t from those found in older children or adolescents
and are meant as punishmen ts to the child in response to
crying or soiling. These bites tend to be co ncentrated on
the cheeks, arms, shoulders, buttocks or genital ia. Bites
in tended as punishment for soilin g are most commonl y
seen in the last two locations.
There is considerable disagreement as to the location of
bi te marks in NAl cases. Some observers take the view that
bi tes on the breast are much more common in females,
whilst bi tes on the arms are more common in males. G3 In the
Cardiff survey, th is was true only in individuals over 16
years of age. The overall frequency of bite marks in children
is not clear, but a three-month study of a sel ected juvenile
popul ation,51 among sheltered children, demonstrated an
inciden ce of 1545 bite marks per 100 000 population. This
incidence is si milar to that of diseases such as go norrhoea. It
seems unlikely that such an incidence wou ld be found in the
popu lati on at large. Bite mark injuries constitute one per
cent of all emergency department visits in the USA and are
often associated with child ab use. 58 A recent report from th e
USA69 looked at 101 bite mark cases that had been referred
to Courts of Appeal. Seventeen per cent of the cases we re in
children and all the male children had suffered bites to the
genitalia, whereas in the femal e children bites were found at
all locatio ns, including face, legs, arms and buttocks. Child
bite marks from two centres in the UK were comp ared and it
was found th at 1- to 5-year- olds were at greatest risk.70 The
location of bite marks was simi lar in the children from both
centres, the most common site of injury being the arm. In the
UK,71 an incidence of child bites of 2.25 per 100 000 children
per year was calculated. Many of these children had bite
- -
- - -
Injunes as isolated findings, with only 25 per cent of the
children having associated injuries. Seventy per cent were
caused by adults and 30 per cent by other children. The
importance of a close working relationship between paedia­
tricians and forensic dentists is emphasized. Other studies
have claimed that quite frequently there are other injuries,
such as pinch marks, bruises or burns.72 Bite marks in the
older child may be less punitive in nature and are more likely
to be related to physical or sexual abuse. Distinguishing
between adul t and child bite marks may not be straightfor­
wa rd . Using receiver operating characteristic analysis, it has
been shown that the best decisi ons were made by senior or
junior experts in the field. General dental practitioners and
police officers were the least able to differentiate correctly
between adul t and child bite marks. The effect of training
was emphasized.73
Types of Bite Marks in Children
Determinin g the aetiology of a human bite mark may be
difficult. There is general agreement that there are two
main patterns of bite marks. The first is an aggressive or
anger bite in which the teeth, in the main, are used alone
a nd the marks on the ski n most cl osely resemble the shape
of the teeth mak ing them. The mark results fro m an attack
or defence bite and is the ty pe most frequently seen in NAI
to the child 72 This kind of bi te is a particularly aggressive
one . No suction of lips and tongue is involved, th e tissues
being bi tten directly between the teeth. On occasions, sub­
stantial abras ion marks may be associated with each tooth
mark, suggestin g movement durin g the biting episode.5
The secon d type of bite mark is on e that is slowly, sadis­
ti cally an d deliberately inflicted . Suction or sometimes pres­
sure is applied to the soft tissues via the lips and to ngue.
These types of mark are more often seen in abuse in older
children with a sexual overlay. The injUly demonstrates
bruising in the central part of the bite and sometimes in the
peripheral areas, together wi th linear radiating abrasi ons
caused by the incisal edges of the front teeth. The outlines
of the teeth are usually quite clearly visibl e. The pressure
exerted by teeth during biting may be considerable and may
be as much as 11 kg (550 kPa) from the incisors an d pressure
from the to ngue may rea ch at leas t Sib per square inch
(55 kPa). Durin g suckling activity,14 suction as distinct from
thrust from the tongue and lips, may reach a negative pres­
sure of 20 mmHg (2.75 kPa) in so me bi tes .75 The circum­
stances of the injury will vary a nd , although the present
remit refers to NAl to th e child, it should be re membered
that children may presen t wi th bite marks fo llowin g fights
(61 per cent in one stud y) or pl ay (26 per cent).76,77 Most
children with bite mark injuries are brought to the cas ualty
or paediatric depal1ment by the parents or adul t guardians.
The person who ab used the child and produced the bi tes is
likely to be among th em.78 This, however, is not always the
case, and the examiner should be aware of the fact that bites
--- - - - -
 Bite marks I 427

may be produced on small children by siblings or play­

mates, either aggressively or as a means of punishment, or
occasionally in sexual play. The methods of distinguishing
bites produced by children from those produced by adults
will be described later.
The injury itself may consist of bmising, abrasions, lac­
erations, punctures or any mixture of these in a single bite
mark. In a study carried out on 320 human bites in children
in the USA,76 75 per cent of all bites were superficial abra­
sions, 13 per cent punctures and 11 per cent lacerations.
This may be important since none of the abrasions became
infected but 38 per cent of the punctures and 37 per cent
of the lacerations did. In the Cardiff survey,62 82 per cent of
bite marks in children were categorized as bruises, nine per
cent as lacerations, eight per cent as abrasions or scratches
and one per cent as amputation of tissue.

Self-inflicted Bites

Self-inflicted bites may be emotional responses to extreme

pain or a type of counterirritation to alleviate pain?9 It has
been suggested that the arms may be pushed into a child's
mouth by an assailant in order to stifle crying, so but self­
biting may also be a type of self-destructive behaviour,
such as has been described in individuals who are mentally
retarded or psychologically disturbed. Self-biting may
occur in adults, and a case has been described in which a
51-year-old man inflicted a bite mark on the left wrist dur­
ing a fatal episode of myocardial ischaemia 79 It is obvious
that self-inflicted bite marks can be made only on parts of
the body that may be placed in the mouth, but one suicide
victim is reported as having managed to self-inflict a bite
upon the left breast. In alleged NAI cases it is clearly of
great importance to determine whether self-infliction has
occurred Bo or whether the bite might have been relatively
innocently acquired during childhood play with a sibling
or friend.
Animal Bites on Children
NAI bites on the child may need to be distinguished from
bites produced by domestic or wild animals, usually on
unsupervised infants or older children. Most studies give the
average age of children attacked by a dog as between 5.4
and 6.2 years. BJ.82 Animals may attack the face of a small
child, producing very serious injury. It is usually reratively
easy for a dentist to distinguish between a human and an
animal bite because of the size of the dental arch, and more
particularly the arrangement and size of the teeth - a carni­
vore such as a dog or a ferret,B3 for example, has large
canines and diminutive incisors in both the upper and lower
jaws. The classical bites from these animals present either as
a four-point puncture wound from the canines or as tearing
of the tissue in more aggressive attacks (Fig. 22.3). Small
Figure 22.3 Dog bites may result in four-point puncture
wounds or tearing of the tissues, as in this case.
animals, such as rats, have extremely sharp, razor-like inci­
sors. These can inflict deep and extensive lacerations which
may be mistaken for injuries using sharp implements (Fig.
22.4).11 It may be important from a medicolegal point of
view to determine which animal has made the bites since
the parents or owners may be held responsible if it is a
domestic animal. However, blame may be apportioned dif­
ferently if, for example, the injuries have been produced by
rats that have not been exterminated by the local authority,
as in the case shown in Fig. 22.4. Distinguishing between
different animal bites requires a knowledge of comparative
dental anatomy, but also a knowledge of the habits of com­
mon animals. Watch dogs tend to bite and hold their vic­
tims, whilst untrained wild dogs may move their heads as
they bite, tearing the tissue and making the bite mark diffi­
cult to evaluate. B4 A case has been described in which the
body of a 13-month-old baby was discovered with more
than 80 puncture wounds, initially thought to have been
caused by a mechanical instrument. It was eventually
shown that the child had been injured and finally killed by
two German Shepherds. 8s Although fatal animal bites are
rare, two cases were described in Germany,BG and in the
USA there were 238 deaths from dog attacks over a
20-year period. Most of the dogs were pit bull terriers and
Rotweilers. 87 The author dealt with a case in south Wales
where a pet German Shepherd savaged a three-month-old
sleeping baby and dismembered the body. All this occurred
in the presence of the sleeping father (Fig. 22.5).
 428 I Dentist's role in chi ld abuse and neglect
Figure 22.4 Multiple bites from small anima ls may result in
serious facial injury extending to loss of an eye. These injuries
were caused by a rat.
Figure 22.5 A sleeping baby was savaged by a dog and then
dismembered by the same animal.
Dogs are probably the most common animal to bite chi l­
dren,88 and most of the injuries involve the extremities. As
a result of their small size, 33 per cent of bites in children
under 5 years of age are to the head and neck, 88 and in
some studies up to 67 per cent have been to the head and
neck. 82 There have been instances of damage to the central
nervous system caused by an English bull terrier, a Dober­
man pinscher and a Bengal t iger89 In nearly all the
reported cases of injury to the head from dogs or similar
- - - -- - -
-- --
animals the child has been younger than 2 years old and
usually less than 1 year of age.
Infections in Human and Animal Bites
Children who survive NAls and who have suffered from
either animal or human bites may develop serious infec­
tions. 9o.9 ! Aspirates from bite wounds in 39 children (21 with
animal bites and 18 with human bites) were cultured for
anaerobic and aerobic bacteria. 92 Aerobic bacteria alone were
recovered in 7, or 18 per cent, of the wounds, anaerobic bac­
teria alone in 3, or 8 per cent, of the wounds and mixed bac­
teria in 29 cases, or 74 per cent. The most frequent isolates
were Staphylococcus aureus, anaerobic cocci and Bacteroides
spp., Pasturella multocida, Pseudomona s fluorescens were
present in animal bites only. Human bites to the hand war­
rant special consideration 9J because of the special risk of
severe infection. Recent data demonstrate that human bites
occurring anywhere other than the hand present no more of
a risk of infection than any other type of mammalian bite.
Cultures of infected bite injuries in chi ldren may yield an
average of five or more micro-organisms. 77
A case of primalY and recurrent herpes simplex infection
was described in a paediatric nurse resulting from a human
bite. 94 In 1988 the issue of transmission of HIV through
human bites and scratches was raised . A total of 198 health
care wo rkers were studied , 30 of whom were traumatized in
this way while caring for an aggressive AIDS patient. 95 This
single violent patient frequently bit his carers, his mouth
contained both blood and saliva and he was HIV antibody
and antigen positive. After 2.5 years of serial follow-up, all
traumatized personnel were clinically normal. It appears,
therefore, that the risk of transmission of HIV through this
route should be low. Two cases were reported in 1996 ; in one
case an HIV-positive female prostitute transmitted HIV to
another individual by biting. In another report, two adult
sisters had a violent fight in which several of the Hrv­
infected sister's teeth were knocked out. She then bit her
HIV-negative sister, who contracted the virus. It appears that
to transmit HIV the biter's mouth must conta in blood. 96
The current position was recently reviewed, and it was
concluded that a bite from an mY-seropositive individual
which breaks the skin or is associated with a previous
injury, carries a risk of infection for the bitten individual. 97
The Investigation of the Bite Marks
Where bite marks are present in a case of NAl, they are usu­
ally to be found on the child victim. However, it should be
rememb ered t hat occasionally they may be on an assai lant,
if the victim has attempted to fight back or is attacking in
self-defence. The same principles of investigation apply in
both cases and, apart from the difficulties of examining a
conscious sma ll child, the methods llsed to investigate bite
----

marks on an adult can be applied routinely in most ch ild
cases. Injuries caused by biting can be thought of as 'tool
marks', the tools in this case being the biting edges of the
teeth. In many ways they are a unique injUly in the sense
that not only do they demonstrate injury which is difficult to
explain away as an accident, but they are also one of the few
injuries that can be related directly to the assailant. Bite
marks in human flesh can be extremely difficult to interpret
because of the dynamic nature of the injury. Distortion may
be caused by the bite itself, by subsequent changes and by
many other factors. It is important, therefore, that an expe­
rienced forensic dentist should be included in the team
investigating cases of childhood abuse. This is still not the
case in many child protection procedures produced by health
authorities in the UK. As a general rule, bite mark injuries
should always be carefully recorded and described before
the teeth of a putative suspect are examined. This is because
even experienced forensic odontologists may introduce an
unconscious bias in to their interpretation if they already
have a particular dentition in mind. The investigator should
be attem pting to fit 'the injuries seen in the bite' to one of a
number of suspects and should not be attempting to fit 'o ne
particular dentition' to a specific injury.
The Bite Marks
There are no detailed protocols for examinations of bite
mark injuries that are acceptable throughout the world,
and individual dentists have developed their own methods.
Guidelines have been laid down by the American Board of
Forensic Odontology.98
Once recognized as such, a bite mark on a child should
always be investigated by a forensic odontologist if abuse
is suspected. Children receive bite marks from adults, other
children or animals, and they are a common problem in
children presenting for medical attention. 99 When child
abuse is suspected, the recognition of bite marks may pose
significant problems for a medical clinician. It may be
difficult to determine whether bruising or lacerations are
caused by a bite because the marks are often incomplete,
distorted and change over time. They may, on occasions, be
confused with skin eruptions .6 ) Collaboration between pae­
diatricians and forensic dentists is a satisfactory solution to
the assessment of bite marks where abuse is suspected. 71
After recording the usual details of the victim, the location
of the bite mark on the body, be it living or dead, is carefully
desClibed . At this stage it is important to note whether the
bite mark is on a flat or a curved surface and whether it over­
lies soft tissue, fat or bone. Bite mark identification entails
several steps: recognition of the wound, documentation, and
interpretation . A human bite mark is usually identified
because of the double arcade shape of the marks. If both
upper and lower arches have left marks, the characteristics
of the resulting bite include an elliptical or ovoid pattern
containing tooth and arch marks,57 which are almost always
Bite marks I 429
produced by the canine to canine of the upper arch opposed
to canine to canine of the lower arch. Incisor teeth tend to
leave rectangular bruises or incisions whe reas the canines
leave puncture or triangular marks. A decision as to whether
an injury is, in fact, a human bite mark depends upon its
shape, colour and size and on the impressions made by indi­
vidual teeth. Bruising may be present not only rel ated to the
biting edges of the teeth but also in the central area of the bite
beca use of tongue thrust or suction by the assailant. This usu­
ally presents as petechial haemorrhage in the central area of
the bite. If the marks of individual teeth in their correct align­
ment can be satisfactorily identified then it is possible to say
at this stage, that the victim has suffered a human bite.
Before the bite is touched, measured or photographed it is
important to obtain evidence of saliva traces left on the skin
surface by the assailant. These swabs may be used for subse­
quent blood grouping, usually limited to the ABO system
only. A, Band/or H substances are secreted by some 75 per
cent of the population. A negative result may therefore be
due to the absence of saliva to a non-secretor status of the
biter or to poor sampling. Nowadays swabs are taken for
DNA analysis. Control swabs must be taken from swabs in
other parts of the body where no bite marks are suspected 100
and it is often convenient to use the contralateral area of the
body. A saline control swab should also be supplied. These
specimens are needed to exclude contamination by secre­
tio ns from either the victim or the investigator, who should,
of course, wear gloves throughout the procedure. The use of
stelile cotton swabs moistened with steriJe water followed by
dly swabs, the so-called double swab technique, is now rec­
ommended. 101 The importance of a standard and satisfactory
tech niqu e cannot be over emphasized.
The Detail of the Bite
Once an injury has been recognized as being caused by a
human bite and saliva samples have been taken, the details
of the mark should be recorded. A human arch mark may
be identified when four or five individual marks of adja­
cent teeth are present. If the tips of the canines in both
upper and lower arches ca n be recognized, then the inter­
canine width should be measu red. In a human bite, the
intercanine width will lie between 2. 5 and 4.5 cm.70 If the
intercanin e width is less than 3 cm, the bite will probably
have been inflicted by a child. In this context, a child refers
to any biter below the age of 12 whose anterior teeth are
either deciduous teeth or mixed dentition. The two arcades
of upper and lower teeth marks should be described and
measured in terms of their curvature and overall size; in
addition, marks left by individual teeth should be individ­
ually measured and described in an attempt to reconstruct
the type of denti tion that could have made the mark. Miss­
ing teeth or areas where no bruises have occurred must be
noted. The alignment of each bruise and the possibility of
them being made by rotated teeth should also be examined.
430 I Dentist's role in chi ld abuse and neglect
In general, the mark caused by the upper maxillary arch
is often more diffuse and wider than that produced by the
mandibular teeth. This results from the greate r surface area
of the incisal edges of the upper teeth, but also because the
maxilla is stable in relation to the skull. By contrast, the
mandible is mobil e and acts as a cutting instnlment once
the skin has been stabi lized against the upper teeth. Before
the bite mark is recorded the forensic dentist should
attempt to establish the circumstances surrou ndin g the
bite. A slowly produced bite usua lly results in stretching of
the skin over the cutting edges of the teeth and pressure or
suction from the tongue. This will be reflected in the extent
of bruising, both in the too th marks and in the surroundin g
and central areas of the bite. It is said that a bite mark
show ing these characteristics was probably produced under
amorous circumstances. However, even so-called ' love
bites' require considerable force to produce and can be
painful to the recipient. A bite mark that consists only of
tooth marks is of an aggressive nature and has probably
been delivered rapidly. This is especially so if teeth have
penetrated the sk in. The observer sho uld be wary of
attributing a particular force or intent, such as aggression,
in any particular bite mark case.
It may sometimes be possibl e to offer a general opinion
as to the timing of the bite. Too th marks that do not break
the skin usually last from several minutes to 24 hours. 102 In
cases where the skin is broken, the borders or edges wi ll last
several days depending on the thickness of the tissue. Thin­
ner areas retain the marks 10ngeL IOJ It is clearly important
that the forensic dentist be called in the early stages of the
investigation of a bite mark case. Many investigators have
recommended that bite marks be photographed at 24-hour
intervals because their detail may improve in the first 2-3
days as swelling decreases and brui sing increases. 57 In add i­
tion, the outline and shape of a bite may change because of
infection, oedema and discoloratio n of the skin. Accurate
determination of the time of a laceration or brui se is diffi­
cult and complicated by individual valiation.
The history of a bruise can be correlated to sequential
colour changes, but is frau ght with difficulty, and only a
gen eral comment should be made on these matters . The
current position has been reviewed recently. 104.105 Quite
different appearam:es will result if the bite was inflicted
immediately prior to or immediately after death . If the bite
mark was inflicted prior to death, providing that sufficient
time has elapsed to allow extravasation of blood into the
tissues, a clear bite mark may be seen that changes velY lit­
tle until the onset of putrefaction. Bites inflicted after death
rarely result in bruising. Very considerable forces indeed are
required to produce an appearance of bruising post mortem.
Recording the Bite Mark
It is important to record the injury for posterity, especially
when legal proceedin gs may subsequen tly occur. In the
majority of cases, good-qu ality co lour and bl ack and white
photographs will be used. Some fo rensic dentists argue that
the photography should be carried out by a professional
police photographer under the direction of the dentist, so
that questions in court regarding scale and reproduction,
colour balance and other photographic questions may be
answered with authority. However, exp erienced forensic
dentists prefer to take their own photographs in a standard­
ized manner as better results are often obtained. A 35-mm
camera with a focal len gth lens of abo ut 100 mm and a ring
flash and/or side flash is the most commonly used equip­
ment and the techniques involved have been described in
detail. lOG Whatever equipmen t is used, photograp hs should
include low-power distance shots of the bite mark and sur­
rounding tissues, which allow for orientation. A standard
colour chart is often in clud ed along with the subject when
colour film is used. In the case of black and white, ultra­
violet filters may be used, which may bring out details that
are not obvious when viewed by conventional lighting. 107
Ultraviolet light requires a grea t deal of careful and expe­
rien ced interpretation as injuries made many years before
the bite may be imaged using this technique a nd may con­
flict with the interpretation of a recent injUly.
Whatever method of photography is used, photographs of
the injury must be taken with a standard, rigid, L-shaped
centimetre scale, arranged as close as possible to the injury,
but not overlapping it. Because injuries are often on curved
surfaces, it may be necessary to reposition the sca le a num­
ber of times in relation to different parts of the bite mark. In
principle, the scale should be arranged in the same plan e as
the portion of the injury being photographed and the plane
of film of the camera should also be parallel to both scale
and the portion of the injury being photographed. Inevitably,
on a sharply curved surface, a number of photographs will
be required around the curvature of the injury (Fig. 22.6).
These procedures are essential to avoid introducing photo­
graphic distortion into a situation that is already distorted by
Figure 22.6 Bite marks are often on cu rved su rfaces, such as
the breast, and each portion of the injury should be photographed
individually. The mark on the right is from the upper arch with
only three incisors present.
---

curved sUifaces, oedema , positional distOliion and possibly
putrefaction and post-mOliem changes.
It is advisable to re-photograph marks on living victims
every 24 hours for a number of days. One-to-on e en large­
ments should be made of both colour and black and white
photographs for measurement and comparison purposes. In
addition, larger prints at twice or four times the normal size
may be prepared for demonstrations in court. lOB On co mple­
tion of photography, the marks should be examin ed care­
fully thro ugh the hand lens to determine whether there are
foreign bodies in the depth of th e ma rks. These may include
fragments of tooth, fractured from the assailant at the time
of the attack, and they should be removed carefu lly and
stored as evidence. 109 It is well recognized that de terioratio n
may occur in a bite mark. This may be caused by dynamic
activity during the bite, curvature and distortion of tis­
sues,JlO chan ge in the tissues after the bite or distortion
introdu ced during photography. Care should be taken to
minimi ze and record any distortion; there are methods
avail a ble to correct the image to some extent. III Some
odontologists recommend digital enhancement of bite mark
images, 11 2 but this shou ld be done to a minimal ex tent, if at
all; the method has been criticized in so me courts.
Impressions of the Bite
Most bite marks in human tissue do not leave a three­
dimensional imprint, but are visible only by way of the
bruising in the und erlying tissues. Impressio ns are valueless
in such cases. If the bite is deeper, and particularly when the
skin has been penetrated, an impression may be useful and
shou ld be taken in a reco mmended si licon rubber-based
dental impression material. It is advisable to back the
impression with a rigid material, such as plaster of Paris, to
minimi ze distortion on removal. Subsequent models should
be cast in high-quality dental stone. They may be useful for
measurement purposes and study models prepared from pos ­
sible suspects may be compared with the cast of the injUiy.
Excision of Skin Samples
In fatal NAI, further information concern in g a bite injUly may
be acquired by excising th e wound and studying it histologic­
ally. Sections can be stain ed for iron and other blood break­
down produ cts an d examined for changes in the elastic and
collagen fibres; they may also show impacted plaque from the
biter's teeth (Fig. 22 .7). Classical features of a bite mark injUly
are abrasion or penetration of the epithelium, compression
and dis tOliion of underlying collagen fibres, oedematous
spaces and extravasation of elythrocytes from the blood
vessels. 11J Attempts have been made to detenlline the age
of injuries by microscopic or biochemical examination. lo5
Enzyme histochemistry using serotonin and histamine, along
with other histochemical methods, have been used. 114
Bite marks I 431
Figure 22.7 Histological sections through a bite indi cate loss of
epithelium, damage to connective tissue and, in this case,
impaction of denta l plaque into the wound (arrow).
It has been suggested that bite marks in human skin in a
victim who does not survive should be excised and preserved
as evidence of the injury. There is a potential problem of
shrinkage and distortion of such sa mples, and a recent
study1 15 has shown that both contraction and expan sion of
bite mark specimens ca n occur using the standard methods
of fix ation and support. It seems, therefore, that standard
techniques for stora ge and preservation of bite mark samples
will not produce reliable dimensional accuracy.
The Suspect
Once a suspect or suspects have been apprehended, it will
be necessalY to examine their dentitions, to take impres­
sions a nd make dental casts . In England, the examination
of suspects is governed by the Police and Criminal Evi­
dence Act. Written consent of the procedure should be
acquired from the susp ect and witnessed, preferably in the
presence of the suspect's solicitor. In Scotland, an impres­
sion made by virtue of a s heriff warrant would be admiss i­
ble evidence. It is advisab le to take a full dental history,
especially if the suspect may have received dental trea t­
ment prior to and following the biting incident. Colour and
black and white photograph s should be taken or ordered of
the extra- and intraoral appearance of the suspec t.
It may sometimes be useful to photograph the suspect with
a scale in place show ing the maximum opening between the
incisor teeth. Suitable blood and saliva samples should be
taken from each suspect when a saliva swab has been taken
from the victim. Comprehensive intraoral examinations
shou ld be undertaken, charting the teeth present and any dam­
age, misalignment or fractures. Teeth tender to biting forces
and any abnormal chewing activity should also be noted.
Impressions in dental alginate should be taken of both
upper and lower arches, and the occlusal relationship
between maxilla and mandible must be record ed using a
suitable technique. These impressions llluSt be attested,
labelled and cast in high-quality dental ston e as quickly as
432 I Dentist's role in child ab use and neglect
possible. The technician making the castings must record
the details of the procedure. The base of the study models
must be marked with the individual's name and also signed
and dated by the technician. Some dentists invite a suspect
to produce a samp le bite into a wax recording medium at
the time of the examination.
Comparison of Evidence from the Bite
and the Suspect
The final comparison between the details of the bite mark
injury and the dentition of a possible suspect is not a simple
matter. Some a uthorities claim that transparent overlays of
the biting edges of a suspect's teeth can be produced and
directly compared with the bite mark. An accurate fit then
warrants an opinion of guilt by a putative biter. However,
transparencies produced by whatever meth od rarely, if ever,
exactly fit the size and shape of a human bite mark. This is
because of the distortion factors, which have been discussed
earlier in this chapter. There is a danger in ass uming that bite
marks are equivalent to fingerprints and that a reasona ble
number of matching points wi ll prove the id entity of the
biter. A more satisfactOlY ap proach is to describe the bite
mark in detai l before looking at the details of a suspect's
dentition a nd then to describe the suspect's dentition and
to note the points of correspondence or deviation. The bite
mark inju ry and the dentition of the suspects are then exam­
ined together, and an iterative process should enable the
observer to determine whether or not the details of the injury
can be explained satisfactorily by relating th em to the den­
tition of the suspect. If they can be so related, the forensic
dentist shou ld base the strength of his or her final opinion on
the si milarity of pattern between the injury and the dentition
a nd should tl1en finally illustrate a nd confirn1 his or her
opinion using transparent overl ays produced at the sa me
scale. The non-scientific nature of this methodology has
been emphasized, but it can work sat isfactorily in most
cases. 98 Unusual features of bo th the mark and a suspect's
dentition that would tend either to elimin ate or incriminate
a suspect in the case should be noted. Thus, a single feature,
such as a missing, rotated or fractured tooth leaving a com­
parable mark in the .bite, may be a much more useful pointer
than any number of matching features in an arch of normal
size and arrange ment (Fig. 22.6). Comparisons of dentitions
and bite marks have used superimposition of colour slides,
denta l articulators and the recording of bites in suitable
media. Sophisticated scientific analysis such as micro­
co ntour mapping, scanning electron microscopy a nd reflex
microscopy have been used. 11 6 In practice, the simpler and
more obvious the method, the more likely the jury memb ers
are to understand and appreciate the significance. The
current vogue is to use computer analysis of the bite mark
photograph and to produce transparent overlays digitally
from sca nned images of the dentition.III, ll l These methods
are sa id to be able to correct for common photographic
distOIiion and size discrepancies, eliminate examiner subjec­
tivity, provide better control of image visualization and pro­
v ide standardization of procedures. However, it should be
remembered that these imaging methods are only as good as
the original evidence and should not be used to enhance
poor-qu ali ty evidence. 11 2 The final opinion is a very serious
issue for possible suspects as well as for the victim and fam­
ily; in few areas of medical evidence is such a respons ibility
placed upon the objectivity and honesty of the expert. No
jury should be led to believe that bite mark analysis is sim­
ple and accurate and the expert must clearly indicate the
pros and cons of each case.
REFERENCES
Knigh t B. The hi story of chil d abu se. Forensic Sci I11t 1986;
30: 13 5- 41.
2 Kempe CH, Silverman FN, Steele BF, Droegemuell er W,
Silver HK. The battered child syndrome. lAMA 1962;
181:17- 24.
3 Welbu ry RR. Ch ild physical abuse. In Andreason JO,
Andreason FM reds) Textbook alld C%ur Atlas of Traumatic
Injuries to the Teeth, 3rd edn. Copenha gen: Munksgaard,
1994.
4 Mea do w RC. The ABC of Child Abuse. London: BMJ
Publis hing Group, 1997.
5 Cameron JM , J oh nso n HR , Camps FE. The battered ch ild
sy ndrome. Med Sci Law 1966 ; 6: 2-2 1.
6 Wagner GN. Bitemark identification in chil d ab us e cases.
Pediatr Dent 1906; 8 :96 -100.
7 Needl eman HL Orofacial trauma in child abuse: types,
prevalence, manageme nt, and the dental profession 's
in volvement. Pediatr Dent 1986; 8:71-80.
8 Wei bury RR, Murphy JM. The dental practitioner's role in
protecting chi ldren from abuse. 3. Reporting a nd subsequent
management of abuse. Br Dent 1 1998; 184:115 - 19.
9 Kirschner RH, Stein RJ. The mistaken diagnosis of chi ld
abuse. A form of medical abuse 7 Am 1 Dis Child 1985;
139 :873-5.
10 John so n CF. Inflicted inju ry ve rsus accid e nta l injury. Pediatr
Clill N Am 1990 ; 37:791-814.
11 Whittaker OK. The principles of forens ic denti stry. 2. Non­
accidental injury, bite marks and arc haeo logy. Dent Upda te
1990; 17:386-90.
12 Fonseca MA, Feigal RJ. Ten Bensel RW. Dental aspects of
1248 cases of child maltreatment on fi le at a major country
hospital. Paediatr Dent 199 2; 14: 152-7.
13 Rasm ussen P. Tannlegens forhoJd tilsykdo m hos barn,
barnhemish andling og omsorgssvikt. In Odontoiogi.
Copenhagen : Munksgaard Publishers, 1992, pp. 81-92.
14 Speig ht N. Bruises and burns: accidental or non -accid enta l?
In Accidents and Emergencies in Childhood. London: Royal
CoJl ege of Phys ic ians of London , 1992, pp. 97 -103 .
15 Welbu ry RR, Murphy JM. The denta l prac titi oner's ro le in
protecting chil dre n from abuse. 2. The orofacia l sig ns of
abuse. Br Dent 1 1998; 184:61 - 5.
16 Simpson K. A Doctor's Guide to Court. London : Butterworth,
1967.
17 Creighton SJ, Gallagh er B. Child Abuse Deaths: Informatio n
Briefing, No.5 (revised). London: The National Society for the
Prevention of Cru e lty to Children, 1988.
18 Schm itt BD. Physical abuse. Specifics of clinical diagnosis.
Paediatr Dent 1986; 8:83.

19 Creighton SJ. The incidence of child abuse and neglect
(Chapter 3). In Browne K, Davies C, Stratton P (eds) Early
Prediction and Prevention of Child Abuse. Chichester:
John Wiley Et Sons, 1988.
20 Golder M. Non -accidental injury in children. Dellt Update
1995; 22:75-80.
21 Fearn 1. Dental aspects of child abuse (Letter). Aust Dellt J 1987 ;
32:227.
22 Cappelleri JC, Eckenrode J, Powers .fL. The epidemiology of
child abuse: findings from the second National Incidence and
Prevalence Study of Child Abuse and Neglect. Am J Publ Hlth
199 3 ; 83: 1622.
23 Hobson P. Child abuse: we can help (Letter). Br Dellt J 1987;
162:53.
24 Child protection in ethics ill dentistry. BDA Advisory Semices,
1995, pp. 45-6.
25 Department of Health. Working Together under the Children
Act (1989), 48-49, London, HMSO, 1991.
26 Depart ment of Health. Children alld Youllg People on Child
Protection Registers, Year Ending Jist March 1995, England.
London: Gov ernment Statistical Service, 1995, p. 19.
27 Macintyre DR, Jones GM, Pinkney RCN. The rol e of the dental
practitioner in the management of non-accidental injury to
children. Br Dellt J 1986; 161:108-10.
28 Kittle PE, Richardson DS. Parker JW. Examining for child
abuse and child neglect. Pediatr Dellt 1986 ; 8:80-2.
29 Casamassimo PS. Child sexual abuse and the paediatric
dentist. Paediatr Dent 1986; 8: 102-6.
30 Anasavice K1. Treatment regimes in preventive and
resto rative dentistry. J Am Dent Assoc 1995 ; 126 :727-43.
31 Ramos-Gomez F. Rothman D, Blain S. Knowledge and
attitudes among California dental ca re providers regarding
chi ld abuse and negl ect. J Am Dent Assoc 1998; 129:340-8.
32 Beck JD , Weintraub JA, Disney JA. University of North
Carolina. Caries Risk Assessment Study. Community Dellt Oral
Epidemiol 1992; 20:313-21.
33 Jessee SA. Physical manifestations of child abuse to the head.
face and mouth: a hospital survey. ASDC J Dent Child 1995;
62:245-9.
34 Rupp RP. Conditions to be consid ered in the differential
diagnosis of child abuse and neglect. Gen Dellt 1998;
46:96-100.
35 Jessee SA. Risk factors as determinants of dental neglect in
children. ASDC J Den. Child 1998; 65:17-20.
36 Kellogg N and the Committee on Child Abuse and Neglect.
Oral and dental aspects of child abuse and neglect. Pediatrics
200 5; 116: 1565-8.
37 Sanders B, Brady FA, Johnson R. Injuries. In Saunders B (ed.)
Pediatric Oml and Ma.xilloJacia1 Surgery. St Louis: Mosby,
1979, pp. 330-400.
38 Fotana VJ, Dqnovan MD, Wong RJ. The maltreatment
syndrome in children. N Eng I J iVIed 1963; 269:1389-94.
39 Skinner AE, Castle RL. A Retrospecti1Je Study. London:
NSPCC, 1969.
40 Willging JP, Bower CM, Cotton RT. Physi cal abuse of
children. A retrospective review and an otolaryngology
perspective. Arch OtolarYl1gol - Head Neck Surg 1992;
118 :584-90.
41 Mehta MN. Physical abuse of abandoned ch ildren in India.
Child Abuse Neg11982; 6:171-5.
42 Becker DB. Needleman HL , Kotelchuck M. Child abuse and
dentistry: orofacial trauma and its recognition by dentists.
J Am Dent Assoc 1978; 97 :24 -8.
43 Feldman KW. Patterned abusiv e bruises of the buttocks and
pinn a. Pediatrics 1992; 90:633-6.
44 Wilson EF. Estimation of the age of cutaneous contusion in
child abuse. Pediatrics 1977; 60:750-2.
References I 433
45 Lenosky EF, Hunter KA. Specific patterns of inflicted burn
injuries. J Tra uma 1977; 17:842-6.
46 Kumar P. Child abuse by thermal injury - a retrospective
survey. Bums 1984; 10:344-8.
47 Prescott PRo Hair dlyers and non-accidental injury. Pediatrics
1990; 86:692-7.
48 Hobbs CJ. Skull fracture and the diagnosi s of abuse. Arch
Dis Child 1984; 59:246-52.
49 Worlock p. Stower M, Barbor P. Patterns of fractures in
accidental and non -accidental injUly in cbildren: a
comparative study. Br Med J 1986 ; 293: 100-2.
50 Kaban LB, Mulliken JB, Murray JE. Facial fractures in
children. Plost Recollstr Surg 1977; 59: 15-20.
51 Kvaal 51. Oral injuries in cases of child abuse. Rev Beige [,vIed
Dentaire 1993; 48:49-53.
52 Carrotte pv. An unusual case of child abuse. Br Dent J 1990;
168 :444-5.
53 Ablin DS. Reinhart MA. Esoph agea l perforation by a tooth in
child abuse. Pedia tr Radiol 1992; 22:339-41.
54 Sperber N. The dual responsibility of dentistry in child abuse.
Mt J Orthod 1981; 19:21-8.
55 Schuman NJ. Child abuse and the dental practitioner:
discu ssio n and case rep0l1s. Quintessence 111t 1987;
18:619-22.
56 Troutman KC. Non-consensu al touching: your input needed.
Am Acad Pediatr Dent 1987; 22:3.
57 MacDonald DG. Bite mark recognition an d interpretation.
J Forens Sci Soc 1974; 14:299-233.
58 Hod son JJ. Forensic odontology in cases of battered baby
syndrome. I Den. Res 1971; 50:6 56.
59 Sims BG, Grant JH, Cameron JM. Bite marks in the ' battered
baby' sy ndrom e. Med Sci Law 1973; 13:207-10.
60 Marr JS, Beck AM. Lugo JA. An epidemiological study of the
human bite. Public Health Rep 1979; 94:514-21.
61 Rawson RD, Koot A, Martin C et al. In cidence of bite marks in
a selected juvenil e population: a preliminary report. J Foren s
Sci 1984; 29:254-9.
62 Liddiard P. An analysis of bite mark cases in foren sic dental
practice, with reference to dental anomalies affecting case
outco mes. MS c, University of Wales, 2002.
63 Val e GL, Nogu ch i IT. Anatomical ciistribution of human bites
in a series of 67 cases. J Forens Sci 1983; 28:61-9.
64 Harvey W. Denf'al Identification and Forensic Odontology.
London: Henry Kinpton, 1976. p. 91.
65 Lowry TMcC. The surgical treatm ent of human bites. Al1nals
Surg 1936 ; 104: 1103-6.
66 Speirs RF. Prevention of human bite infection. Surg Gynec
Obstet 1941 ; 72:619.
67 Sims AP. Non-accidental injury in the child presenting as a
suspected fracture of the zygo matic arch. Br Dent J 1985;
158 :292-3.
68 Liston PN, Tong DC, Firth NA, Ki ese r JA. Bite injuries:
pathophysiology, forensic analysi s and management.
N Zealand Dent J 2001; 97:58-63.
69 Pretty IA. Sweet D. Anatomical lo catio n of bite marks and
associated findings in 101 cases from the United States.
J Forens Sci 2002; 45:812-14.
70 Simper RD, Whittaker OK. An analysis of bite marks
presenting in two centres in the United Kingdo m. J Dent Res
1994; 7J :791.
7J Whittaker OK. Aitken M, Burfitt E, Sibert JR. Assessing bite
marks in children: wo rking with a forensic dentist.
Ambulatory Child Hlth 1997; 3:225.
72 Levine U. Bite mark ev idence. Dent Ciin North Am 1977;
21: 145- 58.
7J Whittaker DK. BrickJey MR. Evans L. A comparison of the
ability of experts and non-experts to different iate between
434 I Dentist's role in child abuse and neglect
adult and child human bite marks usin g receiver operating
characteristic (ROC) ana lysis. Forflls Sc i Tilt 1998; 92 : 11-20.
74 Ky dd WL. Quantitati ve analysis of fo rces o n the tongue.
J Dellt Res 1956 ; 35:17 1- 4.
75 Harvey W, Millington PF, Evans , JH et a l. Bite ma rk s - the
clinical picture; physica l fea tures of skin a nd tong ue.
Sta nd a rd and scan nin g d ec o'on m icrosco py. mt MicroJorlll
J Leg Med 1973; 8:3.
76 Baker MD, Moore SE. Human bites in chi ldren. A six -yea r
expe rience. 1987; 141: 1285-90.
77 Leung AK, Robson WL. Human bites in chi ldren. Pedilltr
fJ-nerg Cllre 1992; 8:255-7.
78 Levine LJ. The solution of a battered ch ild hom icide by dental
ev id ence : report of case. J Am Dellt Assoc 1973; 87:1234-6.
79 Walllick AJ, Biedrzycki L, Russanow C. Not all bite marks are
associated with abuse, sexual activities or h om icid es : a case
study of a self-inflicted bite mark. J Fo rflls Sci 1987; 32:788-92.
80 And e rson W R, Hudson RP. Self- infli c ted bite marks in
battered child syndrome. Forell s Sci 1976; 7 :7 1-4.
81 Ca lkins CM, Bensard DD . Patrick DA. Ka rre r FIVI. Life­
threa tening dog attacks: a devas tating comb in at ion of
penetrati ng a nd blunt injuri es. J Pedilltl· Su rg 200 1:
36 :1115-17.
82 Gandhi RR. Liebman IVIA. Stafford BL, Stafford pw. Dog bite
injuri es in c hildren: a preliminary survey. AIlI Surg 1999;
65:863-4.
83 App lega te JA, Walhout MF. Childh oo d risks from the ferret.
J Emerg Med 1998; 16:425-7.
84 Triratana 1. Bite marks. J Dent Assoc Thllilllilil 1970;
20:259 -7 2.
85 Class RT, Jordan FB, Andrews EE. Multipl e animal bite
wounds: a case report. J forells Sci 19 75; 20: 305- 14.
86 Rhe ul J , Bratz ke H, Fedderse n-Perterso n DU et al. Dea th
ca used by 'attack' dog bites, Arch Kril1lillol 1998;
202: 140- 51.
87 Sacks JJ , Sinclair L, Gilchrist Jet al. Breed s of dogs in vo lved
in fa tal human attacks in the United States betwee n 1979 a nd
1998, J Am Vet Med Assoc 2000; 217:836 - 40.
88 Belardi FG, Pascoe JM, Beegle ED, Pasteurell a l11ultocicla
menin g itis in an infant following occipital dog bite, J fam
P/'Gct 1982; 778:780-2.
89 Ste inb ok P, Flodma rk 0, Scheifele Ow. Animal bites caus ing
ce ntral nervous sys tem injury in children. A repolt of three
cases. Per/ iat!· Neurosci 1985; 12:96-100.
90 Callaham ML. Treatment of cOl11mon dog bites: infect ion risk
factors. JACEP 1978; 7 :83-7.
91 Marcy SM, Infectio ns d ue to dog and ca t bites. Pediatr IIIJecr
Dis 1982; 1:35 1-6.
92 Brook I. Mic robio logy of human and anima l bite wounds in
childre n. Pediotr IIIJecr Dis J 1987; 6 :29-32.
93 Griego RD , Rosen T, Orengo IF, Wol f JE. Dog, cat and human
bites: a rev iew. JAm Acad Dermawl 1995: 33:1019- 29 .
94 FUOItes L, Me lso n E. Prima lY and recurrent herpes sim plex
infection in a paediatric nurse resul ting from a human bite
(Comment). TnJecr COlli Hasp Epidemiol 1989; 120: 295.
95 Tsoukas CM, Hadjis T, Shuster J et al. Lack of transm iss io n
of HIV throug h h uman bi tes and sc ratches. J ACl{uir 11I1JI1!llie
Defic S.\'lId 1988; 1 :505-7 .
96 Anony mous. To tra nsmit HIV, biter mu st have blood in the
mouth. AIDS Policy Law 1996 ; 11:5.
97 Pre tty lAo And e rso n GS, Sweet OJ. Hum a n bites and the risk
of human immunodeficiency virus transmiss ion . A m J
Forells Med Porliol 1999; 20:232-9.
98 Arheart KL, Pretty IA. Results of th e 4th ABFO Bitema rk
Workshop - 1999. Forells Sci 1111 200 I; 124: 104- 11.
99 Sc hweich P. Fleisher G. Human bite marks in children.
Pediatr Emerg Cll re 1985; 1:51-3.
100 Pereira ilil. Possibilities and limitations of sa liva tests in
forensic odonto logy. Br Dellt J 1971; 130:161.
101 Sweet 0, Lorente M, Lorente JA et al. An improved meth od
to recover sa li va from hu man skin: th e double swab
tec hnique. ) Forells Sci 1997: 42:320- 2.
102 Dinkel HI. th e use o f bite mark ev iden ce as a n in vesti gative
a id. J Forells Sci 1974; 19:535 -47.
103 Sper be r ND . Bite ma rk s, o ral and racial injuries - h arb ingers
of seve re c hild abuse 7 Pacdiatriciall 1989 ; 16: 207 - 11.
104 Dail ey JC. Bowers CM. Ageing of bite marks: a lite rature
revi ew. J Forell5 Sci 1997; 42:792-5.
105 Van ez is P. Inte rpreting bruises at necropsy. J Oill Pllthol
2000; 54:348-55.
106 Summers R. Lewi n D. Forensic dental photography. In Cla rk
DH (ed.) Practica l forellsic Odoiliology. Oxrord: Wright,
J 992, p. 188.
107 Krauss TC, Warl e n Sc. The forensic science use of re fl ec ti ve
ultravi o let photography. J forells Sci 1985; 30:262-8.
108 W h ittaker OK. MacDona ld DG. A Colour Atlas oj Forell sic
DentiSTry. London: Wolfe Publications, 1989, p. 108.
109 Whit taker DK. Fo rensic dentistry in the ident ificatio n of
victims a nd assaila nts. J Oill Forells IVIed 1995; 2: 145 -5/.
110 Sheasby DR, Mac Donald DG. A forensic classification of
distortion in h uman bite marks. fore lls Sci II/rem 200 1;
122:75-8.
111 Bowers Cilil. J o ha nse n RJ. Dig ital analysis of bite marks and
human id e ntifi cat ion. Dfllr Clill NAill 2001; 45:327-4 2.
112 Kara za lu s KP. Palm bach IT, Lee He. Digital enhancement of
subquality bite mark photographs. J forells Sci 2001;
46:954 - 58.
113 Millington PF. Hi stological stucii es of skin carrying bite
marks. J Forms Sc i Soc 1973: 14:239-40,
114 Raekallio J. Histological estimation of the age of injuries, In
Perper JA, Wec ht CH (eds) Microscopic Dillgnosis ill Forensic
Parliolog.\'. Springfield, IL: Cha rl es C Thomas, 1980, pp. 3-16.
115 Rothwell BR , Thie n AV Analys is of d isto rtion in preserveci
bite mark sk in. J Forel1s Sci 200 1; 46:573.
116 Lightellll AJ, Coetzee WJC, Va ll Ni eke rk PJ, The identifica tion
o f bite marks us ing th e re fl ex mi c ro sco pe. J Forms
Odolltosrollliltol 1987; 5: 1- 8.
117 Sweet 0, Parhar 1Vl. Wood RE. Computer- based production
of bite mark comparison overlays. J Forel1s Sc i 1998;
43: 1050-5.
-_.
- - - -- -

PAEDIATRIC DENTAL IDENTIFICATION
G Howard Moody
Introduction 435
Comparison 435
Facial reconstruction and dental profiling 440
II'JTRODUCTION
radiographs, with ante-mortem dental reco rds to
Accurate identification of the deceased by the relevant
authorities is a human right enshrined in Article 6 of the
United Nations Declaration of Human Rights (1948) - as
this, for most people, will be crucial in coming to terms with
the loss of a relative or friend, for the resolution of legal or
insurance disputes and for attaining 'closure', enabling the
mourning process and due ritual to proceed.
The identification of individuals from their teeth is not
a recent development. In AD49, Agrippina, mother of the
future Roman Emperor Nero, ordered the murder of Lollia
Paulina, whom she believed wanted to become the wife of
Claudius, as this would have prevented Nero from becom­
ing Emperor. On being presented with Paulina 's head,
Agrippina, to confirm its identity, 'op ened the mouth with
her own hands and inspected the teeth which had certain
peculiarities'.J In mediaeval times, illiterate apprentices sig­
nified thei r bondage to a master by biting into a piece of
wax, leaving a unique mark and becoming 'indentured'.
Worldwide, the task of identification is the responsibility of
either the judiciary or the police; in Scandinavia, the police
are responsible,2 while in the USA id entification is the role
of the medical examiner. J In England, identification is the
responsibility of the coron er and in Scotland of the procura­
tor fiscal. 4 The forensic dentist (odontologist) often has to
provide these legal authorities with an opinion, assisting the
identification using the techniques of comparison or profiling.
• Comparison: the procedure involved is the comparison
of post-mortem dental chartings, including
I CHAPTER 23 I
Age estimation 441
References 444
corroborate or deny an identity. This is the most
common procedure. 5
• Profiling: the process whereby a forensic dentist,
in co llaboration with colleagues, who may include
forensic pathologists, anthropologists, archaeologists
and biochemists, attempts to 'reconstruct' from human
remains the age, sex, height and pOSSibly the ethnic
origin of the deceased with a view to creating a
possible liken ess that may cause so mebo dy to come
forward and offer a possible id entification when the
deceas ed is completely unknown to the authorities 6
COMPARISON
While generally regarded as a common method of identi­
fication, the number of persons, which includes children,
in the UK who are identified by their dental characteristics
in anyone year is not known. There is no central data bank
holding this information, and there is no evidence to
suggest that all forensic pathologists refer evelY dental
identification to a dental surgeon, let alo ne a forensic den­
tist. Corroboration involves the comparison of ante- and
post-mortem records; the certainty of the identification is
related to the degree of concordance between the two
sets of records and whetl1er any discrepancies that occur are
explicable. 7 To be va lid , comparative identification requires
identifiable dental characteristics, the accurate recording of
these characteristics and accurate ante- mortem data .
436 I Paediatric dental identification

Dental Characteristics
Dental characteristics are understood to be those features
that contribute to a unique charting (Fig. 23.1). Such fea­
tures include decayed, missing, or restored teeth and fixed
and removable prostheses (dentures). Restorations range
from small white fillings through large amalgam and porce­
lain restorations, crowns, fillings, root canal fillings and
implants. Given the complete adult dentition of 32 teeth,
each with five surfaces, the wide range of dental materials,
design and construction variations available with respect to
prostheses and the variations in root patterns discernible on
radiographs, the number of permutations is enormous.
The statistical probability of concordant points on ante­
and post-mortem records has been calculated for extrac­
tions,a amalgam restorations,9 root morphologiO and digital
radiographic scanning. ll ,1 2 A single unique dental feature
may, on occasion, suffice,I3 but this is rare, and the number
of concordant points that are considered sufficient to collec­
tively ascertain an identification remains controversial. 14
[n children, however, the situation is very different. In
much of the industrialized world, dental decay is in sharp
decline with a corresponding reduction in the number of
decayed, missing and filled teeth (dmfl. The effect of this is
that the dentition of individual children is becoming clini­
cally less characteristic (Fig. 23.2).
Dental disease, as measured by the DMF, decreased in all
social classes in the UK during the period 1978-98. 15 In
1996, Kas te et al 16 reported the assessment of dental caries in
the USA as part of the Third National Health and Nutrition
Examination Survey (1998-91). This survey yielded esti­
mates for over 58 million children aged 1-17 years. It was
found that, among infants aged 12-23 months, only 0.8 per
cent scored positively for early childhood caries while 62.1
per cent of children aged 2-9 years were caries-free in their
deciduous dentition. One-quarter of the children and adoles­
cents aged 5-17 years with at least one permanent tooth
accounted for about 80 per cent of the caries identified in
permanent teeth. White et al 17 also describes how, in the
USA, the number of children free from dental caries
increased dramatically between 1963 and 1987. By 1987,

Figure 23.1 Maxilla retrieved from a shallow grave. Whilst not

from a child, the teeth clearly demonstrate their value for the
purposes of identifiCation. An ante-mortem chart of this dentition
will demonstrate the appropriate missing and filled teeth. Whilst
the upper left canine may have been extracted before death
(there is loss of part of the bucca l plate consistent with a forceps
extraction) the centra l incisors may either have been extracted or
lost post mortem - possibly shaken out by urban foxes once the
periodontal membrane had sufficiently decayed, rendering the
teeth loose. The presence of the incisors in an ante-mortem
charting would not necessarily invalidate the identification.
Relying so/ely on computer matching is not advisable. Note also Figure 23.2 A caries- and restoration-free complete deciduous
that remnants of the gingiva remain, suggesting that this is not dentition: (a) maxilla, (b) mandible. There are no characteristic
an 'old' burial. Estimation of the time since burial is subjective, clinical features enabling identification. It is unlikely that there
being influenced by soil temperature, bacteria in the soil and pH. will be ante-mortem radiographs since the child is sign-free
These variables are to some extent interdependent. and thus not clinically justified.

---

Figure 23.3 Fragment of a maxilla. It is extremely improbable
that two persons out of a (known) small populat ion of 270
persons wo uld both have a denta l charting correspo nding
in every respect to these teeth and their restorations.
almost half of school children between the ages of 5 and
17 years were caries-free and , among those with caries, the
number of teeth affected had also declined. Flinck et ai, 18 in
Sweden, and Lawrence and Sheiham, 19 in Brazil, report sim­
ila r findi ngs.
Following the Lockerbie air disaster in 1988, 208 of the
273 victims of were identitled solely, or in conjunction with
other methods, by forensic dentists (Fig. 23.3). 20 The victims
were mainly adult. A very different result may be envisaged
if a similar major air disaster were to happen now, but
involving numerous children of the same sex, of similar
age, and who had been exposed throughout their lives to a
non-cariogenic diet, fluoridated water supplies and flu orid e
toothpaste. The observed differences between the individual
dentitions might be minimal and dental records indicating
only satisfactory 'check-ups' would be unhelpful. In this sit­
uation, identification would be achieved by comparison of
DNA with DNA from tlrst-degree relatives. Dental pulp is an
excellent source of DNA, being protected from heat and
trauma by dentine, bone and the facia l muscles as well as
offering a minimal risk of contamination. A useful adjunct
to DNA comparison in this situation is the compa riso n of
ante- and post-mortem dental radiographs; rad iographs
taken at the time of a 'check-up ' may reveal characteristic
root and crown patterns (Fig. 23.4) 2 1
Ante-mortem Procedures and Dental Records
Ante-mortem procedures require the retrieval by the police
of dental records, including radiographs, study models and
clinical photographs, and rely on the clarity and accuracy
of those records.
The successfu l retrieva l of dental reco rds depends upon
diligent police work. In a 'closed' situation such as an air
disaster, the passenger list can be obta ined, relatives iden­
titled and dental records sought. However, an individual
-- ~ -
Comparison I 437
Figure 23.4 Root canal fillings and root cu rvatures identified in
an te-mortem radiographs can be matched against post-mortem
findings. An experienced dental radiographer ca n adjust the
position of th e skull or ja ws to the position in which the ante­
mortem radiographs are most likely to have been taken. This
fa cilitates th e id en tifica tion of not on ly correspo nd ence of
restorations but also root and ca nce ll ous bone morphology.
may visit several dentists over a lifetime. Multiple records
may exist each recording only the work undertaken by one
dentist. These will not necessarily include a record of pre­
vious treatment.
Ret rieval of dental records of victims of an 'open' disas­
ter, such as a fire in a theatre or railw ay station, may take
much longer since there will be no record of who was pres­
ent. Police may have to await contact from anxious rela­
tives enquiring about a person who has not returned home.
The helpline telephone numbers issued by the police on
such occasions help serve this purpose as their opera tors
are trained to seek such information. In the case of the dis­
covery of a body of a long-deceased individu al, identifica­
tion will entail the retrieval of dental records of 'missing
persons'; this may take many months.
On receipt of the dental records, the forensic dentist cre­
ates a single dental chart on a form identical to that being
used for the post-mortem charting (Fig. 23.5). This chart
may be created manually or electronically. In theo ry, this
should be straightforward, but in reality it is frequently very
time-consuming and prone to errors related to the different
methods of ante-mOliem chartin g; Ahlberg 22 records 40 dif­
ferent notations in use throughout the world . Incomplete,
illegible and in accurate records compound the problem.
While some states in the USA make legal provision for
comprehensive charting of the teeth at dental consultation,
there is no such requirement in the UK, where only work to
be carried out need be charted.2J Records frequently do not
include work carried out by a previous dentist and are thus
incomplete. Ireland et al 24 have shown that even basic infor­
mation is often unrecorded, with only 7 per cent of carious
cavities recorded and only 2 per ce nt of general dental prac­
titioners recording the occlusio n. Whatever system is used,

438 I Paediatric dental identification

IDEN TAL Re f No POSTMORTEM Ref No is known as 'toothprints'. This procedure, devised and
SURNAM E FO RENAMES 0. 0 .8 developed by Tesini et aJ25.26 in conjunction with Kerr
AD D RESS Dental, utilizes an arch-shaped thermopl astic warm wafer
Tela,
into which the child bites for 50 seconds. After 2-3 min­
() ~Hv ) D:.~) (Ill})
100
~
~ » 10:' ,I
"'"
". utes' cooling, the wafer is sea led into a pl astic bag, thereby
preserving both a very accu rate impression of the teeth and
DC(~ 0 C~ 0) (v q~ ~ ~ ~D'O (0 (J)~) CD (~··D
a suffi cient numb er of shed ep ithelial cells for fut ure DNA
analysis. Good a nalytica l results after 3 yea rs appear to be
~D (0 0 CDdD G achieved if the wafer is a ll owed to dry prior to sea ling in
the zipl ock bag 27 Longitudinal trials will be necessary to
,-

Figure 23.5 Asimple plastic lam inated dental char t fo r use
at the locu s. An indeli ble pencil/pen should be used to prevent
damp and dirty conditions fr om spoiling 'first sourc e' evidence.
This record should be retained even after a 'cl ea n' copy for filin g,
and later ante -mortem matching, has been created. Close
examin ation , includin g rad iographs of the teeth and ja ws, should
be undertaken in th e labora tory once the remains have been
removed from the locus. Any laboratory findings, includ ing
radiograp hs, may be added to the clean post-mortem record.
-_
~
.........
,
'­
. -~ - ~,
- ....
Figure 23.6 An almost illeg ible ante-mortem dental chart.
illegib ility may m-ake interpreta tion almost imposs ible (Fi g.
23.6). Co mputerized records should red uce this problem.
Among the common ante-mo rtem charting erro rs are
transposit ion of left and right, misreco rdin g of a bu cca lly
extended restoratio n as a ling ual extens ion, a tlrst bicuspid
charted as a second bi cuspid tooth, and first an d sec ond
permanent molar teeth cha rted as seco nd and th ird molars
when on e molar is absent from the arch. Other errors
includ e adding a surface to a resto ration or incorrect
record ing of su rfaces fill ed, listing a tooth as restored when
it is not and om itting to record a unique fe ature such as a
sup ernumera ry too th or an apicectomy.7
A novel ante-mortem development designed to assist in
the identification of child victim s of fires or mass di sasters
,'A
determine how long the wafer remains a valuable source of
DN A. This is important sin ce the de nta l record in the child's
impression will beco me of decreasing value as teeth are
shed, replaced by permanent teeth or restored.
Post-mortem Procedures
The Illajority of denta l identification s are carried out on sin­
gle bodies. It is good practice for a forensic dentist to be
called to the site (locus) where the body has been fou nd since
the teeth necessary for identification may be loose or, follow­
ing a fire, brittle, and requi re expert handing to avoid dam­
age or loss th ereof. When remains are found in a shallow
grave, dogs or foxes Illay have disturbed them, resul ting in
loose teeth or dentures being displaced. A forensic dentist will
be aware of these possibili ties and ensure that the site is
meticulously inspected. If teeth or dentures are found in pl ace
within the j aws, they should be photographed ;11 situ, ca re­
full y bagged, labelled and removed from the scene only with
receipt of permi ssion from the investigating poli ce officer.
In the mortuary, a detailed dental examinati on is
usually best carried out afte r the completion of the post­
mo rtem examin ation, using a forens ic dental kit comp ris­
ing a full range of dental hand instruments, a portab le
water pick to flu sh away debris and fas t-setting adhesives,
su ch as cyan acrylate, to stab ilize tooth fragments. If den­
-
~-
..
~. tures are present, they shou ld be photographed in position.
If the deceased is beli eved to have had den tures but these
are not immediately availab le then impressions of the jaws
shou ld be taken in ord er to create plaster models upon
which a later retrieved dentu re may be fitted.
Making photographic records is importa nt. Thi s may
best be achi eved by utilizing the servi ces of police photog­
raphers under the directio n and in the presence of the
forensic dentists. However, where th is is not possible, a
35-mm single-lens reflex (SLR) digita l camera (10 mega
pi xels or greater) with a dedicated flash facili ty will nor­
mally enable the dentist to obtain satisfactory illustrations.
Accurate dental charting in the mOliualY is best achieved
with the assista nce of a dental nurse or a seco nd forensic
dentist. The author has a preference for Llsing pencil at this
stage as charts may become soiled and ink may run .
Whenever possible, the jaws should be radiographed.
Small whi te (composite) fillings may be very hard to see in
the mortuary but readily identified on radiographs, as are

--- - ----~---~-----------
- ---

Figure 23.7 Victims offire are candidates for dental
identificat ion, be ing vi sual ly unidentifiable but with
teeth in tact.
root canal fillings and unerupted teeth. If suitabl e radio­
graphic facilities are not availabl e in the mortuary, permis­
sion may be granted to remove the jaws at autopsy and take
them to an X-ray department. Permiss ion to do this may
only be given if there is gross facial disfigurement preclud­
ing visual identificatio n. Disarticul ated jaws are simple to
radiograph, multiple views may be taken to reproduce the
orientation of any ante-mortem radiogra phs and thus facil­
itate matching. If permission is granted to remove the jaws
then two criteria must be satisfied irrespective of technique:
minimal further facial disfigurement and preservation of
the apices of the tooth roots within the resection. An excel­
lent technique is described in detail by Whittaker and
MacDonald. 28 Extreme care should be exercised when remov­
ing the jaws of fire victims (Fig. 23.7). Both the police and
relatives of fire victims will want to know whether the
deceased died before or after they were engulfed by fire. If
the victim died before the fire, there often will be no soot
deposit on the palate as evidence of smoke inhalation. It is
imperative that those carrying out the autopsy do not inad­
vertently contaminate the palate with soot adheri ng to their
gloves (Fig. 23.8). It is the author's practice, having gained
access, to change gloves so as not to inadvertently carry
soot onto the palate or tongue. Once the radiographs have
been developed, post-morte m charting may be completed
on a form identical to that upon wh ich the ante-mortem
records have been transposed.
Comparison I 439
.:.~( ' /.......
1r ,-#
Figure 23.8 Fire has destroyed the anterior teeth but the extent
to which the soft tissues around the mouth protect the teeth is
evidenced by the preservation of most of a plastic denture. There
is no carbon trail across the palate. The victim died before the fire
engu lfed him.
The procedure in a mass disaster is essentially the same,
only the circumstances may necessitate working in a tem­
porary mortuary. There may also be a need to resist the
pressure from relatives and the media for fast identifica­
tion, to keep the risk of errors to a minimum.
After preparing the ante- and post-mortem records, the
two records may then be compared. Where there is com­
plete concordance or an obvious gross discrepancy, there is
little probl em. In other cases, an intelligent appreciation
and interpretation of the data is necessary; this precludes
simple 'computer matching'. In addition to the appreciation
of the possible errors already described as possibly occur­
ring in ante-mortem charting, some other pitfalls should be
considered.
When teeth are extracted, a drifting of adjacent teeth
may occur, and the resulting diastema and occasion ally
associated partial rotation of a tooth may not be record ed.
Where an ante-mol1em schematic drawing or computer­
generated icon appears on the record, the result may be a
poor representation of the actual restoration. The number
of restorations seen on post-mortem examination may be
more or less than the number indicated in the ante-mortem
chal1s. If greater, the causes may include additional work
undertaken by a 'new' dentist or a simple clerical error.
Reasons why the number of restorations may be less than
indicated in the ante-mortem records include a post-charting
emergency extraction, whi lst on vacation for examp le, a
clerical error or, in the case of highly aesthetic restorations,
a restoration that was not identified post mortem. Most
forensic dentists have had the humblin g experience of
returning to the disarticulated jaws and finding
such restorations following a closer examination , possibly
aided by ultraviolet light. However, if a filling is unam­
biguously indicated in the ante-mortem chal1s and radio­
graphs and is c1ea rl y absent post mortem, albeit with the
440 I Paediatric dental identification
tooth present, then a mismatch in identification is the most
likely exp lanation.
Human error in recording and genuine disagreement
about the identity of a tooth therefore prevent total
reliance upon 'computer matching'.
Aws 29 compared two computer programs, CAPMI
(computer-assisted post-mortem identification) and DAV1D
(disaster and victim identification), and found that both are
a useful adjunct to comparison, especi ally in quickly iden­
tifying features indicative of a mismatch, but the final
matching process requires a degree of human interpretation.
In children without fillings , the matching of radiographs
is of paramount importance. In a recent experimental eval­
uation, forensic dentists were provid ed with radiographs of
teeth with no restora tions. For the purposes of identifica­
tion, we demonstrated that root morphology and alignment
we re more usefu l than crown morphology for the purposes
of identification. 21 Studies by Wood et a l ii demonstrate the
value of digitized slices made from ante- and post-mortem
radio graphs. The slices were manipulated using a computer
graphics program and a quantitative comparison of the
alignment of normal dental anatomical landm arks was
undertaken. A perfect match was achieved in all 39 speci­
mens stud ied.
In 1999, Wood et al 12 published a study of di gital den­
tal radiographic superimposition at various stages of devel­
opment of the dentition. The technique was app li ed to 25
cases in primary dentition phase, 25 in the mixed dentition
phase and 25 in the permanent dentition phase. They found
that their technique is reliable in the cas e of both primary
and permanent dentitions but less so in the mixed dentition
situation .
Once the match ing process has been completed a repott
must be made to the appropriate legal authority concluding
with a statement of probability. The American Board of
Forensic Odontology30 identifies four categories.
1. Positive. The ante- and post-moltem data match in
sufficient detail a nd there are no irreconcilable
differences.
2. Possible. The ante- and post-m ortem data have
consistent features but, due to the quality of either the
post-m ortem remains or the ante-mortem evidence, it
is not possible t6 positively establi sh identifi cation.
3. In sufficient evidence. The avai lable evidence is
inconclusive.
4. Exclusion. The ante- and post-mortem data are
inconsistent and there are irreconci lable differences.
FACIAL RECONSTRUCTION AND DENTAL
PROFILING
In the absence of dental records of skeletonized remains,
there have been many attempts to derive information from
the skull and jaws with the aim of creating a 'profile ' of the
deceased in li fe.
Facial Reconstruction
The science of facial reconstruction is now very sophisti­
cated, using computer-guided morphometry. The interested
reader is referred to the studies of Koelmeyer,31 Perper and
Patterson,J2 Hill et al, 33 Philips and Smuts,34 Law and Bow­
ers,35 Sha h rom et al 36 and Tyrell et al. 37
While there is a strong scientific basis for morphologi­
cal accuracy, some features, such as colou r and length of
hair or colour and texture of skin, can be based only upon
probability, so that an exact likeness is not possible. How­
ever, in forensic work an 'exact' reconstlUction is not
required. An approximate likeness may stimulate some­
one' s memory and enable him or her to go to the police to
suggest a possible identity.
The direct superimposition of photographs upon skulls in
order to demonstrate a ' match' requires extreme care and
attentio n to detail, especia lly as the quality of the photograph
may not be high or recent. Where resources are meagre, how­
ever, it may be of some value. Brocklebank and Holmgren 38
assisted the Hong Kong authorities in body identification
using this method when identification was required following
gross disfigurement by fire (Figs 23.9 and 23. 10).
Simpso n,39 noting the absence of negative controls in the
Hong Kong cases, selected a skull from the collection of the
Royal College of Surgeons of Edinburgh which, on the basis
of size, morphology and dentition, was probably that of a
middle-aged male Caucasian. Full fronta l photographs of 25
white members of staff of the Edinburgh Dental Institute aged
between 23 and 60 years of age were superimposed upon the
skull using the method described by Brocklebank and Holm­
gren. It was found that in two instances a 'convincing match'
could be achieved. Despite much more soph isticated methods
now being employed, further contro lled studies are required.
Dental Profiling
Bodies fragmented by explosions or burnt beyond recog­
nition may be extremely difficult to identify, Dental
Figure 23 .9 A reassembled skull fo llow ing disintegration in a
fire. The body had been dumped in a skip and the n set alight. By
kind permission of Dr L. Brocklebank.
--
~- ---

Figure 23.10 Superimposition of a photograph of the suspected
victim upon the skull. By kind permission of Dr L. Brocklebank.
characteristics have proved to be of great value in their
identification since the teeth are the most indestructible
part of the body due to their mineralization, their partial
encasement within bone and the low turnover rate of their
constituent tissues. 4o The dental pulp may be an important
source of uncontaminated DNA.
Teeth may contribute information on age, sex, height
and ethnicity.
AGE ESTIMATION
Age estimation in children is conveniently divided into
(1) prenatal and (2) from birth to 18 years of age.
PRENATAL
In practice, it is unusual for a forensic dentist to estimate
the age of a fetus, but an accurate estimation of age may
be required for a prematurely born child when there are
grounds to suspect infanticide.
The examination of histological preparations of fetal
and perinatal specimens provides very accurate estimations
of age. 41 The data are based upon histological evidence of
premineralization sequences coupled with mineralization
and incremental patterns of enamel and dentine formation
(Fig. 23.11).
Radiographs can detect the earliest stages of calcifica­
tion in the developing primary dentition,42 provided the
Age estimation I 441
Figure 23.11 Incremental striae.
fetal material is appropriately handled. Clement and
Kossa 43 made this point succinctly in observing that calci­
fied tooth caps can be retrieved from the fetus throughout
the second and third trimesters of pregnancy. Often these
tooth caps are missing, since putrefaction liberates the
tooth caps from the forming alveoli of the jaws. This is not
surprising. The calcified part of the first permanent molar
at birth, for example, is about the size of pinhead and
weighs approximately 1.0 mg. For this reason, a complete
radiographic survey of the fetus should be made before an
autopsy in order to locate the tooth caps.
However, radiography alone is insufficient to assess cal­
cification. There is a marked time lag in the identification
of the initial stages of cusp calcification between those
examined by histological and radiographic techniques.
Histology (Fig. 23.12) is the more sensitive technique, early
mineralization being detected up to 12 weeks before it
becomes apparent on a radiograph.42
Since the primary and secondary (permanent) dentitions
develop sequentially, it is possible to estimate the age of
the child on any single defined tooth but it is much better
practice to estimate the age from several teeth and average
the result. With little routine experience in estimating age
in this fashion, forensic dentists need to be aware of possi­
ble interobserver variation, the quality of the radiographs
and the histological sections, and be cautious in their
reporting.
442 I Paediatric dental identification

Deciduous dentition

~
2yearS
5 months

in utero
~ (+/-6mOS)
o
7 months

in utero

Prenatal

~ O QwJ

~
-oooo dY1 Birth ~~
~0 .' U .

~
o

'0"1'e =
moo", '" lc
'0000= (+/- 2 mas)
,
~
' 4 years

[./- 9 m"i

Figure 23.12 Histology of a developing deciduous tooth.

~
In the author's view, dental age estimation in the peri­
and prenatal periods is not a routine procedure and should 18 months
be carried out by two forensic dentists working in close
collaboration with expert histologists and radiographers, ~~o
(+/- 3 mas)

Early childhood
initially examining the material independently but ulti­ Infancy (pre-school age)
mately presenting their findings in a joint report.
PFMP Figure 23.13

Figure 23.13 Schour and Massier charts for developing

BIRTH TO 18 YEARS OF AGE
Possibly the most important question to be answered about a
dead neonate is whether or not the child could have been
born alive. An accurate estimate of age at the time of death is
crucial to answering this question, and most authorities agree
that data derived from the developing dentition provide the
most accurate means of age estimation in a chiJd. 44 ,45
Ground sections of enamel reveal cross-striations in
enamel prisms. The intervals between these striations, meas­
uring about 16 ~~m, represent the amount of enamel formed
in each 24-hour period. 46 .47 Disturbances in cellular activity
and calcium metabolism around the time of bilih produce
a grossly accentuated striation, called the 'neonatal line' in
both the deciduous teeth and the first permanent molars. The
presence of the neonatal line there indicates that the child
had been born alive and lived beyond the neonatal period.
Using the neonatal line as a baseline, a count of the sub­
sequent striations should provide the child's age in days
until such time as the enamel ceases to form. However, while
the enamel is still forming, the leading edge is not calcified
and therefore does not show an incremental pattern.
deciduous teeth.
The number of days taken to produce the uncalcified form­
ing front is speculative and the potential error is taken into
account by giving the age in days as between 20 days either
side of the total count.
From birth to 6 months the state of development and
mineralization of the crowns and roots of the deciduous
teeth and the first permanent molars can be determined
radiographically, and the body need not be mutilated. 4s - 5o
From 6 to 30 months, age may be estimated from the
state of emergence of the deciduous teeth, provided that it
is appreciated that there is a range around the mean time of
emergence of about 2 months for the central incisors and
4 months for the lateral incisors, canines and molars. 51
Chronological and dental age do not, therefore, perfectly
coincide and, as there may be interobserver variation in
appreciating what constitutes 'emergence' into the mouth,
the radiographic techniques developed by Demirjian and
his colleagues,50,52,5] used in conjunction with Schour and
MassIer charts (Fig. 23. J3),54 are the preferred methods.

------------ ----------- -- -----------------

----- -

These charts are periodically updated by the American Den­
tal Association (available at: www.forensicdentistl),online.
org/Forensic_pages_l/ eru ption2.htm).
Recently, Maber et al 55 published an extensive and
detailed review of the accuracy of age estimation of radi­
ographic methods using developing teeth. They concluded
that the most accurate method is that of Willems, 55 using a
modified Demirjian's technique. De Salvia et a[57 published a
salutary paper in which they concluded, perhaps not surpris­
ingly, that the greatest limitation to the use of radiographic
methods to determine the dental stage of development is
operator experience. The case for more than one forensic
dentist to make independent age estimation on the same
material is implicit before conferring and submitting a joint
report to the relevant authority.
Age estimation of the living remains relatively uncom­
mon, but a recent paper by Garamendi et al 58 illustrates the
value of dentistry in assisting immigration officials. By
combining skeletal (carpus) radiographs and Demirjian's
method, they were able to place 114 Moroccan males either
above or below the age of 18 years. The results were con­
firmed by the Moroccan Embassy in Spain.
The Schour and Massier charts are self-explanatol), and,
for that reason, popular. Both deciduous and permanent
dentitions are displayed i/1 situ, including root resorption
sequences for the deciduous teeth. The drawings are life
size, enabling direct comparison of either extracted teeth or
radiographs. The chalis do not distinguish between males
and females, but Ciapparelli 59 demonstrated that the corre­
spondence is very good for males, with females on average
being aged 3-6 months younger than they actually are.
Mornstad et al GO have suggested that these 'chart com­
parison' methods may be criticized as too subjective. In an
attempt to address this concern, they made objective meas­
urements of developing teeth and correlated them with the
subjects' chronological age. The structures measured were
crown height, apex width and root length. With the aid of
a multiple regression model, a linear relationship between
some of these distances and age was shown. The method
has a 95 per cent confidence interval (CI) of about ± 2 years
around an estimated age.
In a recent assessment of age estimation methods, Liver­
sidge et al 61 detei-mined that Monstrad's method is currently
the most accurate. Since this paper was published, however,
Foti et al G2 have described equations for age estimation in
both living and dead children obtained with the help of step­
wise ascending multiple linear regression. The equations
should be applied on the basis on the number of erupted teeth
and tooth germs, which are detected on radiographs during
clinical examination and in infant skeletal remains. Foti
claims that this method permits age estimation until 20 years
of age. Beyond the age of 20 years, the most accurate method
of age estimation using teeth is not morphology but a bio­
chemical method, based upon aspartic acid racemization
in dentine. 63 This method relies on the age-dependent
non-enzymatic changes of L-form amino acids to D-form
Age estimation I 443
Figure 23.14 Ground section of tooth. Tetracycline fluorescent
bands are present, each indicative of an episode of tetracycline
medication.
amino acids, a progression which shows a linear relationship
with time.
Analysis of aspartic acid in dentine is amongst the most
reliable and accurate method to date, and work continues
to refine the method yet fUliher. G4 Recently, Yekkal et al 65
published a pilot study, which evaluated the efficiency of
high-performance liquid chromatography (HPLC) as dis­
tinct from gas chromatography. Their work suggests that
HPLC may provide more stable derivatives of aspaliic acid
for the estimation of racemization.
Pioneering work based on the changes in human non­
collagenous proteins from dentine by Sajdok et al 66
demonstrated a linear correlation with increasing age, but
the wide scatter of values in this initial work limits its cur­
rent usefulness.
Should a tetracycline antibiotic be used to treat a febrile
illness at a time when the crowns of teeth are still forming,
it will bind to the hydroxyapatite crystals being laid down
in the forming dentine and enamel. The bound tetracycline
is fluorescent in ultraviolet light and visible in ground sec­
tions. From an examination of these F1 uorescent bands, the
number of courses of antibiotic may be deduced and an
estimate made of the age of the child at the time of admin­
istration. It may be possible to match these findings with
the medical records of a missing child [Fig. 23.14).
Sex Determin ation
Sex determination is normally achieved from the skull
rather than the teeth. Prominent supraorbital ridges and
mastoid processes may suggest masculinity, as do heavy
rugged jaws with a tendency for the angle of the jaw to be
a right angle G7 Other features in male skulls include promi­
nent muscle lines in the occipital area, square orbits with
rounded margins, heavy, lateral arched zygomatic bones
and large broad palates. G8 •59 These features are only sugges­
tive; children and bodies fragmented in fires and explosions
444 I Paediatri c dental identification
may be very hard to sex from their skulls. 7o The cellular
material in toot h pulp can be analysed for X and Y sex
chromosomes/ I •n although vVhittaker et al n also showed
that the determination becomes less reliable as putrefaction
proceeds. Intact cell nuclei are needed in most methods, and
this is clearly of limited use in burnt remains. 73 However,
the protected dental pulp is an exploitable source of DNA
because of its resistance towards physical a nd chemical
exterior conditions. 74 Work by Urbani et al 75 has demon­
strated that dental pulp is an excellent source of DNA for
gender determination, being retrievable and capable of
an alysis from unprotected teeth heated up to 200°C for 15
minutes. With teeth embedded in bone and soft tissue a 100
per cent success rate was obtained from teeth heated up to
350°C for 15 minutes.
Recently, Karaman 76 reported achieving 83 per cent
accuracy in predicting sex usin g diagonal , as distinct from
mesiobuccal, measurements of anterior teeth. However, this
degree of accuracy could only suggest, not prove, the sex
of skeletonized remains.
Nevertheless, more sophisticated methods are not with­
out their problems. Kumar and Hegde,n using polymerase
chain reaction techniques, were able to identify sex with
100 per cent accuracy from the pulps of deciduous teeth,
but only up to 1 month after ex traction. After 6 months, the
method was no longer effective.
Height
Fazekas and Kosa 78 showed that all the bones of the fetus
can be measured and that, by constructing regression curves
for each bone, the size of the bone may be related to the
height at death of the deceased. Clement and Kosa 43 claim
that the basilar part of the occipital bone, the zygomatic and
the temporal bones are excellent for this purpose, but the
best and easiest bone to identify is the mandible. They state
that the length in millimetres of the mandible is equal to the
total crown-heel body length in centimetres, while the max­
illa is 1/20th the length of the body. At birth, with all the
bones of the body much further developed, direct measure­
ments are possible_and these methods are superfluous.
Ethnicity
Prior to the advent of DNA analysis, the determination of
ethnic origins from dental remains was unreliable because
the dental a nd orofacial indicators were very imprecise and
the overlap betvveen racial characteristics was so consider­
able that, realistically, they were of little value in determin­
ing the ethnic origin of anyone set of remains.7o Hurme 79
states that the eruptive process occurs 6 months to 1.5 years
earlier in blacks than in whites, but this trend is of little
value in profiling and it remains thus unless the circumstan­
tial evidence indicating a specific ethnic origin is strong.
-- - -
In addition , neither of these studies makes a distinction
between eruption and emergence. For a highly detailed and
robust scientific study of odontometrics relating to compar­
ative and sexual dimorphism the reader is referred to Human
Adult Odon tom etrics by JA Keisser. 8o
Teeth have more recently caught the imagination of the
Department of Anthropology of the University of Durham,
where Buck and Vidarsdottir81 of the Evolution ary Anthro­
pology Research Group have investigated differences in
the mandibular morphology of 'subadults' (i.e. not adults)
using geometric analyses. One hundred and seventy-four
mandibles from five morphologically distinct samples were
digitized and subjected to Procruste's analysis. The results
showed significa nt morphological differences between the
samples and obtained cross-validation results of over 70 per
cent accuracy in identification of unknown individuals
using the complete mandibl e. They suggest that these tech­
niques could provide a method for the identification of race
in a subadult individual. This appears to be, at first sight, a
useful tool in demographic research, but application to pro­
filin g an individual skeleton can only result in a suggestion
or, a t best, increased probability of gender determination.
REFERENCES
Carey E. Dio's Romall History, vol. 8, book 61. London:
William Heinem a nn, 192 5, p. 17.
2 Solheim T, Lorentsen M, Sundnes P et a l. Th e 'S candinavian
Star' feny disaster 1990 - a challenge to forensic odontology.
lnf J Legal Med 199 2; 104:339-45.
3 Delattre V. Forensic denta l id entification in the greater
Houston area . J Forensic Sci 200 1; 46:1379 - 84.
4 Knight B, Pearl H. The legal system in the British Isl es.
In Clark D (ed.) Pracrical Forens ic Odontology. Oxfo rd:
Butterwo n h-Hein emann, 199 2.
5 Hill IR. Id entification , the dentist and the coroner in Engl a nd
and Wal es . J Foren sic Odontostomatol 1988; 6:77-82.
6 Jones DG. Odontology often is the final piece in a grim
puzzle. J Calif Dent Assoc 19 90; 26:650-1.
7 Dailey J. Charting errors in ma ss disa ster denta l reco rds. In
Bouers CM, Bel l GL (eds) Th e Manual oj Forensic Odontology,
3rd edn. Saratoga Springs, NY: Am erican Society of Forensic
Odontology, 1995, pp. 250-8 .
8 Kiesser- Neilson. Person Idelltijicatioll by Mea ns oJThe Teeth:
A Practical Guide. Bristol: John Wright, 1977.
9 Philips VM. Th e uniqu eness of amalgam restorations for
identifi ca tion. J Foren sic Odol1tostol11atol 1983; 1 :33-8.
10 Smith B. Recons truction of root morphology in ske letonised
remains with postmortem dental lo ss. J Forensic Sc i 1992 ;
37: 176-84.
II Wood RE, Tai CE, Blenkins op B, Johns ton D. Digiti sed slice
interpOSition in forens ic dental radiographic identification.
A m J Forensic Med Pathol 1994; 15:70-0.
12 Wood RE, Kirk NJ, Sweet DJ. Digita l dental radiographic
identifica tion in the paediatric, mi xed and permanent
dentitions. J Forensic Sci 1999; 44: 9 10-16.
13 de Villi ers CJ, Philip s VM. Pe rson identification by me ans of a
single unique dental feature. J Forens ic Odonrostomotol 1990 ;
16:17-19.
14 Nonje CJ. Harris AJIiIP. Maxi llo-facia l radiology in forens ic
dentistry: a review. J Forensic Odon tostol11a tol 1986; 4:2 9-38.
- -~-
---- -

15 Watt R, Sheiham A. In equalities in oral health: a review of the
evidence and recommendation for action. Br Dellt J 1999;
187 :6- 12.
16 Kaste LM , Selwitz RH, Oldakowski RJ et al. Coronal caries in
the primary and permanent dentition of children and
adolescents 1-17 years of age: United States, 1988- 1991.
J Den t Res 1996; 75:631-41.
17 White BA, Caplan OJ, Weintraub JA. A quarter century of
changes in oral hea lth in the United States. J Dent Educ
1995 ; 59:19-57.
18 Flink A, Kallestal C, Holm AK et al. Distribution of caries in
12-yea r-old children in Sweden. Social and oral hea lth­
related behavioural patterns. Coml11l1ni~y De/If Hea ltft
1999; 16:160-5.
19 La wrence HP, Sheiham A. Caries progression in 12- to 16­
year-old school children in tluoridated and deficient areas in
Brazil. Communiry Dent Oral Epidemiol 1997; 25:402-11.
20 Moody GH, Busuttil A. Identification in the Lockerbie air
disaster. Am J Forensic Med Pathol 1994; 15:63-9.
21 Sholl SA, Moody GH. Evaluation of dental radiographic
identification: an experim ental study. Forensic Sci fnt
2001; 115:165-9.
22 Ahlberg JE. We must get our numbers right. FDI Newslerrer
1987; 158:8-9.
23 Ayton FE, Hill CM, Parfitt HN. The dental rol e in the
identification of the victims of the Bradford City football
ground fire. Br Dent J 1985; 159:262- 4.
24 Ireland RS, Harris RV, Pealing R. Clinical reco rd keeping by
General Dental Practitioners piloting the Denplan 'Excel'
accreditation programme. Br Dent J 2001; 191 :260-3.
25 Tesini DA. O'Malley KD, Schwartz S. Development of bite
impression technique for use in identification of unknown and
missing children. J Mass Dent Soc 1985; 34:61, 63 , 69, passim.
26 Tesini DA, Harte D, Crowley K. Dentistry's role in the
identification of missing and unknown children: upd ate on
the dental bite impression technique. J Mass Dent Soc 1999 ;
48:29-34.
27 Tesini DA. Comments on the Toothprints Bite Impression jo r
Search and Identification oj Missing and Unknowl1 Children.
2003. http: //www.kerrdental.com/Toothprints/
Toothprints0f020Position20Paper
28 Whittaker DK, MacDonald D. A Colour Atlas oj Forell sic
Dentistry. London: Wolfe Medical Publishers, 1989.
29 Aws G. An Overview and CompaTison ojCAPMI and DAVID
Programmes. Dissertation for the Diploma in Forensic
Odontology. Hatfield, UK: Uni ve rsity of Hertfordshire, 1999.
30 American Board of Forensic Odontology. Body identification
guidelines. J Am Dent Assoc 1994 ; 125:1244-52.
31 Koe1myer TO. Video ca mera superimposition and facial
reco nstru ction as an aid to identification. Am J Forensic Med
Patho/1982; 3:45-8.
32 Perpe r JA, Patterson GT. Face imaging reconstructive
morphography. A new method for physiognomic
reconstruction. Am J Forensic Med Patho11988; 9:126-38.
33 Hill B, Macleod RM, Crothers A. Rebuilding the face of
George Buchanan (1506 - 1582). J Audioli Media Med 1996;
19: 11-15.
34 Philips VM, Smuts NA. Facial reconstruction: utilization of
computerized tomography to measure facial tissue thickness
in a mixed race population. Forensic Sci Int 1996; 83:51.
35 Law CA, Bowers CM. Radiographic reconstruction of root
morphology in skeletonised remains: a case study. J Forensic
Sci 1996 ; 41: 514-17.
36 Shahrom AW, Vanezis P, Chapman RC et aJ. Techniques in
facial reconstruction: computer-aided facial reconstruction
using a laser scanner and video superimposition. Int J Legal
Med 1996; 108:194-200.
References I 445
37 Tyrell A, Eviso n, MP, Chamberlain AT, Green MA. Forensic
three-dimensiona l facial reconst ruction: historical review and
contemporary developments. J Forensic Sci 1997 ; 42 :653-61.
38 BrockJebank L, Holmgren CJ. Development of equipment for
the standardization of skull photographs in perso nal
identification. J Forensic Sci 1989; 31: 1373-9.
39 Simpson JDR. The investigation of an unknown skull with
superimposition of known faces. Dissertation for a Master of
Science Degree. Edinburgh: University of Edinburgh, 1998.
40 Malaver PC, Yunis YY. Different dental tissues as source of
DNA for human identification in forensic cases. Croat Med J
2003; 44 :306-9.
41 Calonius PEB, Lun in M, Stout F. Histologic criteria for age
estimation of the developing human dentition. Oral Surg
1970 ; 29:869-76.
42 Kraus BS, Jordan RE. The Human Dentition Bejore Birth.
Philadelphia, PA: Lea and Febiger, 1965.
43 Clement JG, Kosa F. The fetal skeleton. In Clark DH (ed.)
Practical Forensic Odontology. London: Wright, 1992,
pp. 43-52.
44 Ashley KF. Identification of children in a mass disaster by
estimation of dental age. Br Dent J J 970; 129: 167 -9.
45 Ciapparelli L. The chronology of dental development and age
assessment. In Clark DH (ed.) Practical Forensic Odontology.
London: Wright, 1992, pp. 22-42.
46 Schour L, Hoffman MM. Studies in tooth development, 2. The
rate of ap position of enamel and dentine in man and other
mammals. J Dent Res 1939; 18:161-75.
47 Noble HW. The estimation of age from the dentition. In
Harvey W (ed.) Dental Identification and Forensic Odontology.
London: Kimpton, 1976, pp. 28 - 35.
48 Moorrees CFA Fann ing EA, Hunt EE. Formation and
resorption of three deciduous teeth in children. Am J Phys
A l1thropol 1963; 21 :205-13.
49 Moorrees CFA. Fanning EA, Hunt EE. Age variation of
formation stages for ten permanent teeth. J Dent Res 1963;
42: 1490-502.
50 Demerjian A. Postnatal grow th. In Falkner F, Tanner JM (eds)
Dentition in HU1110n Growth. London: Bailliere Tindall , 1978,
pp. 413 -44.
51 Koski M, Gam SM. Tooth eruption sequence in fo ssil and
modern man. Am J Phys Anthropol 1957 ; 15:469-88.
52 Demiljian A. Goldstein H, Tanner JM. A new system of dental
age assessment. Hum Bioi J973 ; 45:211-27.
53 Demerjian A, Goldstein H. New systems for dental maturity
based upon seven and four teeth. Ann Hum Bioi 1976 ;
3:411-21.
54 Schour L, Massier Iv!. The development of the human
dentition. JAm Dellt Assoc J941; 28: 11 53 - 60.
55 Maber M, Liversidge HM, Hector MP. Accuracy of age
estimation of radiographic methods using developing teeth.
Forensic Sci Int 2006; 159:S68-73.
56 Willems G, Van OImen A, Spiessens B, Carels C, Dental age in
Belgian children: Demirjian's technique revisited. J Forensic
Sci 2001; 46:893-5.
57 De Salvia A. Calzetta C, Orrico M, De Leo D. Third mandibular
molar radiological development as an indi cator of
chronological age in a European population. Foren.sic
Sci Int 2004; 146:S9-1 2.
58 Garamendi PM, Landa M1, Ballesteros J, Solano MA .
Reliability of the methods applied to assess age minority in
living subjects around 18 years old. A survey on a Moroccan
origin population. Forensic Sci Illt 2005; 154 :3 -12.
59 Ciapparelli L. The assessment of dental age in Essex school
children using panoral radiographs with forensic app li cations.
Dissertation for the Diploma in Forensic Odontology. Lond on :
Hospital Medical College, 1985.
446 I Paediatric dental identification
60 Monstrad H, Staaf Y, Welander U. Age es timation with the aid
of tooth developm ent : a new method based on objecti ve
measurements. SCll1l d J DCIII Res 1994; 102: 137 - 43.
61 Liversiclge Hrvl, Lyons F, Hector MP. The accuracy of three
methods of age estimation usin g rad iog raphic meas urements
of developing teeth. Forellsic Sci il1t 2003; 131 :22- 9.
62 Foti B, La lys L, Adalian P et al. New forensic approach to age
es timation in ch ild re n based o n too th erup ti on. Forel1sic Sci
ill I 2003; 132:49- 56.
63 Ritz -Timme S, Ca ttaneo C. Collins rvlJ et al. Age est im ation:
the state of the art in relation to th e spe cific demand s of
forensic pra cti ce.illl J Lego! M ed 2000 ; 113:1 29 - 36.
64 Arany S, Ohtani S, Yoshioka N. Gonmori K. Age esti mation
from asparti c acid racemi za ri on of root dent in by internal
standard meth od. Forellsic Sci 111 1 2004: 141: 127-30.
65 Yekkal R. Meers C. Van Schepdae l A et al. Rclc emi za rion of
aspartic acid from human dentin in the estimation of
chronologica l age. Fomlsi c Sc i i1l12006: 159:S89 - 94.
66 Sajdok J, Pilin A, Pudil F et al. A new method of age
est im ation based on the changes in human non -col lageno us
prote ins from dentin. ForeHsi(' Sci i ll l 2006; 156:245-9.
67 Aitchison J. Sex differences in teeth, jaws and skull s. Delli
Prerel 1963 ; 14:52-7.
68 Biggerstaff RH. Craniofacial characteristics as dete rminants of
age, sex and race in foren sic dentistry. Del/r Clil/ North Am
J 997 ; 21 :85- 97.
69 Borha CT. Cra ni o facia l characteri stics as determ inants of age.
race and sex in forensic dentistry - a hand s on guid e.
J Forel/ sic Od OI110S(0))/lIto/ 1991; 9:47-61.
70 Harvey W. Effec ts of sex, race, hereditary and systematic
disea se on oral tissues. In Harvey W [ed.) Del/ta! Jdel/rijicotiol/
III/d Forellsic Odol/to!ogy. Lond on: Henry Kimpron. 1976.
pp.36-4 3.
71 SenD M. Ishi zu H. Sex iden tification of a human tooth.
illl J Forel/sic Dml 1973 ; 1:8 - 1I.
72 Whittaker DK, Ll ewelyn DR , Jones RW. Sex determin ation
from necrotic pu lp tissue. Br Dellt J 1975; 139:403- 5.
73 Murakami H, Yamam oto Y, Yosh ito me K et al. Forensic study
of sex determination using PCR on tooth sampl es. Acta Med
OkaYII/JI1I 2000 : 54:2 1-3 2.
74 Ga ill a rd H. Lucl es B, Kaess B, Ma ngin P. Th e use of the teeth
in genetic fin gerprinting. Bull GrouJl i nt Reclr Sci SI011111tal
OdO lltO / 1994; 37 :65-70.
75 Urbani C, lastrucci RD, Krame r B. The effec t of te mp era ture
on sex determination usin g DNA-P CR analysis of dental pulp.
J Forel1 sic OdOI/IOS I0111I1to!1999: 17:35-9.
76 Karaman F. Use of diagonal tee th measurem ents in predicting
gender in a Turk ish population. J Forensic Sci 2006;
51 :630-5.
77 Kumar MG, Heg de AM. Sex ide ntification from exfoliated
primary reeth - a PCR study. J Ciil1 Pedilltr Den12005 :
30:19 - 21.
78 Fazekas lG, Kossa F. Determ in ation of the body leng th and
age of fe tuses on the basis of th e dimen sions of the facial
bones. Chapte r X . In Fa zeka s IG, Kossa F (eds) Forensic
Fetll! Os re% m'. Budapest: Akade miai Ki ado, 1978,
pp. 16 5-96.
79 Hurm e va. Time and sequ ence ot- tooth el1lption. J Forel/s
Sci 1957 ; 2: 371 -88.
80 Ki esser JA. HUIIIan Adulr OdOIIlollIclrics. Studies in
Biol og ica l An thro pology. Camb rid ge: Cambridge Uni ve rsity
Press, 1990.
81 Buck TJ, Vidarsdottir US. A proposed method for the
id ent ifi ca tion of race in sub- adult skeletons: geo metric
morpho metr ic analysis of ma ndibular morph ology.
J Forel/ sic- Sci 2004; 49 : 11 59 -64.

THE EXPERT WITNESS AND EXPERT
TESTIMONY
Anthony Busuttil
Introduction
Mission statement of the expert
Claim to expertise
Professional witnesses
Opinions
Yes or no?
Admissibility of expert evidence
Communications from the expert witness
Physicians are not ordinary witnesses, but give judg­
ment rather than testimony.
Digest - Emperor Justinian (AD483-565)
INTRODUCTION
The courts, by the very diversity of cases that are brought
to their attention, wi ll, from time to time, invariably require
assistance from those who possess intimate and profound
knowledge of specific and-specialized subjects, of which the
court is not expected to have direct first-hand knowledge.
This is particularly the case in the context of an adversarial
system of law, as practised in Great Britain - and in most
Commonwealth countries - and the USA, whether the case is
being tried in the civil or criminal courts. For example, if the
matter under consideration by the court is the safe construc­
tion of a bridge, it is unlikely that the court would know
much about the specific matters related to this, unless
mechanical engineers, architects, steel erectors and quantity
surveyors ass ist them by providing them with basic details of
constructional requirements, good practice and building
techniques. Such persons are referred to as eJ.pert witnesses
*The mal e gender that is used throughout th is chapter also refers to the female gender in all instances.
I CHAPTER 24 I
447 Declaration by the expert in the report 452
448 In the witness stand or box 453
449 Pre-trial communication 453
449 Conflict of interest 454
450 Rules of evidence 454
450 Conclusion 454
450 Recent developments 454
452 References 455
in that they are assisting the court through (Latin ex.) their
acquired specialized skill, craft or profession (Latin periNa).
WG Dickson stated that: 'Another exception to the general
rule against examining witnesses on matters of opinion
occurs wherever the issue involves scientific knowledge, or
acquaintance with the rules of any trade, manufacture or
business with which men of ordinary intelligence are not
likely to be familiar. A foundation for such an examination
must be laid by ascertaining whether the witness is a person
of skill or an "expert" [the English term], under which is
included those who have a theoretical acquain tan ce with the
subject, as well as men who speak from practical know ledge:]
Medical practitioners who are summon ed to the courts,
criminal or civil, in order to assist with medical matters can
therefore appear as expert witnesses. In criminal matters,
their ex peliise is mainly required for the interpretation of
traumatic changes and causes or modes of death; in civil
cases, medical expertise is requ ired for cases of alleged
medical or industrial negl igence and in li tigation relating
to compensation for injury and disability.
As with all other evidence given to the courts, the testi­
mony is taken under solemn oath or affirmation; thereby, it is
empowered with a further, more robust, sea l of assurance as
to its veracity, totality, professional integrity and impartiality.
448 I Expert witness/expert testimony
As Cicero stated, 'the court is the temple of truth, and the
search for truth is critical towards the conduct and conclu­
sion of a fair trial'; this should be the tenet that underlines
the involvement of the expert witness.
Similarly, in the preparation of cases for trials, be these
criminal or civil cases, it is often useful for both the prose­
cuting auth orities and legal defence team (criminal) or, for
the legal team putting the case together for one or other
of the litigants involved (civil) , to acquire information on
subjects with which they are not familiar. This would guide
them with respect to the strength or weaknesses of the case
and in coming to considered decisions as to whether to
plead guilty to the charges, to settle out of court, to instruct
further investigations to be carried out prior to the trial
or to contest the trial. A very onerous duty and a heavy
responsibility are placed on the expert. Initially he has to
consider very carefully whether or not he possesses the
qualities appropriate to fulfilling the expectations placed
on him by the courts or by legal professionals in the partic­
ular matter under review.
MISSION STATEMENT OF THE EXPERT
By its very nature, the criteria that expert evidence has to
satisfy are the following:
1. It refers solely to specialized medical matters and
themes, and not simply to matters that can be deduced
from a common-sense or ordinary knowledge approach
or by recourse to simple logic. It is on such specific
issues, that the court - namely the ordinary 'ladies and
gentlemen of the jUly' and/or the presiding judge(s) ­
may require to be enlightened in order to reach the
more appropriate verdict.
2. It has to be delivered only by those who are
appropriately qualified to do so, both by virtue of an
appropriate duration of first-hand specific experience
in the matter under review and by having acquired
their expel1ise by periods of study - these academic
degrees and diplomas shou ld be related to the subject
areas being addressed as, an expert. The direct expertise
which is being sought should be kept in mind by the
expert who is instructed, and, if it is felt that the
subject is not within his direct professional knowledge
and specific experience, then he should decline to act
in the capacity of an exp e11 in the particular case. In
the recent past, a trend developed in civil court rooms
in which persons with a sound general expeliise spoke
on more specific matters of which they had little first­
hand knowledge, and thereby brought the role of the
expert witness into a certain level of disrepute. For
example, a general surgeon cannot, under ordinary
circumstances, be expected to speak from a vantage
point of expertise on intricate stereotactic neurological
procedures; if he does, he is quite likely to be
overstretching his expertise. However, he is in a
position to give excellent expert advice on
herniorrhaphy and its complications. The expert must
appreciate the boundaries of his expertise and
knowledge, and always ensure that he is not making
pronouncements on subjects outside his experience.
3. The delivery of the evidence has to be such as to
enable the court to appreciate and understand the
specialized information that is being imparted to them;
if this evidence is written down in the form of the
report, it must be user-friendly, devoid of long,
rambling sentences, and bereft of a surfeit of jargon
and statistical calculations. It should be laid out
logically, in sequentia l chapters, and it cannot be seen
to be muddled, perhaps focally inconsistent, with
conclus ions in it that are not fully justified and
substantiated by wi?at appears earlier in the evidence.
The information that is being imparted, albeit often
quite complex and difficult, would somehow have to
be titrated to a level that is readily comprehensible by
the jury. The lawyers - and failing them the judge
himself - will usually steer the expel1 in the correct
direction while on the witness stand, and may ask him
to elaborate on certain passages or chapters of his
evidence, perhaps by giving analogies and examples
that will assist the jury's comprehension.
4. This evidence has to be, as much as possible, f1awlessly
dispassionate, totally unbi ased, non-partisan, carefully
balanced and comprehensive. It must not contain
views tailored to the advantage of the side instructing
the expe11. In an adversarial system, whichever side
calls on the services of the expert, the evide nce given
to the court should be unsullied by leanings in favour
of one or other side. Experts - including medical
expe11 witnesses - who fail to achieve this level of
probity have been variously referred to in a derogatory
fashion as 'hired guns' and, worse, as 'gobs for hire'.
The expert must always keep in mind that he is not
there to win cases or to ensure that his 'side' triumphs
against all odds; he is there to assist the court, to
the best of his capabilities, in reaching the most
appropriate decision or verdict, and no more. On
many an occasion, the temptation may be very
pronounced to veer away from this rule - but on the
expert's head be it if he is found to have taken sides in
this manner.
5. The data adduced in the expert's evidence must be
accurate and up to date, both in terms of listing all the
information given to him in his instructions by the
legal personnel and in quoting the current peer­
reviewed literature. Using stale and superseded
references in either books or journals that are later
discovered to be so by the courts discredits the expert's
testimony and his expertise. If the information given
originally was somehow incomplete or in accurate, at
least in the light of the evidence actually stated in
court, and the conclusions reached are no longer

tenable in the face of what was actually heard from the
witness box. then the expert should honestly inform
the instructing side that, as a direct consequence of
this, his opinion has had to change. The courts, and
eventually the instructing laYlryers , will accept ­
sometimes quite reluctantly - that honesty is indeed
the best policy. This implies that there may be
occas io ns when it is expedient and indeed essential
that, with the COUlt's permission. the expert actually
sits in the court room and listens in to the factual
evidence as it is being delivered.
6. All the data referred to by the expert must be properly
acquired at clinical examination or necropsy or from
available documents. To do otherwise would be
fraudulent and. if the case proceeds to court, may be
tanta mount to perjury. In this respect, the infamous
instance of an expelt clinical neurologist who gave
evidence on the findings of a clinical examination that
he had never act ually personally conducted is perhaps
a major sa lutary and admonitory lesson 2
The genera l judicial requirements of the expeli were
summarized in the English civil case, National Justice Com­
pania Naviera SA v. Prudential Assurance Co. Ltd, tile
Tk.arian Reefer (1993) 2 Lloyds Reports [2] 68, 81-82, and,
after appeal, in (1995) 1 Lloyds Reports 455. These are par­
aphrased as fo llows:
• Expert evidence must be an independent product of the
expert uninfluenced by the exigencies of the litigation.
• Expert evidence must be of independen t assistance to
the court by way of objective unbiased opinion within the
expert's expertise, wi th the expert never assuming the
role of an advocate.
• The expert witn ess must state the Facts or assumptions
for the opinion expressed, never omitting material
facts.
• The expen witness must clearly indica te when a matter
falls outside his ex pertise.
• The expert witness must say if the data available are
insufficient or if he cannot assert that the report
contains the whole truth.,
• If the expert witness changes his opinion, he must
co mmunicate this to the court and to the parties
involved in the action.
• If any photographs or other material are used to
compile an expeli repoli, all of this must be made
available to other parties.
CLAIM TO EXPERTISE
Within the forensic arena, no matter how eminent the expert
witness, aU aspects of the claimed expeliise and the testi­
mony given will be challenged by the legal parties involved
in the trial - quite often themselves aided by 'experts' whom
they have specifically hired to assist them and who may
Professional witnesses I 449
indeed be present when the testimony is actually taken. The
expert is never immune to criticism and to adverse com­
ment; he can, and most certainly will, be challenged in open
court through searching quest ions and cross-examination.
Evidence from other experts may be adduced to demonstrate
that in the evidence given to the court, th ere were short­
comings, inadequacy of knowledge, flawed interpretations, a
failure to consider all po ten tial options, and so on. All these
matters should be considered prior to agreeing to prepare an
expert report.
There is no scope for arrogance or ex cathedra dogmatic
pronouncements of the ipse dixit style; the court will be
sure to query and to probe, and the expeli has to cultivate
an unperturbed style that enables him to parry such ques­
tions and indicate why he disagrees with the propositions or
scenarios that are being put to him , a nd to remain cool and
collected, even under what may appear to be serious provo­
cation and verbal pressure from those who have been
trained so to do.
Thus, the opening gambit in couli is for th e expert to
establish, to the satisfaction and probing of the court, the
breadth and depth of his specific expertise, in order to pro­
vide reassurance. This is always the case, whether or not spe­
cific criteria are actually laid down by the courts delineating
who can and who cannot be an ex pen. In continental juris­
dictions, the court has access to a list of persons who have
been considered, vetted and accepted by the courts as pos­
sessing the appropriate standing to qu ali fy them as experts.
In the USA, the Federal Rules of Evidence (Rules 104,40\-3,
702-6 - the Committee on the Judiciary, House of Represen­
tatives, US Government publication) establish the authority
of the court to determine whether a person can be considered
as an expert, whether his testimony can be heard, the type
of evidence that can be admitted in court and the permitted
scope of such testimony.] In the light of the Woolf Report,
the position in Britain may already have changed in some
respects towards the continental system.
PROFESSIONAL WITNESSES
A fine line of distinction needs to be drawn between the
so-called professional witnesses and expert witnesses.4 Any
professional person, including a medical doctor, who is
required to give testimony a bout professional matters is
considered as a professional witness . Thus, for example, a
young ARE department doctor w ho gives evidence about
when he saw a patient wh o self-referred to his department
- describing the nature of the injuries that he observed ,
with their measurements and locations, the number of
sutures that he inserted, the presence and absence of frac­
tures - is giving professional evidence. Yet he should not
be expected to be able to inform the court, from a position
of expertise, whether the injuries he described were likely
to have been caused in any particular manner or by any
specific weapon, as this is unlikel y to be within his direct
450 I Expert witness/ expert testi many
knowledge and ex perience. He is thus a professional and
not an expert witness.
This is a problem that has to be guarded against, both
from the point of vi ew of the individual witness - who
should have no hesitation at refusing to accept an expert's
role if he feels unable to fulfil it - and from that of the
courts - who should not expect every professional witness
to transform himself into an expert witness on the spot
when the court feels inclined to require this from him,s
OPINIONS
The ability to assist the courts in matters not directly
observed or personally done by the expert and the giving of
op inion s highlights the only major co ncessions and privi ­
leges that are given to the expert witness by the court; it is
ass umed that, given his stated expertise, he can assess and
weigh the information that he is provided with and present
to the court balanced opil1io11s with regards to consistency,
possibility, probability, ceI1ainty and likelihood of certain
matters. He is not there to usurp the function of the court
and take decisions, but sim ply to give evidence. Nor is he
there to lecture and harangue tlle COU lt, no matter how emi­
nent his status and how sophisticated the subject matter that
he is addressing. He cannot present ilie jury and judge with
ready-made conclusions or take decisions on behalf of the
court; he must at all costs remember that he is there only to
assist them with infolmation , albeit of a specialized type and
with opinions that will en able them to make up their own
minds. Thus, the expert cannot proffer an opinion on the
'ultimate issue', i.e. the crux of the case, namely the central
issues iliat will be deci ded by th e jury. In the light of R v.
Stockwell (I 993} 97 CR App R 260, an expert in DNA profil­
ing cannot say to the court what he thinks the likelihood was
that the defendant deposited blood or seme n at the crime
scene as 'this requires considerati on of factors other than
those within his area of expertise', and indeed this opinion
wi ll have to be based on evidence that has been given in
courts by other pa rties, which would have occurred when he
was not present.
This was brought to the fore when expert evidence was
given relating to the degree of intoxication in a case of a
woman who was alleging rape; there had been an allegation
that she had been under the influence of alcohol. The expert
referred to the complainant's level of intoxication as being
commensurate witll that of 'someone who was not in control
of her behaviour and judgement'. Although the expert was
fully entitled to give this opinion on tlle matter of substance
use in this case, the judge should have directed the jury that
they were not bound by the expert's opinion and that they
could make up their own minds (R v. Ugoh Et Others [2001]
EWCA Clim 1381). In tlle case Re Hand R (Minors), the
judge stated that, although the medical evidence was 'q uite
indispensable', the 'final decision in the case is the judge's,
and his alone'.
YES OR NO?
It is a fact with medical matters that often there are many
possible ca uses or associations, although some of these are
more probable than otllers, and some almost impossible; it is
a matter of balanced judgement as to how a specific question
about likelihood can be addressed and answered comprehen­
sively. It is therefore frequently the case that a cut-and-dried
answer to a particular question asked of th e expert witness
cannot be given, eith er in the affi rmative or in the negative.
The answe r de pends on a numb er of facts and other factors,
and can vary through a considerab le spectrum of possibili­
ties in the light of other prevailing and contributory circum ­
stances. The expert should explain this to the court.
A clear yes or 110 answer cannot be given with the best will
in the world. Lavvyers may therefore ten d to consider doctors
as being obtuse and sh ifty when they cannot provide the
direct and curt answers that are being sought of them.
Knock ing square pegs into round holes for the benefit of giv­
ing the lawyers ilie trite and curt responses requested by tllem
does no on e any favours, particu larly the expert him self; one
shou ld be as helpful as possible, bu t in many instances the
telegrammatic answers required cannot honestly be given.
Any bUllying by lawyers in this context should tllerefore be
strenuously, but politely and resolutely, resisted, and if over­
persistent, the judge should be asked to intervene.
ADMISSIBILITY OF EXPERT EVIDENCE
Paraphrasing C] King, the judge in the Australian case of
R v. Bonythol1 1984 38 SASR 45, admissibility of expert
evidence was held to involve the fo llowin g main issues:
• Does the witness claiming expelt status possess
sufficient knowl edge and experience to render his
opinio n of substance and value in assisting the court
to resolve the issues before it?
• Does the subj ect matter of the evi den ce fall within the
class of subject upon which expelt testimony sho uld be
permitted, i.e. forms a part of a body of knowledge or
human experience th at is sufficiently organized and
recognized as to be accepted as reli able?
• If new and unfamiliar techniques or technology have
been used by the witness at arriving at his conclus ions,
can the court be satisfied th at such techniques or
technology have a sufficient scientific basis to render
the result arrived at reliable and acceptabl e?
• Will the judge, without experience in the subject
matter on which the opinion is being sought, be ab le to
for m a sound judgement without the assistance of the
'expert ' witnesses possessing spec ial knowledge and
experi ence of that subject matter?
In the US federal cou rts, adm issibility of expert evidence
was, to a maj or degree, guided by the Frye test (Frye v.
United States, 54 App. D.O. 46, 293 F. 1013, 1923), namely

that the courts adm itted expert evidence deduced fro m
weJl-recognized scientifi c principles or discoveries: ' the
thing from whic h the ded uction is made must be suffi­
ciently establ ished to have ga ined general acceptance in
the particul ar field in which it belongs'. The legal grounds
of acceptab ili ty of expert evidence are even h igher a nd
more demand ing than those of the expert's discipline 6 This
'general acceptance test' changed after the Daubert v.
Merrell Dow Pharmaceuticals Inc (1993) 113 S Ct 2786 7 case
and was replaced by a num ber of more flexible guidelines :
• Can the theory or techni que on which t he evidence is
based be tested independ ently? Has it been tested?
• Can a determination be made of its fa lsifiab ili ty, i.e.
can it be proved to be wrong if it is in fact wrong?
• Has the theory or technique been subj ected to peer
review and pub lic atio n?
• What is currently known of its potential error rate?
• Is the theo lY or technique generally accepted?
The Kumbo Tire Co Ltd v. Carmichael (1999) 526 US 137
case perhaps enhanced further the flexibility of the test by
stating that 'Daubert's list of specific factors neither neces­
sarily nor exclusively applies to all experts in all cases'. The
US supreme co urt in Daubert's perhaps sum s up the 'proof
of the pudding' in relation to expeli ev idence, namely
'vigoro us cross-exam ination, presentat ion of contrary evi­
dence and carefu l instruction on the burden of proof as
being the t raditional means of attacking shaky but admis­
sib le evid ence'.
In Brita in, detailed tests of admissibility of expert evi­
dence have not exercised the judiciary to any major extent;
admissibility, to a large extent, has been based on a prag­
matic approach from the bench, erring on the si de of accept­
ance and acquiescence rather than otherwise, with evidence
being ruled to be admissible if it is considered to be relevant
and its probative va lue is considered to outweig h its prejudi­
cial effect. s Thus, in R v. Robb 93 Cr App R 161, the COLlli of
Appea l upheld the trial judge's admiss ion of evidence given
by a phonetician on voice recognition as expert, even
though misgivings were expressed about the scientific basis
for such ex pertise. , expert,6 a single, joint expert who may be given instructions
In the English family division, in the Re AB (child abuse;
expert witnesses) (1995) 1 FLR 18 1 case the judge had to
weigh the evidence before him in terms of fractures sus­
ta ined by a young child. An expeli for the defence claimed
that a form of brittl e bone disease akin, but not identical ,
to the va rious types of genetica Lly documented brittl e bone
disease existed that did no t have simila r hereditalY a nd
familial characteristics of osteogenesis imperfecta but could
account for the spontaneous occu rrence and accidenta l
infliction of injuries, and thus exclude criminal damage and
assau lt. 9 This 'temporary' brittle bone disease had not been
substantiated by standard research studies or by pUbli ca ­
tion s in peer-reviewed journals. In particular, the judge, in
his criticism of this expert, indicated that he lacked obj ec­
tivity, omitted factors that fai led to support his theory and
Admissibility of expert evidence I 451
continued to cite cases, altho ugh they had been t he subject
of decisi ve judicial findings of non-accidental injUlY inflic­
tion. In summary, the judge considered this evidence to be
unrel ia ble and inexpeli and he threw out his 'expert'
evidence.
In the context of child ab use, particularly sexual abuse,
a considerabl e debate has arisen, particularly in the USA,
concerning the weig ht tha t should be attribu ted to testi­
mony of such historical ab use because the subj ect could
not recall spontaneously a ny such occurrences; expert psy­
cholo gists and psychotherap ists were required to gather
information through repet itive interviewing and sessions
under hyp nosis (sometimes combined with use of medica­
tion). These in te rviews we re carried out in alleged attempts
to tap the so-called ' rep ressed-memory', ' recovered mem­
ory' or 'false-memory syndrome'. 10.11 The coulis a nd , per­
haps more teJlingly, the insurance companies who refused
to pay compensatio n to those alleged ly stressed and trau ­
matiz ed by suc h repressed occurren ces have not taken a
very favourab le view of the entire mat te r.
In civi l cases, in parts of Great Britain other than Scot­
land, there has been a move to ration alize the use of ex perts,
particularly in personal injury compensation, and in a con­
sideration of the moves towards cond itiona l fees by law
firms, namely the 'no win, no fee ' situation. In this field, the
lawyers act ing for the claimant under the pre-action proto­
col set-up prepare the case, obtain a ll the relevant medical
documentation abou t their client and submit this to the firm
of so licitors acting for the person/company who will be
defending the act ion. If possible, the matter is resolved at
this point, before court proceedings are ini tiated. If there is
no agreement reached at this stage, then the Civil Procedure
Rules (SI 1998/3132 Part 35) and Practice Direction [1991]
1 WlR 1124, apply. Basically, these infer that the ex pert evi­
dence must be reasonably necessary (Rule 35.1), with the
overriding duty o f the expert being to the co uli.) The court's
permission is sought for the expert evidence4 a nd is then
obtained as a w ritten report. 5 At this stage, the opposing
party may put written questions to the expert. 6 The cOUli
may direct that evidence will be required from only one
by both parties in th e litigation. s If the court believes that
certain documentation should be further disclosed to the
expert then it may direct SO.9 Such disclosed evidence may
be adduced in court, I I but not undisclosed evidence. 12 Th e
expeli may seek direction from the court. 13 The Daniels v.
Walker [2000] 1 WlR 1382, CA, Lord WoolfMR case lai d out
the proper approach for the cOUli .
In the criminal cOUlis, the prosecution and the defence
wi ll usually instruct their own experts, and it is here, perhaps,
that the ogres of bi as and partisanship become manifest and
have to be guarded against in the cut-and-thrust of the foren ­
sic courtroom atmosphere and the persistent questioning by
barristers and advocates. In some instances, th e judge wi/ I
have seen, prior to the trial, the written repOlis from other
experts, and he may decide that there is a sufficient agreement
 452 I Expert witness/expert testimony
and similarity between the two that perhaps a joint statemen t
or minute of the points agreed to by both experts should be
drafted with only the points of disagreement coming to the
jury under their direction for assessment.
In this instance, there should not be any disagreement
about facts and the information collected as evidence. If this
is shown to exist, there is a fundamental evidential problem
that has to be resolved. Thus, for example, if there is a dis­
pute about the substantive clinical findin gs in a patient who
has been examined separately, albeit on different occasions,
by two do ctors or missed limb fractures in a utopsy findin gs
then th ese matters require earl y and definitive resolution .
Another fundamental problem relates to the breadth of
the informatio n collected. A skeletal surveyor a blood clot­
ting scree n that has not been carried out by the prosecu tion
expert when it is thou ght to be essentia l, omission of co l­
poscopic examination in a child abuse case, lack of histo­
logical examination on a utopsy blocks, failure to instruct
toxicology or microbiol og ica l investi gations are all funda­
mental flaws that never come before a jUly but should be
resolved prior to the court proceedings commencing.
COMMUNICATIONS FROM THE EXPERT
WITNESS
By and large the expert will communicate w ith th e instruct­
ing la wyers and with the courts in two ways: through a
report and , eventually, in person (when evidence is given in
court). In all instances, there should always be absolute clar­
ity in such communicatio ns, to the extent of using very ele­
mentary co mprehensibl e language.
The ex pert 's reports should be dated, properly typed and
form atted, with numbered pages and, when approp riate,
numbered paragraphs. It should start by telling the readers
of th e report'in whi ch matter this report is being produced,
and who has instructed it and w hen. The reader would then
want to be introduced to the report wliter in terms of how
he claims expertise in the matter under revie w, and all the
releva nt qualifications and experience should be listed;
some prefer to attach t his a~ a mini-curriculum lJitae, which
is entirely acceptable, and some suggest that it is preferable.
If the a uthor has been instructed previously in simil a r cases
and has also given evidence in court, this should be stated.
If there are relev ant publications in the name of the expert
these should be cited and, when relevant, copies of these
publications should be attached to the report.
Date of qualifi cation , General Medical Council (GMC)
number and th e memberships a nd fello ws hips of va rious
royal colleges, institutes and learned societies sho uld all be
cited with the date of entry. The cunent post(s) held and
previous relevant posts, with dates of appointments, should
be given. (In Scotland , the date of bilth of the ex pert is a
useful, but not essential, adjunct.) The report writer shou ld
give some deta ils of his expertise in de al ing with th e
specific case, for example the number of cases of child
---
sexual abuse in which he has examined the patient with a
colposcope.
The report will then have to cite what documentary
and other evidence (photographs, histology slides, sketches)
have been available for the compilation of the repOlt. If other
inform ation has been used as background, then this should
be included in the list. The various findings, if relevan t, in the
clinical a nd any other reco rds ava ilable are then pn~c i sed or
cited in fulJ as the case dictates. Obvious errors and inconsis­
tencies have to be referenced, for example if, in a particular,
the nurse's notes refer to a left-sided pneumothorax but the
doctor's record mentions a right-sided pneumothorax. It may
be necessalY to indicate that the evidential bundle available
is inco mplete, for example clinical photographs alluded to are
not included, a sample sent for alcohol analysis has not been
reported on or X-ray reports are not included. It is important
to remember that those with legal training will have already
gone through the notes page by page; it is not co nfidence­
inspiring if the la wyers point out the omission in yo ur repOlt
of what turns out to be a vital bit of information that you
have missed from you r perusal of the clinical notes.
Having laid out all the facts , the co nclusion s should be
derived logically and consistently. Of various co nclusions
that are possibl e, there should be an indication of which
one is the more likely. It should be kept in mind th at when
referring to wounds one must be alert to the fact that
lawyers may wish to know how muc h force was required,
whether the wo und could be considered as life-threa tening
or disfiguring, or whether the wo und could have been acci­
dentally inflicted or indeed se lf-inflicted. The options
should be ranked and not mixed up in random order. One
will be doing a major disservice to the instructing solicitors ­
a nd to oneself - if the conclusio ns are jumbled up in any
order and one has to admit in open court that the probability
of one scenario is 1 in 1000 while tha t of another is 1 in 5.
In coming to conclusions, one may have to cite the lit­
erature - joumals and books (Jatest edition) - to bolster up
the arguments that are being made. These have to be cited
in full. Ph otocopies of releva nt articles and chapters in
books should be appended to yo ur report.
The report has to be si gned, and it is a useful practice, to
indicate that one has read the report carefully before sub­
miss ion and cOITected any typing and other errors in it, to
sig n each p age individually. In spite of this - but obviou sly
much more so without a check - a numb er of howlers do
survive, even after the most meticulous trawling for errors
in the rep ort, and one has to accept these candidly and
apologetically.
DECLARATION BY THE EXPERT IN THE REPORT
It is essential that the ex pelt witness shows that he has
accepted his responsibilities and sbict duties to the COUlt full y
and, as such, it is important that a decl aration is appended at
the end of th e report to enunciate these principles.
- ~.

This ' declaration' should include a statement to the
effect that:
• The report submitted, and any subsequent oral
testimony which may have to be given later under
oath or affirmation, is being given solely to assist the
court and that such an overriding obligation takes full
precedence over any obligations and loyalties the
expert may have to those who have engaged him.
• The report is a tru e and complete re port, and any
opinions expressed in it are correct.
• The report has been produced with all reasonable care
to ensure that it meets the high standards expected
by the courts.
• In further reference to the completeness of the report,
th e expert should indicate that he has cited all the
sources of inform ation on which his report is based,
whether it favours his arguments and opinions or
othelwise.
• The report is independent, unbiased and uninfluenced
by the views of others.
• It is ap preciated that th ere could be cross-examination
on it.
• Finally, it has to indicate that any professional fees due
for expert services rendered by the witness, i.e. for
compiling the report and subsequently presenting
evidence, does not depend in any way on the final
outcom e of the cas e. There have been regrettable
instances in North America of experts producing
favourable reports in return for large remuneration,
particularly in high-profile cases, or in oth er instances
where very large monetary settlements were in the
sights of the lawyers. 12 - 14
IN THE WITNESS STAND OR BOX
It is said that lawyers require experts who can articulate
opinio ns clearly and pe rsuasively in court, can produce a
polished but not patronizing presentation and who do not
simply spout out raw technical detail. A speaker informs,
and indeed persuades, by his, words, by his body language
and through the modulations of his voice at the time in
which he is speaking, with the words actually spoken per­
haps being the least effective communication. Body lan­
guage, in terms of facial expressions, posture, eye contact
and hand movements, is important and conveys signa ls to
the jurors and to the lawyers.
To be able to do this well, you have to be very familiar
with your script, and that includes all the other documen­
tary productions that may have to be referred to in your
evidence. It impresses the jUly tha t you ac tually find imme­
diately the relevant paragraph in your report that refers to
a particular question put to y ou , or th e page in the case
records on which a particularly significant item has been
recorded and, what is more, to be able to assist the court by
helping court officials find the right place.
.. ~ - ­ ­
Pre-trial communication I . 453
Think carefully before you a nswer a question and indeed
allow an appropriate pause. Speak clearly and slowly
enough that the judge's pen can keep up with you if he is
recording the evidence in long hand. If you were un ab le to
hear the question for one or other reason, or you wish the
qu estion to be repeated or rephrased, do not hesitate to say
so. Be careful to a nsw er the specific questions asked and not
the qu estions that you would rather had been asked. The
judge may intervene if you do not play by his rules.
Whatever happens, keep your 'cool' and do not exhibit
bad temper: do not scowl, sneer or sulk. You may be pro­
voked and tried ; thus, this advice is more easily' imparted
than heeded. At the risk of being too nice, do not press your
advantage by chea pening the opinions of others out of hand
or by making cheap points against them, even if they deserve
this. If, however, the question is objective and one is required
to be critical of others, be fair in expressing your opinion, and
give cogent an d competen t reasons for such criticisms.
There is no place for arrogance and overconfidence ; be
patient and courteous. Lord Avonside, in Assessor for Lothian
Region v. Wilson, states that 'It can't be overemphasized that
no committee or, for that matter, no court is in any way bound
to accept expert evidence or to be dictated to by experts'.
There is rarely scope for humour; the old adage that the
judge is the only person in the court room who can make
jokes, and that it is indeed mandatolY to laugh at them, is
well made. The following two utterances by learned judges
may illustrate this matter.
In Rolland v. Lothian Health Board, Lord Ross stated: 'I am
not disposed to place much weight on the evidence of Or H. He
struck me as a witness who in this instance was unable to
approach any question obj ectively; he was intent on defend­
ing his hospital and his staff and I found it difficult to accept
some of the answers which he gave. Professor F gave his evi­
dence in what might not unfairly be described as a didactic
manner; most judges, I imagine, dislike being lectured by w it­
nesses, and in tills respect I count myself among the majority'.
In McVey v. Central Regional Council, RG McEwan sit­
ti ng as a tempo rary judge stated: '... He is a very experi­
enced surgeon and experienced witness. Whether this
makes him unable or unwilling to accept that he could ever
be wrong I know not. However, I find his at ti tud e less tha n
helpful. He was patronizing almost to the point of discour­
tesy with the counsel for the pursuer and at times sarcastic
ad personam over certain proper questions. Wisely counsel
did not allo w herself to be provoked and responded only
with courtesy. I might not have been so patient. For these
and other reasons I did not have any co nfiden ce in his tes­
timony and where it differed from the other medical wit­
nesses, I had no hesitation in preferring them.'
PRE-TRIAL COMMUNICATION
The expet1, to a much lesser extent but increasingly so - and
long may the trend continue - will engage in pre-trial
454 I Expert witness/expe rt testimony
conferences with the side which has instructed him. There, as
in other communication, he should be absolutely dis­
passionate and ruthless but, without fear or favour, should
give all the points in favour and those against. Holding back
or camouflaging the problems at this stage will not do his
side, or the case, any good; when the question is asked in
court, he will have to concede (and blurt out what he may
have politely, but reluctantly, tried to gloss over during a
chummy cup of tea in chambers). The certainty, or otherwise,
of the opinion given should be fully aired with counsel at this
stage; this will go a long way in saving on expensive court
time and, even better, preventing miscarriages ofjustice.
Yet counsel should not assist the expert in any way in
the preparation of his report. Lord Wilberforce at 256 in
Whitehouse v. Jordan stated: 'While some degree of consul­
tation between experts and legal advisers is entirely proper,
is it not necessary that expert evidence presented should
be, and should be seen to be, the independ ent product of
the expert, uninfluenced as to form or content by the exi­
gencies of li tigation. To the extent it is not, the evi dence is
likely to be not only incorrect but also self-defeating'.
CONFLICT OF INTEREST
The 'other' side in any forensic case may also wish to speak
with you before the trial. Answer all the questions asked
by them truthfully and honestly, and be willing to admit, if
required , that certain aspects of the case had not been pre­
viously familiar to you and that on the basis of this fresh
information there may be scope for revising the opinions
already offered. The expert may find that he may thu s
inadvertently be sucked into a conflict of interests l 5 . 16
between the parties involved. Perhaps the earthy statement
of the Unites States Supreme court in Williams v. Florida
should be kept in mind in this respect: 'The adversary sys­
tem of trial is hardly an end in itself; it is not a poker game
in which the players enjoy an absolute right always to con­
ceal their cards until played '.
In some other instances, there could be a more genuine
conflict of interests in the m,ind of the expert, for example
when giving evidence in a child abuse case knowing
fuJJ well that the evidence may point to a member of the
child's family, also a patient of the practice, and that there
may be a disruption of that family unit or an unfavourable
divorce settlement. The expert, thus, comes to lie on the horns
of a clilemma; should his allegiance to the child override loy­
alty to other patients - the parents? The child's welfare may
have to override obligations of medical confidentiality to the
child (and his family) in the interests of public safety.
RULES OF EVIDENCE
In the courtroom certain evidence can be excluded in
certain well-defined circumstances. The policy reasons for
such exclusions are basically ones in which confusion and
unfairness can be avoided; this will also prevent the judge or
jury, especially the latter, from falling into error. Over the
years, the rules of evidence have been relaxed; in Scotland, it
was only in 1898 that the accused could give evidence before
thejury trying him. In this respect, doctors often utilize infor­
mation obtained at second hand from others in coming to
their decisions, i.e. hearsay evidence. As a generality the
court will accept only the 'best evidence'. What constitutes
hearsay is a legal matter to be determin ed by the judge.
CONCLUSION
Perhaps the classic statement of Lord President Cooper as
found in Davie v. Magistrates of Edinburgh sums up best
the scope of expert evidence:
Expert witnesses, however skilled or eminent, can give
no more than evidence. They cannot usurp the func­
tions of the jury or judge sitting as a jury, any more
than a technical assessor can substitute his advice for
the judgement of the court ... Their duty is to furnish
the judge or jury with the necessary scientific criteria
for testing the accuracy of their conclusions, so as to
enable the judge or jury to form their own independent
judgement by application of these criteria to the facts
proven in evidence. The scientific opinion evidence, if
intelligibl e, convincing and tested, becomes a fa ctor
(and often an important factor) for consideration along
with the whole other evidence in the case, but the deci­
sion is for the judge or jury. In particular, the bare ipse
dixit of a scientist, however eminent, upon a contro­
versial issue, will normally carry very little weight, for
it cannot be tested by cross exa mination nor independ­
ently appraised, and the parties have invited the deci­
sion of a judicial tribunal and not an oracular
pronouncement of an expert.
RECENT DEVELOPMENTS
In Kariing v. Purdue (2004 SLT 1067), the Scottish courts were
asked to consider a claim for damages against an expert wit­
ness. Mr Karling had been convicted of murder in 1995. Dr
Basil Purdue was the defence expert asked to carry out a
post-mortem examination of the victim. Dr Purdue's findings
were consistent with those of the prosecution case. Mr Kar­
ling was convicted and sentenced to life imprisonment. In
200], fresh scientific evidence came to light indicating that
there was no sound evidence on which to conclude that the
victim had been suffocated. In an appeal, Mr KarJing's con­
viction was quashed. Subsequently, Mr Karling issued civil
proceedings against the expert. He sought damages of
£75000 on the grounds that, had the expert performed the
post-mortem examination with sufficient care and compe­
tence, there would have been a reasonable chance that he
wo uld have been acquitted of the charge. Purdue defended

the actio n on the grounds of his abso lute immunity from suit.
He argued that, in the absence of malice, no li ability could be
attached to him, regardless of how incompetently th e post­
mOliem or the subsequent report was carried out.
Mr Karling's respon se to this pl eading was that, by indi­
cating in his rep Oli that he was attaching his conclusions
sepa rately, Dr Purdue was effectively adm itting that part of
his report was intended for Mr Karling's le ga l team only. He
argued further that this suggested that the expeli was act­
ing as an advisor, therefore owing Mr Karling a duty of
care that was capable of standing separately from Dr Pur­
due's role as expert witness. In rejecting these arguments,
the co urt said that the link in tim e and function with the
criminal proceedings was more than suffici ently close to
conclude that, in relation to his engagement to perform
professional or expert services, Dr Purdue was immune
from suit in Mr Karling's instance. The test imposed by the
courts was whether the expert was preparing to give evi­
dence in proceedings. It was the proximity of th e criminal
proceedings that, in effect, made Dr Purdu e's 'advisory'
capacity indistinguishable from the work done in the pro­
vision of services as an expert witness.
Another more recent impoltant development in the field
of expert evidence was the decision of the Court of Appeal
in ThE GEn eral Medical Council v. ProfEssor Sir Roy
Meadow (Her Majesty's Attorney Generol iJ1ter[Jel1ing)
(GMC v. Meadow) . In November 1999 , Sally Cl ark was tried
for the murder of her two sons. Th e Crown relied. in palt,
upon Professor Meadow's evidence to refute the proposi­
ti on that Mrs Cl ark 's children may have died fro m sudden
infant dea th syndrome (SIDS), or co t death. Professor Sir
Roy Meadow ventured into the field of statistics in givin g
hi s opinion on the likelihood of occurrence of a second cot
dea th in the same family; this invo lvement, sa id to have
been erroneous and outs ide his field of expertise. resulted
in proceedings before the fi tn ess to practise panel (FPP) of
the GMC, with Professor Mead ow being found gUilty of
serious professional misconduct and res ultin g in his era­
sure from the medical regis ter. Professor Meadow appealed
to the High Court, whe re his appea l was a llo wed by Collins
J. The finding of serious professional misconduct was
qu ashed . Additionally, the judge commented that expel1
witnesses, by their status, enjoyed not merely immunity
from civil suit, but also a wide, albei t not absolute, immu­
nity from disciplinary, regulatory or fitness to practi se pro­
ceed ings in relation to statements or ev id ence given by
them in or for the purpose of legal proceedings. These com­
ments resulted in the Attorney General's interventi on to
the COUlt of Appea l; it held unanimously that the immu­
nity identified by the judge did not exist.
REFERENCES
Di ckson WG. A Treiltise all the Law of Eliidel1ce ill Scorlaml,
3rd edn. Edinburgh, UK: Scottish Law Commiss ion, No. 1887,
137 .4, paras 397 el sel].
References I 455
2 Lav in JH. Ev elyon e believed the pl ainti ff except his doctor.
Med Ecol/ 1991; 68:34-4 1.
3 Kaufmann HH. Th e expert witn ess. Neither Frye nor Daubert
so lved th e prob lem: what ca n be don e? Sci Jmtice 200 1;
41 :7-20.
4 Calvert-Smith D. Medica l evidence in criminal prosecutions.
l'vI ed Leg J 2001 : 68 :11 7-2 9.
5 Ca lvert- Smi th D. Medi cal evidence in criminal prosecutions.
Med Leg J 2000: 68:117.
6 Starrs JE. In the lan d of agog : an allegory for the exp ert
witn ess. J forellsic Sci 1985; 30:289-307.
7 Annas GJ. Scientific ev iden ce in the court room; the death of
the FI")'e rule . N £lI gl J Med 1994: 330 : 101 8- 2 1.
8 Mi tch ell J. Scientific evi den ce: pri nciple or pragmatism .
Aied Leg J 200 1: 69: 117- 19.
9 Ed itorial - Osteo genesis imperfecta, non-accidental injury and
temporary brittle bo ne disease . Arch Dis Child 1995: 70:169.
10 Brandon S. Recovered memories; some aspects of the
controversy. Med Leg 199 9; 67: 25 -34.
1J Brahams D. 'Repressed memories' and the law. La ncet 2000;
356:358.
12 Katz J. The fallacy of the impart expert rev isi ted . Bull Am
Aca d PS),ciliatr Law J9 92 ; 20:141-5 2.
13 FacUo D. Buccia relli R. Peer review of the expe rt wi tness: an
opportunity to improve our medical liability system. J Child
Neu rol 1995: 10:403-4.
14 Fisher CWo Dombrowski MP, Jaszczak SE et al. Th e expert
witness: reo l issues and suggest ions. Am J Ob ster GYl1ecol
1995: J 72: 1792-800.
15 Dona ldso n L. Kap lon C, Leung W-c, The medica l expert
witness : time to regulate conflicts of interes t. Med Sci Law
1999 : 39 :1 1- 16.
16 Ma narin B. Assess in g the expen:: a ca ll for reciprocal
di sc lo sure in Canada. Med Sci Law 1999 ; 39: 17 -22.
Further Reading
CASES
• Nariollal Justice Campania Nau iera SA V. Prudelltial Assurance
Co. Ltd, rlre Ikar iall Reefer (1993) 2. L1 0yds Reports 68, 8 1-82
and (19 95) 1. L1 0yds Reports 45 5.
• R V. Srockwell (1993) 97 CR App R 260.
• R v. Uljoh Et Orhers (2001) EWCA Crim 1381.
• Re H an d E (Minors) CO Ult of Ap pea l (1996), 21 May.
• R V. BO II,\'rllOli (1984) 38 SASR 45.
• flye V. United Srates. 54 App. D.D. 46, 293 F. 1013 . 1923.
• Dauberr v. iVlelTeli Do /U Phollllll ceu ricilis Inc (1993)
113 S Ct 2786.
• KUlllbo Tire Co Lrd V. Carmichael (1999) 526 US 13 7.
• R v. Robb 93 Cr App R 161.
• Re AB (cliild abuse; expert lVili/esses) (1995) 1 FLR 181.
• Daniels v. Walker (2 000) 1 WlR 1382, CA.
• A ssess or fo r Lothian Region v. Wilso11 (1979) SC 34 1 lord
Avonside at 349.
• Rollalld v. Lothian Health Board, 27th August 1981 - unreported.
• McVey v. Central Regiona l Co ull cil, 4th Ma rch 19 93 - unreported.
• White/lOuse v. Jordan (198 1) 1 WlR 246.
• Williams V. Florida (1970) 399, US 78.
• Davie V. Magistrates of Edillburgh (1953) SC 34 at 40.
• Karling v. Purdue (2004) SLT 1067 Coun of Session, Edinburgh.
• Tire Gelleral iVIed ical Council V. Professor Sir Roy Meadow (Her
Majes t."·s Artomey General ilitemel1iIJg) (GMC v. Meadow)
Neutral Orar ion Nu mb er: (2006) EWCA Civ 13 90 Case No:
CO/5763/2005.
 I
I
APPENDIXA

CHILD PROTECTION

EXAMINATION FORMS

Form 1: Examination consent form

CHILD PROTECTION EXAMINATIONS

CONSENT FORM

Child's name ..... .... ......... ... .. ............ ... .. .. ..... ..... ....... ........ .... .. ...... .. ...... .... ..... .. ...... .... .OOB ...... .. ... .... .. ...... .. ........... .

Address ... ............ ... ... ......................... .. .......... ... ... ......... ... ........... .. ................ ............ ... ... ....... .. ... ............... ...... ..... .

Unit no ..... ..... .... .... .. ..... ...... ... ...... ...... .. .... ......... ....... .. .. ... ....... ......... .. .... ..... .. .............. ... ....... .. ...... .. ........... .. .......... .

Permission must be obtained from parent(s) or other(s) with responsibility for the child, and from the child where
appropriate.

I give permission for:

1. Medical examination Yes No N/A
2 . Collection of specimens for laboratory tests Yes No N/A
3. Photography of clinical findings Yes No N/A
4. CO of genital findings Yes No N/A

Photographs and CDs will be stored as part of the clinical records. They may be used to support clinical evidence
of injury and may need to be shared with another Doctor involved in any court proceedings.

I give permission for photographs and CDs to be used to support clinical evidence in court proceedings.
Yes No N/A

Photographs and CD recordings can be used for teaching and training of other professionals working in Child
Protection proceedings. Photographs and CDs used for this purpose are anonymized.

I give permission for anonymized photographs and CD recordings of my child to be used for teaching
and training of other professionals. Yes No N/A

I understand that the information from the medical examination will be shared with:

Verbal report Letter

Social services Yes/ No Yes/No
Police Yes/No Yes/No
GP Yes/ No Yes/No
Health visitor/school nurse Yes/No Yes/No

The procedure has been fully explained to me and I understand that I have the right to withdraw my con­
sent at any stage during the procedure.

Signed .......... ........ ....... ..... ........ ....... ..... .. ... ...... ....... ..... .. .. ........ .. ...... .. ... ... ............ Date ........ ... ...... .. ... .. .. .. .. ......... ... ..

Name ............. ...... .... ................ ..... ..... ..... .... ....... ...... ... ....... .. .... .... ....... ..... .......... ..Parent/carer/professional

Examining doctor(s) ... .... ........ .. ..... .. ......... .. .............................................. .. .. ......... ... ............ .. ... .. ...... ... ............. .... ..

Signature .... ....... ..... ....... ... ............ ... ......... ...... .. ......... .... .... ...... ............... ...... ....... Oate .......... .... ....................... .... .. .

458 I Appe ndix A: Child protection examin ation forms

Form 2: Examination form

Confidential Medical Information

EDINBURGH AND LOTHIANS

CHILD PROTECTION OFFICE

Joint Paediatric/Forensic Examination/

Special ist Paediatric Examination of

Child who may have been Abused or Neglected

Place of examination Name of examiner(s)

Child 's surname J Forename(s)

Known as

Address Date of bi rth

Sex

CHI no.

GP Date of examination

Address
Time of examination

Child accompanied to clinic by

(please tick)
Mother D Father D sw D Police D Other D
School/nursery attended

BACKGROUND INFORMATION
(e.g. previous concerns reo developmental delay, poor growth, possible episodes of NAI)

Family/social history
N
o
o
N
<D
.c
o
U
o
;?
~--------------------------------------------------------------------~
Page 1 of 12 1471
 Appe ndix A: Child protection exam ination forms I 459

ACCOUNT OF CIRCUMSTANCES LEADING TO REFERRAL

(a) From referrer :

Name: Position:

(b) From accompanying adult:

Name: Position :

(c) From child:

Child 's terminology
In the presence of:
Breasts
Vagina
Anus
Urethra
Abdomen
Semen
-
Penis

Scrotum

CONSENT

(source, i.e . parent, young person, person holding parental righ ts)

Date: N
o
o
N
Witnessed by: Q)
.D
.8
Name Position u
o
L ­_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _~~
o

Page 2 of 12 1471
460 I Append ix A: Child protect ion exa mination forms

PRESENTING SYMPTOMS

Has the child exhibited any behavioural symptoms:

Masturbation D Secondary wetting D Aggression/anger D Sexualized play D

Soiling D Conduct disorder D Other D

SYSTEMATIC ENQUIRY

N
o
o
N
Q)
.0
(Genitourinary enquiry) o
t5
see over o
~--------------------------------------------------------------------------------~~
Page 3 of 12 1471
 Appendix A: Child protection examination forms I 461

GASTROINTESTINAL/GENITOURINARY ENQUIRY

Comment
Bowels Frequency

Consi stency of stool

Pain on defaecation Yes/No

Bleeding on defaecation/blood on wiping Yes/No

Con stipation Yes/No

Soiling Yes/No

Mucus Yes/No

Other Yes/No

Urinary tract Previous UTI (suspected) Yes/No

(con firmed ) Yes/No

Frequency Yes/No

Dysuria Yes/No

Urgency Yes/No

Wetting - day/night Yes/No

Other Yes/No

Genital Bleeding/blood on pants Yes/No

Di scharge Yes/No

Itch Yes/No

Rash/inflammation Yes/No

Circumcised Yes/No
-
Menarche age

Menstrual cycle

Date of last period

Routine sanitary protection Tampon/towel/both

N
o
o
N
Q;
.0
o
ti
o
o
~

Page 4 of 12 1471
462 I Appendix A: Child protection examination forms

CLINICAL EXAMINATION

People in attendance during examination

General physical appearance of child (note especially any evidence of infection, neglect or injury)

Demeanour/behaviour/impression of maturational status

Measurements

Weight kg centile

Height cm centile

Head circumference cm centile

Findings on externa l physical examination

(Tick appropriate boxes and where possible record findings. Omit if Preliminary Medical Assessment Form
has already been completed.)
Comment
(a) Skin and hair

(b) Teeth

(c) Eyes

(d) Ears, nose and throat

(e)
- Cardiovascular system

(f) Blood pressure (if applicable)

(g) Respiratory system

(h) Alimentary system

(i) Genitalia/testes

(j) Nervous system

(k) Locomotion/posture Ig
(I) Other I~
Q;
.c
0
0
0
0
~

Page 5 of 12 1471

--~. ­
 Appendix A: Child protection examination forms I 463

.....--­ ,,----­
®' I@
~
l.\
.....c:::::'..: "4
,---,

C\l
o
o
C\l
Q;
.0

-----------------------------------p~~~~~--------------------------------~~

...
Page 6 of 12 1471 .
464 I Appendix A: Child protection examination forms

(\J
o
o(\J
Qj
.0
o
o
o
o
~--------------------------------------------------------------------~
Page 7 of 12 1471

-----­
- --­ - -----------­--------------­
 Appen dix A: Ch ild prot ecti on examinati on forms I 465

-
'"-'

N
o
o
N
Q;
.0
o
t5
o
~------------------------------------------------------------------------------~~
Page 8 of 12 1471
466 I Appendix A: Child protection examination forms

Mons pubis

Prepuce of clitoris

Clitoris

Frenulum of clitoris

Orifice of urethra
\ Labium majus (labia majora)
\
I
I
Vestibule (area between
labia minora)

) Labium minus (labia minora)

/ Hymen (covering introitus

of vagina)

Vest ibular fossa

Fourchette

Posterior commissure

Genital examination
Position
Method
Magnification __________

Video Yes/No

Identification:

Video no. I
Serial no.
.'.~:- '
J
C\J
o
o
C\J
OJ
.D
o
tl
o
o

Page 9 of 12 1471

-- - ---
 Appendix A: Chi ld protection examination forms I 467

u\
Genital examination

Position
Method
Magnification
Video Yes/No
Identification:
Video no.
Serial no.

Examination of anus

Position
Method
Magnification

u
Video Yes/No
Identification:
Video no.
Serial no.
(\J
o
o
(\J

Q;
.D
.2
u
o
o
~--------------------------------------------------------------------------------~~
Page 10 of 12 1471
468 I Appendi x A: Child protection examin ation forms

INVESTIGATIONS

D Photography

Taken by:

Police:

Processed at:

Medical staff:

Identifier:

D Radiographs
D
Coagulation profile

Routine specialist:

D Serum HIV/hepatitis

D Swabs

Site:

D Urine
D
Pregnancy test

D Other

CONCLUSION/OPINION

TREATMENT/ACTION REQUIRED

C\J
o
o
C\J
Q;
.D
o
t5
o
o
~------------------------------------------------------------------------------~~
Page 11 of 12 1471
 Appendix A: Child protection examination forms I 469

INFORMATION SHARING

Time Name Present tel. no.

Parent PresenUcontacUno contact

Social worker PresenUcontacUno contact

Police Present/contacUno contact

General practitioner ContacUno contact

HV/school nurse ContacUno contact

AGREED PLAN

1. Further investigation of possible abuse.

2. Need for further assessment of medical/developmental problems.

(NB: If child is looked after/accommodated, Part 2 of the 'Initial Medical Assessment' will be required.)

3. Other action required: 1. Admit to hospital

2. Refer for Child Protection Order
Circle

3. Refer for Child Assessment Order

as 4. Refer to Reporter

appropriate
5. Discuss with Procurator Fiscal
6. Contact GP
7. Other (specify)

4. Report sent to: 1. General practitioner

2. Social worker
Circle 3. Reporter

as

appropriate 4. Procurator fiscal

5. Hospital records

Signed Date

Name in block letters Designation Time completed

Signed Date

Name in block letters Designation Time completed

AUDIT

(\J
o
o
(\J
Q)
.a
o
tl
o
o
~--------------------------------------------------------------------------------~~
Page 12 of 12 1471
Table 1 Fetal autopsy standards at 72-27 weeks of gestation'
Age (GW) FL CR CH HOC Body Brain Liver Lungs Heart Thymus Spleen Kidneys Adrenals

Maceration" 0-3 0-3 0-3 0-3 0-3 0-3 0-1 2 3 0-1 2-3 0-3 0-1 2 3 0-1 2-3 0-1 2-3 0-1 2-3
(mm) (em) (em) (em) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g)

~
12 Mean 9 7.4 9.8 7. 1 29.6 4.8 1.5 1.4 1.3 0.6 0.9 0. 10 0.D3 0.01 0 .25 0.19 0.04 0.11
SO 3 7.7 7.7 7.7 74.9 7.4 7.2 7.2 7.2 0.9 0.9 0. 14 0.06 0.02 0.15 0.75 0.78 0.18
13 Mean 12 8.7 11.8 8.5 37.4 6.5 2.0 1.7 1.2
to
SO
1.7 1.2 0.20 0.04 0.02 0.08 0.3 0.2 0.17 0.17
r
3 7.2 7.8 7.2 74.9 7.4 7.2 7.2 7.2 0.9 0.9 014 006 003 0.03 01 0.1 0.18 0.18 m
14 Mean 15 9.9 13.7 9.8 53.0 9.1 2.9 2.4 2.3 2.0 1.5 0.3 0.05 007 0.05 0.Q4 0.14 0.4 0.3 0.3 0.2 Vl

15
SO
Mean
3
18
7.2
11 .1
7.8
15.6
7.2
11 .1
74.9
76.5
2.5
12.7
7.2
4.2
7.2
3.3
7.2
3.2
0.9
2.9
0.9
2.1
0.1
0.5
0.06
0.07
0.06
0.08
0.06
0.06
0.04
0.06
0.04
0.17
0.7
0.6
0.7
0.5
02
0.5
0.2
0.3
o
'"T"j
SO 3 7.2 7.8 7.2 78.5 3.9 7.2 7.2 7.2 0.9 0.9 07 0.06 0.06 0.06 0.06 0.06 0.3 0.3 02 0.2 Vl
16 Mean

~~
21 12.4 17.5 12.4 108 17.3 5.9 4.5 4.2 3.9 2.7 0.8 0.11 0. 12 0.09 0.09 0.17 0.9 0.8 0.6 0.4
SO 3 7.3 7.8 7.3 47 5.4 7.5 7.5 7.5 7.2 1.2 0.2 0.06 0.06 0.06 0.08 0.08 0.4 04 0.3 0.3
17 Mean 24 13.5 19.3 13.6 147 22.9 8. 1 6.1 5.4 5.1 3.5 1.0 0.18 0.18 0.12 0.13 0.16 1.3 1.1 0.8 0.5
SO 3 7.3 7.9 7.3 53 6.9 3.0 3.0 tJ~
18 Mean 27 14.7 21.1 14.8 194 29.4 10.7 7.9
3.0
6.8
7.7
6.4
1.7
4.4
0.4
1.4
0.06
0.3
0.06
0.3
0.06
0.2
0.12
0. 19
0.72
0.15
06
1.8
06
1.5
04
1.0
04
0.7 »m
19
SO
Mean
SO
3
30
3
7.3
15.9
7.3
7.9
22.9
7.9
7.3
16.0
7.3
65
249
78
8.4
37.0
9.8
4.5
13.8
6.0
4.5
10. 1
6.0
4.5
8.4
6.0
2.3
7.9
2.8
2.3
5.4
2.8
0.5
1.7
0.7
0.2
0.4
0.3
0.2
0.4
0.3
02
0.3
0.3
0.17
0.3
0.2
0.17
0.15
0.22
0.8
2.4
1.0
0.8
2.0
1.0
0.4
1.2
0.5
0.4
0.8
05
b6 I ---'

20 Mean 33 17.0 24.6 17.2 312 45.5 17.2 12.5 10.2 9.5 6.5 2.1 0.6 0.5 0.3 0.4 0.17 3.0 2.5 1.4 1.0 ~><
SO 3 7.4 7.9 7.4 92 77.3 7.5 7.5 7.5 3.4 mto
»
3.4 08 0.4 0.4 04 0.3 0.29 1.2 1.2 0.6 0.6
21 Mean 36 18.2 26.3 18.3 382 55.0 21.1 15.2 12.3 11 .2 7.8 2.6 0.8 0.7 0.4 0.5 0.22 3.8 3.1 1.7 1.2
SO Vl
22 Mean
3
39
7.4
19.3
20
28.0
7.4
19.4
707
461
728
65.4
9.0
25.5
9.0
18.2
9.0
14.5
4.0
13.1
4.0
9.2
1.0
3.1
0.5
1.0
0.5
0.9
0.5
0.6
0.4
0.7
0.36
0.3
7.4
4.6
7.4
3.8
0.7
1.9
0.7
1.4 c
23
SO
Mean
3
41
7.4
20.4
2.0
29.6
7.4
20.5
722
547
74.3
76.9
70.4
30.2
70.4
21.6
70.4
16.9
4.6
15. 1
4.6
10.7
1.7
3.6
0.6
1.3
0.6
1.1
0.6
0.7
0.4
0.9
0.4
0.4
1.6
5.5
1.6
4.6
08
2.2
08
1.6
~
SO 4 7.5 2.0 7.4 722 75.8 77.9 77.9 77.9 5.3 5.3 1.3 0.8 0.8 08 0.5 0.5 7.9 1.9 08 0.8 ~
24 Mean 44 21.5 31.2 21.6 641 89.3 35.4 25.2 19.5 17.3 12.4 4.2 1.6 1.3 0.8 1.1 0.6 6.5 5.5 2.5 1.8
m
SO 4 7.5 2.0 7.5 737 77. 2 73.4 73.4 73.4 5.9 5.9 7.4 09 0.9 0.9 0.6 0.6 2.1 2.1 09 09 Z
25 Mean 47 22.6 32.8 22.6 743 103 41.1 29.1 22.3 19.6 14.1 4.9 1.9 1.6 1.0 1.4 0.8 7.6 6.4 2.8 2.0 --l
Vl
SO 4 7.5 2.7 7.5 754 79 74.9 74.9 74.9 6.6 6.6 1.6 1.1 1.1 1.1 0.7 0.7 2.4 2.4 7.0 1.0
26 Mean 50 23.6 34.3 23.6 853 11 7 47.1 33.4 25.3 22.0 16.0 5.6 2.3 1.9 1.2 1.7 1.1 8.8 7.4 3.1 2.3
SO 4 7.5 2.7 7.5 77l 20 76.4 76.4 76.4 7. 3 7.3 1.7 7.2 7.2 7.2 0.9 0.9 2.7 2.7 1.1 1.1
27 Mean 52 24.7 35.8 24.5 971 133 53.6 37.9 28.6 24.6 18.0 6.3 2.6 2.2 1.4 2. 1 1.4 10.1 8.4 3.4 2.5
SO 4 7.6 2.7 7.5 788 22 77.9 77.9 77.9 8.0 8.0 1.8 1.4 1.4 1.4 1.0 1.0 3.0 3.0 1.2 7.2

CH, crown-heel lengt h; CR, crow n-rump length ; FL, foot length ; GW, weeks of gestation; HOC, head circumference; SO, standard devi ation.
"0, none; 1, mild; 2, moderate; 3, marked.
 ""
-.j
I\)

Table 2 Fetal autapsy standards at 28-43 weeks af gestatian 1

»
"0
"0
Age (GW) FL CR CH HDC Body Brain Liver Lungs Heart Thymus Spleen Kidneys Adrenals C1l
:::J
Q

Maceration' 0-3 0-3 0-3 0-3 0-3 0-3 0-1 2 3 0-1 2-3 0-3 0-1 2 3 0-1 2-3 0-1 2-3 0-1 2-3 x'
rn
(mm) (em) (em) (em) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g)
n;i
IT
C1l
(fl

28 Mean 55 25.7 37.3 25.5 1096 149 60.6 42.7 32.0 27.4 20.2 7.1 3.1 2.5 1.6 2.5 1.8 11.4 9.6 3.7 2.8 S­
en
SO 4 1.6 2.2 1.6 206 23 19.3 19.3 19.3 8.7 8.7 2.0 1.6 1.6 1.6 1.1 1.1 3.3 3.3 1.3 1.3 ill
:::J
29 Mean 57 26.7 38.7 26.4 1230 166 67.9 47.8 35.6 30.2 22.5 7.9 3.5 2.9 1.8 3.0 2.2 12.9 10.8 4.1 3.1 Q
Cll
SO 4 1.6 2.2 1.6 225 25 20.8 20.8 20.8 9.5 9.5 2.1 1.8 7.8 1.8 1.3 1.3 3.6 3.6 1.4 1.4 0.
30 Mean 60 27.7 40.1 27.2 1371 185 75.7 53.3 39.4 33.2 24.9 8.7 4.0 3.3 2.1 3.6 2.7 14.4 12.1 4.5 3.4 :3
C1l
Cll
SO 4 1.6 2.2 7.6 244 26 22.3 22.3 22.3 10.2 10.2 2.3 2.1 2.1 2.1 1.4 1.4 3.9 3.9 1.4 1.4 (fl
C
31 Mean 62 28.7 41.4 28.1 1520 204 83.9 59.0 43.4 36.3 27.4 9.6 4.5 3.7 2.3 4.2 3.3 16.0 13.4 4.8 3.8 CD
:3
SO 4 1.7 2.2 1.7 264 28 23.8 23.8 23.8 lJ.0 lJ.0 2.4 2.3 2.3 2.3 1.6 1.6 4.3 4.3 1.5 1.5 C1l
:::J
32 Mean 64 29.7 42.8 28.9 1677 224 92.6 65.0 47.6 39.6 30.0 10.6 5.0 4.2 2.6 4.8 3.9 17.7 14.9 5.2 4.1 u;­
SO 4 1.7 2.3 1.7 285 29 25.3 25.3 25.3 lJ.8 ll.8 2.6 2.5 2.5 2.5 1.8 1.8 4.6 4.6 1.6 1.6
33 Mean 67 30.6 44.0 29.7 1842 245 102 71.3 52.1 43.0 32.8 11.6 5.6 4.6 2.9 5.5 4.5 19.5 16.4 5.6 4.5
SO 4 1.7 2.3 1.7 306 31 27 26.7 26.7 12.6 12.6 2.7 2.8 2.8 2.8 1.9 7.9 5.0 5.0 1.7 1.7
34 Mean 69 31.6 45.3 30.5 2015 268 111 77.9 56.7 46.6 35.7 12.6 6.2 5.1 3.2 6.3 5.2 21.4 18.0 6.0 4.8
SO 4 1.8 2.3 1.7 328 32 28 28.2 28.2 13.5 13.5 2.9 3.1 3.1 3.1 2.1 2.1 5.4 5.4 1.8 1.8
35 Mean 71 32.5 46.5 31.2 2195 291 121 84.8 61.5 50.3 38.7 13.7 6.9 5.7 3.5 7.2 6.0 23.3 19.6 6.5 5.2
SO 5 1.8 2.3 1.8 350 33 30 29.7 29.7 14.3 14.3 3.0 3.3 3.3 3.3 2.3 2.3 5.8 5.8 1.9 1.9
36 Mean 73 33.4 47.7 31.9 2383 315 132 92.1 66.5 54.1 41.9 14.8 7.5 6.2 3.8 8.1 6.7 25.4 21.4 6.9 5.6
SO 5 1.8 2.4 1.8 373 35 31 31.2 37.2 15.2 15.2 3.2 3.6 3.6 3.6 2.5 2.5 6.2 6.2 2.0 2.0
37 Mean 76 34.3 48.9 32.6 2580 340 142 100 71.7 58.1 45.1 16.0 8.2 6.8 4.2 9.1 7.5 27.5 23.2 7.4 6.0
SO 5 1.8 2.4 7.8 397 36 33 33 32.7 16.1 16.1 3.3 3.9 3.9 3.9 2.7 2.7 6.6 6.6 2.1 2.1
38 Mean 78 35.2 50.0 33.2 2784 366 154 107 77.2 62.2 48.5 17.2 8.9 7.4 3.9 10.1 8.3 29.8 25.0 7.8 6.5
SO 5 1.9 2.4 7.8 421 38 34 34 34.2 17.0 17.0 3.4 4.2 4.2 4.2 3.0 3.0 7.1 7.J 2.2 2.2
39 Mean 80 36.1 51.1 33.8 2996 394 165 116 82.8 66.5 52.1 18.5 9.7 8.0 5.0 11.2 9.1 32.1 27.0 8.3 6.9
SO 5 1.9 2.4 7.9 446 39 36 36 35.6 18.0 18.0 3.6 4.6 4.6 4.6 3.2 3.2 7.5 7.5 2.3 2.3
40 Mean 82 37.0 52.1 34.4 3215 422 177 124 88.6 70.9 55.7 19.8 10.5 8.6 5.4 12.4 9.9 34.5 29.0 8.8 7.4
SO 5 1.9 2.5 7.9 471 41 37 37 37.1 18.9 18.9 3.7 4.9 4.9 4.9 3.4 3.4 8.0 8.0 2.4 2.4
41 Mean 84 37.8 53.1 35.0 3443 451 190 133 94.6 75.4 59.5 21.2 11.3 9.3 5.8 13.7 10.7 37.0 31.1 9.3 7.9
SO 5 7.9 2.5 1.9 497 42 39 39 38.6 19.9 19.9 3.9 5.3 5.3 5.3 3.7 3.7 8.4 8.4 2.5 2.5
42 Mean 86 38.6 54.1 35.5 3678 481 203 142 101 80.1 63.4 22.5 12.2 10.0 6.2 15.0 11.5 39.6 33.3 9.9 8.4
SO 5 2.0 2.5 2.0 524 44 40 40 40 20.9 20.9 4.0 5.6 5.6 5.6 4.0 4.0 8.9 8.9 2.6 2.6
43 Mean 88 39.4 55.0 36.0 3922 512 216 151 107 84.9 67.4 24.0 13.1 10.7 6.6 16.4 12.2 42.2 35.5 10.4 8.9
SO 5 2.0 2.5 2.0 551 45 42 42 42 27.9 21.9 4.2 6.0 6.0 6.0 4.2 4.2 9.4 9.4 2.7 2.7

CH, crown-heel length; CR, crown-rump length; FL, foot length; GW, weeks of gestation; HDC, head circumference; SD, standard deviation.
'0, none; 1, mild; 2, moderate; 3, marked.
 Appendix B: Tables of sta ndard measu rements I 473

Table 3 Mean weights and percentiles for singleton placentas 2

Gestational 90th 75th Mean singleton 25th 10th No. of cases

age (weeks) percentile percentile placental wt. percentile percentile

21 172 158 143 128 114 3

22 191 175 157 138 122 6
23 211 193 172 151 133 7
24 233 212 189 166 145 9
25 256 233 208 182 159 19
26 280 255 227 200 175 14
27 305 278 248 219 192 9
28 331 302 270 238 210 16
29 357 327 293 259 229 11
30 384 352 316 281 249 12
31 411 377 340 303 269 14
32 438 403 364 325 290 24
33 464 428 387 347 311 30
34 491 453 411 369 331 32
35 516 477 434 391 352 44
36 542 501 457 412 372 36
37 566 524 478 432 391 32
38 589 547 499 452 409 62
39 611 567 519 470 426 103
40 632 587 537 487 442 193
41 651 605 553 502 456 87

700,-----r----,-----,----,-----r----.----~----r_-- __----~----~
90th percentile
600~----r----+----~----+---~~---+----~----~--_4~~~~--~ 75th percentile
Mean singleton
500 ~----r----+----~----+---~-----+----~~_=~--~~~~~--~ 25th percentile
10th percentile

200~--~~~~~~~~~q-----~---+----~----+_--~----_+-----

100~----+----1----~----+-----~---+----~----+_--~----~----~
22 24 26 28 30 32 34 36 38 40
Ges tat ion [weeks)

Singleton placental weights and ranges.

 474 I Appendix B: Tables of standard measurements

Table 4 Umbilical cord length from 20 to 47 weeks' gestation 3

Gestational age n Umbilical cord

(weeks) length (cm)

20-21 16 32.4 :!: 8.6

22-23 27 36.4 :!: 9.0
24-25 38 40.1 :!: 10.1
26- 27 59 42.5:!: 11 .3
28- 29 80 45.0 :!: 9.7
30-31 11 3 47.6:!: 11.3
32-33 337 50.2 :!: 12.1
34-35 857 52.5 :!: 11.2
36-37 3153 55.6 :!: 12.6
38-39 10083 57.4 :!: 12.6
40-41 13 841 59.6 :!: 12.6
42-43 4797 60.3 :!: 12 .7
44-45 1450 60.4 :!: 12.7
46-47 492 60.5:!: 13.0
Total no. of cases 35779. Data represent mean ::': 1 SD. n = no. of cases.

---- -------- =-~-:~ ~ -

-+

Appendix B: Tables of standard measurements I 475

Table 5 Body weight, height and head circumference, boys, birth to 1 year. (Courtesy of the Child Growth
Foundation. Printed by Harlow Printing Ltd, Maxwell St, South Shields, NE33 4PU.j

32 31 3'6 3'814'0 4 2 -~4 46

O-lyr 28 3
I '7 ' 8 • I rfj}
I j -1
, II
8 5 0 52

I 51
NAME ..... ..................... ................ ..

D.O .B. ..... .I ...... ! ...... II I I I I I i.

I
II __ - :: :: - - -llJ-1 ____
99.•• , "

98 th -
.1SI -1
50
49
Correct by......... ..... for prematurity , I
---l1:=.: - ­
:~: -, :~
46
45 u
. 3'2 34 3,6 3'8

. 1
2

J
6

'1 j 1 ,~-1 ___ .

8 10 12

l'
)-
I 1.-- -1

. I
I· I .•
.' I
I
f i'
( I ;
~.
'~_ 2:: -I'
2nd -
46
45

:; 111
41 pre-term ,
~/ I ;--<~1 ::r:l> --.
11 j / .
, ; . -......1"
< . "
40
39 II 1~ . ·
'/l /!.y m '

" f/1)1-~' I
I
99 .• ' b
84
82
38 I / / t IYI/ I II, "I _1~1~1--' 9;1h
:: t~ /, 1 /{1 ;1/ :i ::1-:1 ~;:~ ~ ;~:~ ::
I I I !I : I .-1---1
80

:: IVt)jfl ./r~,~ rl~: ' ' Iii J J.3 1:1 ~ -i:~~f __ ~~;:
~ L1- -~ - ___~-
32
31
30
I' "

I'
t
j/! / I :,
/1 /'
. /V'1/ ' I

'I

I t;1.<
, , '

;.-- ~ ""......,
1/1; t.-V: .: : ; . _> >-_:'--/--1-= - .-
1 t /-:--1-::'//1: ,'''',::_ --,=1--'-' -
. !.-' --I -J -] _' J --
. ' ---- . --1- --1- 0.4," 68
66
64
29
28
27
11 1 / I /t/
.(1
1/ Cf/'
f/. «( ~
I

j y l,.
-'

­.
'
62
60
58
26 ' I, / / I , / f ()--,
5lt 1· / / -. 1/ '/ (" --
5~;S1'
):" ./ /~
25 1 14
24 I ;,--- , , I
-.--l 9 9.6 1b. 13
I
l4~ ' /1/
I
'1/ y
..(
j , 98 tb
12
~ 1;-( Slit·

:~ I
11
7Slh

10 50 t b - 10

~5th
9 9
9th
-~-

81' .- .
2 nd
8

71 -' O.4tb

_i ...-'
61 pre-term 6

5 5

3~ 32 3 4 36 38 40 42~4 46 + 8 50 52
8 9

28
D a Le Ag t.: ~ tcasllrem("n1 ~ ;J. m t·
3 *
:
2
:
~eeks /? months :
, 1

-'-~8·~'
~, I O 1~~8 ! :
.-
11 1'4 20 1 2 2'4 :
476 I Appendix B: Tables of standard measurements

Table 6 Body weight and height, boys, 7-5 years. (Courtesy of the Child Growth Foundation. Printed by
Harlow Printing Ltd, Maxwell St, South Shields, Nf334PUj
I
,
1 2 2'/, 3 3'/, 4 4'/, 5

130

1-5yrs years
--l- 125

996th
120

98th ­

115
gist 115

75tb
11 0

50th - 110

105
25tb 105

Sib

100
2.d -- 100
­
95

-- a.4th
95

90

85

/'

80
---~
I
- 9 9,6th 26

./
75
.. /
25

70
./
'"
T '" 24

98th

65
I
L /~
./
23

60 I 22

~--
I "'" /"
9 l st
r

1... _ _ _._---+-_ _ _ _ _ _ _
--t 21

20 20

75t h

19 19

18 ~ -­ .-+ ---- 18

5 0th

17 25 th
17

16 16

9t h

15 15

2nd

14 , 14

OAth

13

12

11 11

10 10

-~---+- 7

6 f---.­
years

-- -+
-......J
1 1'/, 2 2'/, 3
3'/, 4 4 ' /, 5

 Appendix B: Tables of standard measurements I 477

Table 7 Body weight ond height, boys, 5-78 yeors. (Courtesy of the Child Growth Foundation. Printed by Harlow
Printing Ltd, Maxwell St, South Shields, NE33 4PU.;

20 5 2 05
5 6 7 8 9 10 11 12 13 14 15 16 17
200 - - - + - - ­ 200

195 ' ­ 5-18yrs

With provision for school /""-1l--
..- _ - 99.6t h - :

1
195

190 ­ reception class

./ I --- - 98th - 19 0
/ /'
185 NAME .... .... . ...... ..... ... ........ .... .

~~./ 91&t - 185

D.O.B. ... ... 1 .. . . .1 .. .. ..
/ - / . 75th '
180 r--t>' .' .... 180

1 75
h / /' . .. 5 0th 17 5

/- : 1/ / 25 . h
170 77 170

16 5 4/ . ~-
9th

2n d
165
/ /
160
/
+7
v'
....--::i - -
./ r-. __ 0.4'h --'
160
/
155 .L-/ / /-I~ 155
/ /' / ;- /,V
150 L--/' /- 7'
,( / // / - y
/ /
145 /~ .4- /' 105
/ .-/, / 1-/
.L __/ --'
140 ~ .., 99.6' h - 100
/' /
/.-
/'
/
/ .-.­
135 /t-/
././~ .
/ 4 _.- - + - - 95
/ ,/
,Lo/ ./ ./ /'
130 r- ~ 17' / 98th -- 90
/ / ./ ./
L/ /
125 85
/ /
120
L. /
/ '7
115
/ /
/ 75
75th
110 . 70

105 50t h
--- 65

100 25th 60

95 9t h ---- 55

90 2nd
50
O.4th
45

40 40

35 35

30 30

25 ---
.....--. 25

20 20

15 15
years
10 10
5 6 7 8 9 10 11 12 13 14 15 16 17
4 78 I Appendix B: Tables of standard measurements

Table 8 Body weight, height and head circumference, girls, birth to 7 year. (Courtesy of the Child
Growth Foundation. Printed by Harlow Printing Ltd, Maxwell St, South Shields, NE334PU.j

~8 h 3~
O-lyr 14 16
Q)
20
0
241' 28 3b 32
C!; 0 ®
36 38140 42 h4 46 ,f8 50 52

@ @ 6

50

NAME weeks /6 :months 9 9.Sl h

49

O.O.B. . .. f .... .I ... . 9 8th

91sl 48

Correct by .. . .. .. .. for prematurity

7 3th
32 3436 38
2

3
50th

25th
- 47
46

45 · 45

9tl1
44
20d 44

43 , O.-1lh

43

42
42

41

40

39

pre-t e rm
.// 84

82

38
/ .'. 99 .6 1h 80

37
/ BSIII
78

36

/ 9Ist
75th
'76
35
74

50th
34 /
2Gth 72

,
33 /
1 Olll
70

2 tl d
32 / O.4t h
68

31 f 6

30 ,
29 /
f
I·
64

62

28 / ; 60

27
58

26

25

24 52

/ 56

54

52

50

48
13

46/

9fU;:h
.;.4

12

42
9~l h

4()
11

!J ls.t

10
75a; 10

50tb
9
9

25 th

~){ h
8
8

2nl!

7
O.·l lh
7

6 pre-term 6

4
/
/ 85052

,/

3
,. / 3

r
,
;'
r
2

/ 2

t­
i
weeks / j> m onths

~ 8~ 12~14 ~8
1

32 34 36 38 EO D 2 6
10 16 20 l2 2 4 t-,- ---+---I--+-----+­
o ___ T o o' ____ __

Appendix B: Tables of s ta ndard measurements I 4 79

Tabl e 9 Body weight and height, girls 7-5 years. (Courtesy of the Child Growth Foundation. Printed by
Harlo w Printing Ltd, Maxwell St, South Shields, NE33 4PU.j
, ,
1 2 2 '/, 3 3'/, 4 4 '/, 5
1-­

1-5yrs years
99.6 tb 120
25

NAME ............... .. ......... . .

98 th
D.O.B. .. .... / .. .... / .... .. 91 . t 115
75 th
110 110
5 Qth

105 28 th
105
9t h

2nd 100
OAth
95

90

27
9 O.6t h

26

25

98tb 24
60 23
55 22
9Is t

21
20 20
75t h

19 19
18 50t h 18

17 .-'/
17
16 ----- ..... --- 2:,l h

~
16
9 th
15 15
14 2nd
14
13 OA t h
13
12 12
11 11

10 10

,.. ,..
9 9
8
-- ;;-­
8

7 7
6 6
years *Measurement: H i e =H c~ d Circtlmfcrt'llc('. L = LCIl!!t1 l. W = W cl~ hl. H = liCl,~il!

1 1 '/,
2 2 '/, 3 3 '/, 4 4 '/, 5
 480 I Appendix B: Tables of standard measurements

Tabl e 10 Body weight ond height, girls, 5-78 yeors. (Courtesy of the Child Growth Foundotion. Printed by Harlow
Printing Ltd, MaxwellSt, South Shields, NE334PU.j

195 ---, 195

5
6 7 8 9 10 11 12 13 14 15 16 17

190
. 190

I
,
185
5-1Syrs
With provision for school
-; 185

180
f reception class 99 .6th .=, 180

175
NAME .. ...... ... ... ... ...... ... ... .. ... .. . _ --' 98th
175

O.O.B . ...... 1...... 1 ..... .

170

165

--­ __ ~ g lst _

76th
170

165

_ - - - -- 50th _

160
25th 160

9t h
155
155

/ /
2nd
150
150

/ /
- O.4th
145
145

/'
/'
140
/ .7
/ /
135

130
95

125
L 90

/'
_99.6 t l>
120
. 85

115
80

98th
110
75

105
..-:._ 91st - 70
/'

100
,; 65

75 th

95
/ 60

50t h ­
90
55

/ 50

9th
,;
2nd _ 45

"
40
_ 0 .4th
40

./

35
35

/' Y
/' /' ,,/,,/
30
30

.-----.-
/' /' /"
/' //' /

~ -- ------...- --
25
/ // ,.'
25

~
~ ~ .' ~-­
• / __ - - - - - - -- _lo- -­

20 - 20

. ..' . .....­
~ -~--::-­
15 :.:.:--:--:::- -::: .5. - . . 15

-~-­
,...
years
10
10

5 6 7 8 9 10 11 12 13 14 15 16 17

~­ ---
 Appendix B: Tables of standard measurements I 481

Tabl e 11 Organ weights in infants, newborn to 72 months 4

Organ No. in group Mean weight (g)

Newborn
Brain 13 325
Heart 16 18
Right lung 12 29
Left lung 12 24
Liver 16 109
Spleen 16 8
Right kidney 14 11
Left kidney 14 11
Thymus 16 11

1 month
Brain 69 489
Heart 65 27
Right lung 63 55
Left lung 63 46
Liver 74 176
Spleen 74 16
Ri ght kidney 74 17
Left kidney 74 17
Thymus 68 21

2 months
Brain 102 569
Heart 100 31
Right lung 94 61
Left lung 94 52
Liver 102 193
Spleen 103 17
Right kidney 104 19
Left kidney 104 19
Thymus 90 28

3 months
Brain 81 651
Heart 80 33
Right lung 83 67
Left lu ng 83 57
Liver 85 223
Spleen 85 19
Right kidney 84 21
Left kid ney 83 22
Thym us 76 31

4 months
Brain 44 691
Heart 40 33
Right lung 37 73
Left lung 37 60
Li ver 44 232
Spleen 44 22
Right kidney 43 21
Left kid ney 43 21
Thym us 38 29
(Continued)
482 I Appendix B: Tables of standard measurements

Tabl e 11 (Continued)

Organ No. in group Mean weight (g)

5 months
Bra in 26 682
Heart 24 34
Right lung 23 69
Left lung 23 56
Liver 25 246
Spleen 26 22
Right kidney 26 23
Left kidney 26 26
Thymus 22 25

6 months
Bra in 17 766
Heart 15 37
Right lung 16 72
Left lung 16 61
Liver 15 280
Spleen 17 24
Right kidney 18 22
Left kidney 18 22
Thymus 16 32
7 months
Brain 12 763
Heart 11 37
Right lung 10 70
Left lung 10 57
Liver 12 285
Spleen 12 26
Rig ht kidn ey 12 24
Left kidney 12 27
Thymus 9 32
8 months
Brain 8 852
Heart 8 45
Rig ht lung 7 74
Left lung 7 63
Liver 8 340
Spleen 8 31
Right kidney 8 30
Left kidney 8 29
Thymus 6 32
9 month s
Bra in 12 925
Heart 12 45
Right lung 11 87
Left lung 11 79
Liver 12 342
spleen 12 45
Right kidney 12 28
Left kidney 12 28
Thymus 10 26

(Continued)
 Appendix B: Tables of standard measurements I 483

Table 11 (Continued)

Organ No. in group Mean weight (g)

10 months
Brain 10 986
Heart 10 47
Right lung 9 89
Left lung 9 76
Liver 10 369
Spleen 9 36
Right kidney 10 30
Left kidney 10 35
Thymus 8 28

11 months
Brain 8 935
Heart 9 52
Right lung 8 104
Left lung 8 104
Liver 9 388
Spleen 9 39
Right kidney 9 35
Left kidney 9 35
Th ymus 9 33

12 months
Brain 3 920
Heart 3 51
Right lung 3 95
Left lung 3 86
Liver 3 405
Spleen 3 42
Right kidney 2 30
Left kidney 2 33
Thymus 2 16
Oa ta from ref. 4.

Sudden infant death syndrome (SIOS) and non-SIOS aggregated; no significant difference found between

groups in original analysis.

484 . Appendix B: Tables of standard measurements

Table 12 Major organ weights (g) by age (7 -7 8 years) and sex:5

Age (years) Brain Lungs Heart Liver Spleen Kidneys

(right + left) (right + left)

M F M F M F M F M F M F

1 971 894 107.3 175.3 54 48 400 390 35 34 72 65

2 1076 1012 245.9 244.3 63 62 460 450 42 41 85 75
3 1179 1076 304.7 265.5 73 71 510 500 48 47 93 84
4 1290 1156 314.2 311.7 83 80 555 550 53 52 100 93
5 1275 1206 260.6 319.9 95 90 595 590 58 57 106 102
6 1313 1225 399.5 357.5 103 100 630 635 62 62 112 112
7 1338 1265 365.4 404.4 110 113 665 685 64 67 120 123
8 1294 1208 405.0 382.1 122 126 715 745 68 71 128 135
9 1360 1226 376.4 358.4 132 140 770 810 73 77 138 148
10 1378 1247 474.5 571.2 144 154 850 880 82 85 150 163
11 1348 1259 465.6 535.0 157 168 950 960 91 93 164 180
12 1383 1256 458.8 681.7 180 188 1050 1080 101 103 178 195
13 1382 1243 504.5 602.3 202 207 1150 1180 111 112 196 210
14 1356 1318 692.8 517.0 238 226 1240 1270 127 120 212 222
15 1407 1271 691.7 708.3 258 238 1315 1330 135 127 229 230
16 1419 1300 747.3 626.5 282 243 1380 1360 145 134 244 236
17 1409 1254 776.9 694.5 300 247 1450 1380 152 140 260 240
18 1426 1312 874.7 654.9 310 250 1510 1395 157 146 270 244
Data from ref. 5.

F, female; M, male.

References 3 Naeye RD. Umbilical cord length: clinical significance.

Maroun LL, Graem N. Autopsy standards of body parameters
and organ weights in non-macerated and macerated human
fetuses. Pediatr Del! Pathol 2005; 8:204-17.
2 Pinar H, Sung CJ, Oyer CE, Singer DB. Reference values for
singleton and twin placental weights. Pediatr Pathol Lab Med
1996; 16:901-7.
J Pediatr 1985; 107: 278-81.
4 Thompson WS, Cohle SO. Fifteen-year retrospective study of
infant organ weights and revision of standard weight tables.
J Forel1sic Sci 2004; 49: 1- 11.
5 Altman PL, Dittmer OS. Growth: Including Reproduction and
Morphological De1!elopment. Washington DC: Federation of
Am erica n Societies for Experimental Biology, 1962.

--
 I INDEX I

Note: For purposes of conciseness, the terms 'abuse' and 'non-accidental' imply suspected and/or actual abuse. Figures and tables
are comprehensively referred to from th e text. Therefore, significant material in figures and tables have only been given a page
reference in the absence of their concomitant mention in the text referring to that figure. Abbreviations: SIDS, sudden infant
death syndrome; SUD], sudden unexpected death in infancy.

abando ned body (baby or fetus) 140-1

prolonged , isolated 86
abnormalities 240

identifying 413
ac ute metabolic autopsy 119-20
infections 240, 241

abdome n (and contents)

acute phase response, ante-mortem
obstruction 207

maternal trauma as cause of fetal

109
post-mortem examination 155

death
acyl-CoA deficiency )16
post-mortem exa mi na tio n, live

blunt 190-2, 192-3

long cha in 244
birth determination 185

penetrating 192
medium chain see medium chain
see also inhalation; respirato ry tract

neonatal post-mortem examination

acyl-CoA deficiency albumin, post-mortem evaluation 108

160
short chain 244
alcohol (ethanol)

abortion
Addison's disease 113, 128
fetal exposure to 267,274

definition 181
adnexa, ocular, non-accidental injury
fire deaths relating to 318, 321

induced/deliberate 181
125-8
immersion victims, testing 347

offence of procurement 181

adolescen ts/teenagers/ older chi ldren
post-mortem diffus ion 270

abortuses 141
drug ab use 256-7 , 341-2
road traffic accident

abscess, retro ph ary ngeal 240- 1

anabolic steroids 258
testing child 389

abuse
testing 272 -3
testing driver 388

child see child ab use

volatile substances see volatile
alleles 39 7-8

drug see drugs, recreational

substan ces
allergy see an aphylaxis ;

acdde ntal injury 20, 70, 336-44

epilepsy and bath-tub drowning
hypersensitivity

bone, vs non-accidental 20, 70, 71

355
Allitt, Beverley Gait 27 4, 374

causes and mechanisms 338

see also school-age children
alloimmune thrombocytopenia,

consequen ces 338

adrenal glands
neonatal 98

death due to
fetal weight 451
Alport's syndrome 95

bath tub drowning 3 53-4

injury 68
alternating current injury 32 7

epid emiology 336, 337

insufficiency/hypoplasia 113, 238
alveoli

epidemiology 336-7
lipid dep letion and SIDS 212
in immersion 350

falls see falls

adulteration, biological specim en SUD! 210,215,216

fundus ha emorrhage due to 131

269-70
amel ogenin 400-1

head see head

adverse drug reactions 367
amino acid disorders 115

in hosp ital 372-3

age
sudden death 117

prevention 341
dental estim ation 441-3
am niotic fluid toxicology 267

asphyxial deaths 334

gestational see gestational age
amphetamine tests, false-positives 268

role of carers 341

agricultural injuries 341,393-4
amyloid -beta precursor protein see

proneness 33 7-8
air bags 386
beta-amyloid precursor protein

road traffic-related see road traffic

air em bolism 364
ana bolic steroids 258, 271-2

accidents airways
anaemia, aplastic 99

use of term 'accident' 33 7

lower, narrowing in asthma 243
anaesthesia 370-1

acetaminophen poisoning 149

obstruction
dental 371

acho ndroplasia 236

foreign body 330-]
manslaughter 367

acromion fracture 59
imposed 332-3
anaphylaxis 109-10, 239-40

activated partial thromboplastin time

upper 207
drug-induced 367-8

(aPTI/PTIK) 80,81
peripheral, resistance, increase with
ana to mical variants (normal)

with other haemato logica l

immersion 350
genitalia 29-36

abnormalities 89
upper 240-1
skeletal 70

486 I Index

ane urysms, intracranial 236

aspirin 99
factors affectin g 107

an hidrotic ectodermal dysplasia,

assessme nt, clini ca l and medical, in
ge nera l eva lu ation 107-8

X-linked 235
abuse see child abuse
immersion victims 347

animal bites 42 7- 8
asthma 342
infants/young children 149-50

anterior segment (eye) injury 126

acute 243
technical considerations 117- 20

anthro pometry see body

'at-risk' register, death of child on
birth

anticoagulants 98
146
concealment of delivery or 181, 182,

anticonvu lsant (antiepileptic) dru gs

atherosclerosis, premature coronary
352

243
249
live, determination 183-7

antidi uretic hormone, syndrome of

atrioventricular block 232
see also de livery; labour

inappropriate sec retion 366

atrioventricul ar node, endod erma l
birth trauma 173-6

a ntiepiieptic drugs 243

heterotopia 233
neurological 173-6, 289, 300-2

antiplasmin deficiency 93
attention deficit hyperactivity disorder
mechanisms of brain injury 304

a ntiplatel et drugs 99
(ADHD), methylphenidate in
sp inal cord 176,302,312

a nus and perianal area, examination

257- 8, 272-3
post-moltem examination 161- 2,

11-12,36-7,39-40
autopsy see post-mortem examinatio n 188-90

acute/healing/healed trauma 37- 8

axon al damage/injury 168-9
retinal haemorrhages 133

boys 27, 467

diffuse, road traffic victims 392
skeleta l see skeleton

girls 26
nerve root 176
SUD Tand 218

sig ns of abuse 39-40

bite marks 425-32

in sum 2 17
babies see infan ts and babies
detail of 429-30

aortic stenosis, congenital 227

babygram (whole-body radiograph) 48
impressions with see impressions

aplastic anaemia 99
sudden death (infancy/early
pattern/distribution 18, 425-6

apolipoprotein A genotype and

childhood) 146
recording 430-1

traumatic brain injury 312

bacteria l in fect ion
self-inflicted 19, 427

apoptosis, brain cells (in early life)

ante-mortem 110
suspect 431-2

167
bites 428
comparison evidence from bite and

apparent life- threatening events

trachea 24 1
suspect 432

(ALTE; near miss events)

bacteriological samples, sudden death types 426-7

haemosid erin as marker 216

148-9
bladder, STDS 210

respiratory cha in complex deficiency

'b angungu t' 249
Bla nd-White-Garl and syndrome

246
basal gangl ia, hypoxic-ischaemic
231

SUDI/STDS preceded by 202, 202 - 3

damage 172
bleedi ng disorders (inc!. diatheses) 80,

arousal shakin g, retin al haemorrhage

bath sa lt flu orescein 353
82-101

132
bath tub and spa pool drownings
bruising 76-7

arrhythmias, sudden death associated

353 -6
drug-induced 98 - 9

wi th 232, 365
accidental 353-4
evaluation/m easurements 82-94

epilepsy and 243

non -acc idental 354- 6
of primary haemostasis 82

in hospital 365
battered child syndrome see physical fundus haemorrhages 132

due to adve rse dru g reactions 367

ab use
bleeding patient, evaluation 82-6

immersion 350
Bayley's Motor Scale 17
see also blood, loss; haemo rrh age

infants (SUOT) 207, 232

bean - bags, polystyrene-filled 204
bleeding time 82

arrhythmoge nic right ventri cu lar bedding and SUDl/STD S 140, 203
blood

cardiomyopathy 229- 30
bed-sharing see co-sleeping
clotting see coagulation

arsenic exposure 233

behaviour in abuse
flow, cerebral (CBF), mon itorin g 291,

arte ritis, coronary 231

assessing 460
292, 292-3

in stantaneous infant death 205-6

negative 14-15
loss, estimation in sudden death 152

ascorbic acid (vita min C) deficiency

Bernard-Soulier syndrome 78, 81, 82,
sa mples

and scurvy 71,94

94
acute metabolic autopsy ll9

asp artic ac id, dentine 443

0- ox id ation see fat ty acid 0- oxidation
collection 11 8

asphyxial deaths 329-35

defects
to xicology 263

immersion in non-water media 349

beta -amyloid precursor protein 169
tests see haematological tests
preve ntion 333 -4
tra umatic brain inju ry 3 12
blood vessel see vasculature
sleeping infants see sleeping
bicycles see cyclists
blunt abdomin al t rauma, matern al,
enviro nment biochem ical tests on post-mortem
causing fetal death 190-2,

aspiratio n see inhalation specimens 106-26, 149-50

192-3

~ ~r:.
- --
- - -
 Index I 487

body brain injury, trau matic 63-7,

capillalY electrophoresis and capillalY
in fet al /perin ata l death,
282-3 17
zone electrophoresis 261

measurements (inc!.
genetic factors influencing
capilialY reactions, bra in 168

anthropometlY) 187
recovery 312
caput succedan eum 302

in fire fatalities 319-20

non-accidenta l head injury and
car(s) an d other vehicles

iden tifying 320, 412

63-7
accidents and collisions see road
hei ght and weight see heigh t; weigh t
neu rological presentations 295-6
traffic accidents

identifying see ident ification

ocular injury associated with
entrapment in 330

in sudden unexpected death 139

127-8 , 130-1
inspection following collision

cavities s ee ab domen; chest

primary mechan isms 302-7
388

clothing see clothing

secondary mechanis ms 307-11
car seats, child 386-7

family wish to view 142, 389

see also head injury
carbohydrate meta bolism, sudden

measureme nts (inc!. anthropometry)

brainstem
death 117

150
damage 172
carbon monoxide

road traffi c accid en t 389-90

bi rth -rel ated 302
haemoglobin and 320, 322 -3

to tal surface area, burns and 323

SIDS, histology 211
poisoning 139

bone breastfeeding
carboxyhaemoglobin 320, 322-3

brittle see brittle bone disease asphyxial death 204. 372

cardiomegaly. myocardial infarc tio n

facial, fetal hei ght estimation from infant poiso nin g 257
w ith 233

444
breech presentation 188
cardiomyop athy, su dden death 155,

injury
brittle bone disease
156, 228, 228-9, 229 -30

accidental vs non-accidental 20,

inherited see osteogenesis
infant 206-7

70,71
imperfecta
cardiopulmonary resuscitation see

fractures see fractures

temporary 73
resuscitation

post- moliem ex amination 163

bronchiolitis, acute 241
cardiovascu lar disease/disorders see

metabolic disease 7l, 94

bronchopulmonary dysplasia 210, 248
heart and entries above

bone marrow failure sy ndromes

hospital de ath 372
carers and hea lth professionals

99-100
bruises
perpetrating abuse or death 274,

bone scan, ab use 51

diseases causing 76-7
373-5

bowel see intestine

in fetal/perinatal death 189
preventing accidents 337-8

b.ox jellyfish 358

non-accidental 294
carnitine deficiency 244

boys
assessment for 17 - 18, 19
carnitine palmitoyltransferase type 11

growth charts 475-7

importance of sha pe 4
deficiency 244

sex ual abuse, examination for 467

in sudden death
catheter-related deaths 368

technique 27
infants 214-15
cell death, brain (in early life) 167

brain
observed at scene 140
central venous catheter-related deaths

damage see brain injury;

bucket, drowning in 353
368

neurological lesions
bucket-handle fracture 53 , 54, 57
cephalic presentation 188

herniations 311
bullous impetigo 20
cepha1(o)haematoma 218, 301

hypoxia see hypox ia

burns 5, 323-5
cerebral artery

imaging see neuroim aging

eyelid 125
infarction 176-7

infarction 310-11
pathological changes 324-5
middle (MCA), in head injury

in early life 176-7

patterns 20
flo w velocity assess me nt 293

malformations 169-71,237
skin conditions simulating 20
occlusion 290

oedema see oedema

spasm 293

post-mortem exam ination and

C-reac tive protein, post-mortem 109
cerebral blood flow (CBF) monitorin g

findings 156-7
calcification, infantile coronary 232
29 1, 29 2, 29 2-3

neonatal 160
callosal agenesis 170
cerebral function an alys ing monitor,

SlDS 211
callus formation 52
head injury 292

weight see subheading be/ow

calorie intake, post-mortem evaluation
cerebral hae morrhage

temperat ure, head injury 292

108
birth-related 161, 190

weight 484
Camino fibreoptic pressure monitoring
sudden death 236-7

fetal 471
291
cerebral oedema see oedema

neonatal and infa nt 481-3

CAMPI 440
cerebral palsy 167

see also intracranial path ology and cancer/malignancy, bone marrow

cerebral perfusion, infarction due to

entries Linder cerebral infiltration 100

impairment of 310-11

488 I Index

cerebral perfusion pressure (CPP) in

child-killing immersion sy ndromes complexes 79

neurotrauma 309
351,352- 3 deficien cies 80, 83, 84,90, 92 - 3,

monitoring 291
see aIso filicid e; homicide; 96-7

cerebral sinovenous thrombosis 177

infanticide; neonaticide
tests revealing 86, 87, 88, 89

cerebrospina l fluid , post-mortem 120

chi ld protection forms II , 17, 458-69
inhibitors, tests show ing 86, 87 , 88,

cerebrovascular autoregulatory
Child Protection Orders 7
89

mechanisms 291
Child Protection Services 2
coca ine, environmental exposure 268,

certification of death, SillS/SUm 218-19

Children Acts (inc!. Scotland) 1, 7
272

cervica l spine and spina l cord trauma

sig nificant harm in 4
prenatal 274

311-12
chime rism 410
codeine, breastfeeding mother 257

perinatal 193, 311

chinon 161
coll oid cyst, 3rd ventricle 237

road traffic victims (=whiplash

chlamydial infection 42
colposcopy 11-12,27,39

injury) 393
chloride leve ls, post-mortem 110, III
video 7, 8, 11 , 27

cervicomedullary syndrome (whiplash

chromatography
combined paternity index 409

shaking injury) 283, 306, 312

gas see gas chromatography
compression head injury 302

hyperacute 295
liquid see liquid chromato graphy
computer-assis ted dental identification

Chediak-Higashi syndrome 81, 92 , 95

micellar electrokinetic capi llary
systems 440

chemical asp hyxi a 333

(MECC) 261
computer tomography (CT) of head 12,

chemical burns 324

chromosom es 393-4
50, 61-2

non-accidental 326
see also X chromosome;
accidental injury 293

chemical ionization 262

Y chromoso me
mild 289-90

chest (thorax) viscera

chro nic disease, sudden death 226
non-accidenta.l injury 12, 50, 61-2,

post-mortem examination and

cigarettes
299-300

dissection 154
burns 326
conception, prod ucts of 141

neonatal 160
smoking see smoking
confi dentiali ty 8

traum a in abus e 68
Clauss fibrinogen assay 81, 82
confl ict of in terest, expert witnesses

child abuse 1-105, 125-36,457-69

clavicular fracture 59
454

alerting signs 2- 3
at birth 71 , 161,218 congenital infections

clinical and medical assessment

clinical assessme nt, abuse see child brain 177

J-23
abuse syphilis 71

~ow chart 5-6

clinical presentation see presenting conge nital malformations/anomalies/
forms 11 , 17,457-69
history abnormal ities

formu lating an opinion 2 1

clothing of decedent brain 169- 71 , 237

interpreting findings 17- 20

bagging for police in road traffic genitalia 41

need for comprehensiveness 4

accident 389
heart 226-7

rol e 7
in hospital , road traffic accident 388
cyanotic 83 - 4, 101

definition 2
at scene of sudden death
drowning and 3 56

dentist and see dentist

remo va l from body ]]9 instantaneous infant death due to

documentation see docum entation

road traffi c accide nt 388
205

dru g-facilita ted 273

clotting see coag ula tion
perinatal death due to 193

expert witnesses see expert

clozapine 272
rapid infant death due to 205

witnesses
coagulation (clotting) 79-80
in-hospital neonatal death 372

investigations 12
activation 100
upper airway 193, 240

legal issues 1-2, 4, 7 -8, 22

classical path way 80
congenital platelet disorders 94-5

medical conditions simulating see

defects 96 -7
con ing 310, 369

medical conditions
pe tech ial haemo rrhages and 329
conj unctival petechiae, SUD! 213 -14

ongoing health care 21

diss eminated intravascular (DIC) 90,
consciousness, assessing level of 282

prevalence 2
101
conse nt (inci. informed .consent) in

types 3
inhibitors, tests showing 86, 87, 88,
abuse to examination 8, 457, 459

neglect see neglect

89
sexual abuse 27

physical see physical abuse

initia tion and amplification 79-80
contaminatio n of sampl es,

sex ual see sex ua l abuse

normal, and normal platelet count,
environmental

Child Assessment Orders 7

bleeding with 90-4
DNA samples 403

ch ild destru ctio n

platelet enhancement of 79
hair samples 268, 268 -9

definition 181
tests 77
contusions

law 181
coagulation fa ctors 79-80
brain 294

 Index I 489

birth-related 302
cyanide (fires) 320, 322, 323
degradation, drug, post-mortem 270-1

imaging 63
cyanotic congenital heart disease
dehydration, ante-mortem 110-11

spinal cord 311

83-4, 101
deliberate acts see non-accidental acts

copper deficiency 71
cyclists/bicyclists 388, 392
delivery

copy number
helmets 338-9, 392-3
concea lment of birth or 18], 182,

low see low co py number ana lysis cyst

352

variation (CNV) 398

brain 168, 171-2, 237
instrumental see instrumental

core temperature, head injury 292

laryngeal 207
delivery

co rneal damage 126

cystathione B syn thase deficiency, intracrani al haemorrhage and mode

coronary artery pa thology in SUDI

thromboemb olism 235
of 174,17 5

227,230-1,232-3,364
cystic fibrosis 397-8
unattended 182-3

anomalies 227, 230-1

cystic medial necrosis 235
dental identification 435-46

embolism 364
cytochrome P450 (CYP) enzymes in
comparison technique 435, 435-49

instantaneous infant death 205-6

toxicology 258, 263
profiling technique 435,440-1

occlusive disease 232-3

cytokines and burns 325
dentine, aspartic acid 443

petechial haemorrhages 210

cytomegalovi rus
dentist and child abuse 420-34

premature atherosclerosis 249

parotitis 250
accused of abuse 425

coroner 143, 190, 225, 345

SIDS/SUDI 211
examini ng for/reporting of abuse

hospital death 375, 377

420-2

corpus callosum, agenesis 170

Dandy-Walker 237
dentistry, deaths in 371

cortex, cerebral , hypox ic-isch ae mic

dating of fractures , radiological 59-60
Denver Developmental Screening Test

damage 172
skull 60-1
14

co-sleeping and bed-sharing 203,

DAVID 440
depressed skull fractures 294

204-5
'dead in bed' syndrome and insulin­
depression, post-natal 352

overlaying risk 203, 215, 331

dependent diabetes mellitus 245
dermati tis, napkin 217

wedging risk 204, 331

death(s)
developmental assessment (in abuse)

see also 'dead in bed ' syndrome

accidental see accidental injury
12-15

cot 203
adequa te cause of, in SU D! 200-1
diabetes mellitus 112-13,243-4

baby found dead in (and 'cot death')

asphyxial see asphyxial deaths
sudden death 243 -4

140, 198
certifi cation in SIDS/SUDI 218-19
diaphragmatic hernia, co ngenital

abandoning of term 'cot death' 199

dyadic see dyadic deaths
239

pathology 206, 208

feta I see fetus
diaphyseal fractures 52, 53

bedside 204
head injury 283, 284-8
at birth 71

characteristics 202
heat-induced see fire fatalities;
dia to ms, immersion victims 347

see also sudden infant death

thermal injury or death
diazepam and breastfeeding 257

syndrome
immersion see drowning
diffuse axonal injury, road tra ffi c

cotinine, neona ta l hair testing 273-4

infant see infant
victims 392

court, expert witness see expert

mUltiple 141- 2
diffusion-weighted MRI of head 50,

wi tness perinatal see perinatal death

299-300

Cozart® drug detection system (DDS) scene of see scene

digital dental radiographic

260
time of see time
identification 440

crack (cocaine), hair testing 272

traumatic, relative ranking order of
digi tal injuries 59

cranium see skull

346
dilated cardiomyopathy 229

creatinine 114
see also cell death; ch ild-killing; dilute Russe ll's viper venom time 87

creeks, drowning 357-8

eutha nasia; filicide; homicide; direct current injury 327

crime scene manager 138-9

infanticide; manslaughter; disaster and victim identification 440

Criminal Hearing System (Scotland) 7

neonaticide; serial killing; suicide diseases see medical conditi<;>ns

Criminal Procedure (Scotland) Act

decedent dissection, infants/older children

(1995) 7
body see body 152-6

crimping bruising 19
clothing see clothing disseminated intravascu lar coagulation

Crohn's disease, anogenital 41

deciduous teeth, age estim ation 442,
(DIC) 90, 101

crown-heel length (CHL), fetal autopsy

443
distribution, drug, post-mortem 270

471-2
decomposition, electrolyte
diving reflex 349-50

crown-rump length (CRL), fetal

abnormaliti es 111
bath epilepsy and 355

autopsy 471-2
decongestants, nasal, affecting
DNA 396,397

Cushing phenomenon 291

amphetamine tests 268
databases 414

490 I Index

DNA (continued) drugs emotional trauma see psychological

mitochondrial see mitochondrial recreational (illicit drugs)
stress

DNA accidental ingestion 342

encephalitis 238

DNA tests 395-419

coercive administration 272
encephalopathy

dental identification and 437

environmental exposure to 268,
cardiac arrest 172

dental pulp, sex determination 444

272
haemorrhagic shock (syndrome of)

fatty acid oxidation defects 117,

fetal exposure to 257, 267, 273-4
238

246 fire fatalities relating to 318,321

pancreatitis (acute)-related 248-9

history and techniques 398-402

older children see adolescents
of Reye's syndrome 238-9

identifying body (inci. missing

road traffic accident 388-9, 389
traumatic/non-accidental 283,

persons) 411-13
therapeu tic use
295-6

immersion victims 347

death due to errors 266-7
endocardial fibroelastosis (EFE) 229

samples for 119, 402-3

paediatric pharmacology 258-9
sub-EFE 207

mixed 406-7
see also toxic substances
endocrine disorders 112-14

problems and solutions 406-7

'dry' drowning 348
endodermal heterotopia, AV node 233

sexual abuse 403-7

dura
endotoxin, bacterial 110

twins and 396

brain
endotracheal tubes 368

see also molecular methods

haemorrhage see intradural
energy metabolism, disorders 114, 116

documentatio nl records
haemorrhage
England, road traffic casual ties and

in abuse 16-17
in sudden death 157
fatalities 390

babies/infants/young children
spinal cord, in sudden death 158
enterocolitis, necrotizing 212

145-6
dyadic deaths (homicide-suicide)
entrapment asphyxia 330

bite marks 430- J

141-2
envenoming and drowning 358

fundus haemorrhages 133-4

drowning of child 352,355 environment

scene of sudden death 138,

dysfibrinogenemia see fibrinogen drug exposure from 268-9, 272

145-6
dysrhythmias see arrhythmias sample contamination see

dental
contamination

post-mortem 438-40
ecchymoses 88, 89
sleeping

retrieval of patient records 437-8

echo viral infection 209
good 203

expert witness 452

ectodermal dysplasia, X-linked
suboptimal 203-4

declaration by expert in report

hypohidroticlanhid rotic 235
enzyme assays

452-3
eczema (dermatitis), napkin 217
in fatty acid oxidation defects, liver

road traffic collision 387

EEG, head injury 292
116-17

see also photographic evidence;

electrical inj ury/burns
fibroblast cultures see fibroblast

reporting; video recordings

child 326
cultures

dodecanoic acid 116

pregnant woman 192
enzyme immunoassays (ElAs) for

dog bites 427, 428

electroencephalography, head injury
drugs in oral fluids 259

doping agents 258,271-2

292
enzyme-linked immunosorbent assays

screening for 27 J-2

electrokinetic capillary methods
for drugs 259

Doppler, transcranial, head injury

269
enzyme-multiplied immunoassay

292-3
electrolytes, post-mortem 110-11
technique for drugs 259

dosage errors 366-7

immersion victims 347
epidural haematoma see extradural

drains, drownings 358

electron transport chain (oxidative
haematoma/haemorrhage

drowning and near-drowning

phosphorylation; respiratory
epiglottitis, acute 240

(immersion) 332, 345-61

chain) defects 116, 244, 246-7
epileptic seizures see seizures

aftermath 358-9
electrophoresis, ca pillary 1ca pillary
epiphyseal plate injury, radiology 56

causes 345-6
zone 261
epistaxis (nasal haemorrhage)

classification 352-9
electrophysiological monitoring, head
sum and history of 203

by site 352-9
injury 292
SUD! and presence of 214

unlawful child-killing 352-3

electroretinography 134
erythrocyte sedimentation rate,

epidemiology 337, 345

ELISA, drugs 259
post-mortem 109

immersion syndromes 346,

embolism
escharotomy 325

351-9
hospital fatalities 364
ethanol see alcohol

investigations 349
pulmonary 235, 364
ethnicity, dental remains 444

medium 346-9
embryo transfer mix-ups 410
euthanasia by bath tub drowning 354,

pathophysiology 349-50
emotional abuse, assessment 15-16
355-6

-
~-~
_.

~ -:, " .- I , - _.
evidence (forensic)
from bi te and suspect, co mparison of
432
expert witn ess see expert witness
scene of sud de n death 138
road traffic collision 387
sex ual abuse 27- 8
see also sa mpl es
Exclusion Orders 7
expert witnesses (a nd testimon y)
447-55
abuse
physical 45 1
sexual 43, 45 1
admiss ibili ty of evid ence 450 - 2
claim to expert ise 449
communica tion s from 452
pre-trial 453 - 4
conflict of interes t 454
mission statement of expert 448-9
opinion 450
rece nt devel op ments 4 54-5
rul e of evidence 454
in witness sta nd /box 4 53
yes or no answers? 450
external exa min ation
in fetal/perinata l death 189
non-acciden ta l head inj ury 294- 5
in sudden dea th 150 - 2
fi ndings of questionable
significance 21 7
neonatal 158- 9
at scene 140
SlDS vs no n- accide nta l injury
212-15
see also physical exam ina tion
extradural (epidural)
haema tom a/haemo rrhage
bra in 6 1
neonatal 160
neo natal bhih-related 176, 302
post-mortem exami na tion 156, 160
traumatic (non-b irth-rela ted) 294
sp ina l cord , birth-related 176
eye 125- 36
non- accidental inju ry 125- 36
scope 125-8
post- mo rtem examination 157-8
eye witness, roa d traffi c co llision
387
firearm deaths 337
eyelids, non-acciden ta l inj ury 125
fi reworks 32 4
fa bri cated/in d uced/sim ula ted ill ness
(FlI ; Munchausen's sy ndrome by
proxy) 3
assessment 16
by carers 274
in hospi ta l 373
face
feta l height estima tion from bones of
444
injury in abuse, dentists and 42 2-3
reco nstructio n 440
fa ctor V (a nd Va) 79
defi Ciency 89, 97
fac tor VI I (an d VIla) 79, 80
defi ciency 80, 89, 97
factor VIII (and VIlla) 80
deficiency see ha em ophilia, type A
inhibi to rs 88
factor IX (an d IXa) 79, 79-80, 80
defi ciency see haemophili a, type B
factor X (and Xa) 79, 80
detlci ency 89, 97
factor XI (and Xla) 80
deficiency (hae mophilia C) 80 , 84,
88, 97
factor XII 80
defici ency 80, 88, 92-3, 97
factor XIll deficiency 83, 84, 88, 90,
93
faeces (stool s), live birth determination
185
failure to th ri ve (n on-organi c),
assessment of 13-14
falls, accidenta l 33 8, 339-40
from height 338, 33 9-40
mortaliti es 33 7
short fall s 298, 338 , 339
false-negative drug screens 269
false-posi tive dru g screens 268-9
fa lx, neona ta l exa mination 160
fa mili al searchi ng 4 13-1 4
fa mily see pa rents/immediate family
fa mily history
ab used child 10
bl eeding child 84
dea th (perinata l/in fan t/young chil d)
146
fa rmin g injuri es 34 1, 393 -4
fasc iitis, necro tizing 248
fat embolism 364
see aIso lipid
fat her, alleged see paterni ty testing
fa tty acid f)-o xidatio n defec ts 11 4-17 ,
244-6
sudd en death 114-1 7, 244- 6
infa nt 114, 208
fatty acid ethyl esters 26 7
fatty change, liver, SID S 211-12
Fechtner's synd ro me 95
fe male children see girls
Index I 491
femoral fracture 49, 54, 55
SIDS 215
fertili ty treatment mix -ups 410
fetus (prenatal child)
abandoned remain s see abandoned
body
age see gestati ona l age
damage (= intrauterine damage)
169- 73
brain 300
roa d traffic accid ents 192-3, 193,
393
deat h (=intrauteri ne death) 180-97
defini tion 180
giving cause of 193 - 4
maternal death as ca use of 190-3
natural causes 19 3
post- mortem exam ination see
post- mortem exa mination
see also abortion ; abOIi uses
drug exposure 257,267,273 -4
viability assessmen t 187
see also entries under co nge nital
fever (py rex ia) in brain tra uma
292
flbreoptic pressure moni torin g of
intracra nial pressure 291
fibri noge n
abnormal levels (dysfibrinoge nemia)
81, 88, 89-90, 97
co ngenital causes 97
measurements 81-2
fi brob last cultu res (for enzyme
analysis) 119
fa tty acid oxidation defects 117,
208
fibroelastosis, endocardial see
endocardial fi broelastosis
fi bro ma , cardiac 23 4
fibromus cul ar dysplasia 233
fibular fra ctures 57
filicide 353
in hospital 373- 5
fi re fata lities (house fire deaths)
31 8- 27
dental id entifica tion 439
epidemiology 3 18 - 19
pathologiSt's role 3 19-27
scene 141 , 3 19- 20
fi rst-degree burns 32 3
fixation, organs 163
FLAIR MRI, head 64
fl as h injuri es 326
flotation test 185
492 I Index

fluid administration
types 128-9
glutaric aciduria 70

burns 325
see also retinal haemorrhages
glycoproteins, platelet 78

deaths associated with 368, 371

gpIb deficiency 93

fluid-attenuated inversion recovery gamma-hydroxybutyric acid 273

gpIIb-llla complex deficiency 93

(FLAIR) MRI, head 64

gas chromatography (GC) 261
glycosylated haemoglobin 112-13

fluorescein, bath salt 353

and mass spectrophotometry (GC­
gonococcal infection 42

fluorescent polarization immunoassay

MS) 261-2
grasping marks/injuries 295

(FPlA), drugs 259

gas embolism 364
grey platelet syndrome 95

follow-up skeletal survey 48

gastrointestinal tract
Griffiths Development Scales 14,

foot length (FL), fetal autopsy 150,

injury 68
17

187,471-2
sudden death relating to 238
growth

forcers), shaking injury to head 303

Gaucher's disease 100
assessment in abuse 12-15

forceps delivery, intracranial

General Medical Council and abuse 8
charts 475-80

haemorrhage 174
genes 397-8
growth plate (epiphyseal plate), in

forearm fractures, bilateral 53

mutated 410-11
abuse, injury, radiology 56

foreign body inhalation, asphyxial

genetiC (inherited) disease

death 330-1
bleeding due to 84, 90-3, 94, 94-5,
haemangioma, cardiac 234-5

forensic , meaning of term 7

96-7, 99
haematological disease 20, 76-105,

forensic specialists in abuse 22

metabolic see metabolic disease
132

joint working with paediatricians thrombophilias 364

haematological tests 81-2, 86-90

6-7,8-9, 11,458-60
genetic (inherited) factors, recovery bruised child 77

fossa, posterior, damage at birth 176

from traumatic brain injury 312
first-line 85-6

fossa navicularis 29, 31

genetic material, inheritance 395-8
patterns of abnormal results 86-90

signs of abuse/injury 38,39

genetic variation 396, 397-8
haematoma (and associated
fourchette, posterior 39
genitalia (and their examination)
haemorrhage) see
normal findings 29-30
in abuse 11-12
cephalohaematoma; intracranial
fractures 51-60
acute/healing/healed trauma 37-8
haematoma/haemorrhage
birth 71 , 161-2, 174, 189,218
boys see boys
haemoglobin

head injury and non-skull fractures

child protection examination form
carbon monoxide and 320, 322-3

298
461,466-7
HbAlc 112-13

non-accidental
girls see girls
haemophagocytic Iymphohistiocytosis

accidental vs 20, 70, 71, 72

mimics of abuse 40-1
84-5

facial bones, dentist's role 423

normal anatomy and variants 29-36
haemophilia 84, 85, 96-7

post- mortem examination for 154,

in sudden death 152
family histolY 84

155, 163
Germany, child car restraints 386-7
type A (factor VIII deficiency) 80,

radiology 50,51-60
germinal matrix haemorrhage 172-3
88,96-7

dating see dating

gestational (fetal) age
sex and 83

long-bone 52-6
autopsy standard measurements by type B (factor IX deficiency) 80,

sudden death (infant/early

471-2
96-7

childhood) 146
dental estimation 441-2
sex and 83

SIDS 215,218
facial bone estimation 444
type C (factor XI deficiency) 80, 84,

see also specific bones

girls
88, 97

France, child car restraints 387

growth charts 478-80
Haemophilus influenzae 248

frenulum, labial, injury 151,213,423,

sexual abuse, examination 29-36,
type b and acute epiglottitis 240

424
466
haemorrhage

fresh water, immersion in 347-8,357-8

findings 38-9
fundus see fundus

frontal collisions 393

technique 26
in hospital, fatal 363

head-on 393
Glanzmann 's thrombasthenia 78, 84,
intracranial see intracranial

froth, post-immersion formation 350

93-4
haematoma/haemorrhage

fult-thickness burns 324

Glasgow Coma Scale 282-3
nasal see epistaxis

fundus [manifestations of abuse) 126-7

gliosis 168
petechial see petechiae

haemorrhage 128 - 34
glucose
pulmonary, sudden death and 240

differential diagnosis 131-3

brain supply, inadequate 310
SIDS 215-16

examination with 133-4

levels, assessment 112-13
retroplacental 192

forces required 130-1

vitreous humour 1.12, 113, 150, 244
subaponeurotic/subgaleal 161,

mechanisms 129-30
see also hypoglycaemia
173-4, 301

- - ----- -
~ - -.=
 Index I 493

subconjunctival 126
health professionals see carers and high-performance liquid

subperiosteal 161
health profess ion a ls chromatography 260-1

sudden death du e to 240

heart coupled with mass

see also bleeding

arrest
spectrophotometry 262

haemorrhagic disease of the newborn

with dysrhythmias 365
high-tension electrical injuries 326

98
encephalopathy 172
histiocytoid cardiomyopathy

haemorrhagic shoc k encephalopathy

perioperative 369,370,371
206, 230

syndrom e 238
arrhythmias see arrhythmi as
histology (microscopic ap pearance)

haemosiderin 168
cond uction abnormalities 232
of brain injury in early life, timing

haemosideros is, idiopathic pulmonary

histological examination 162
by 167

215
examination in sudden death
immersion victims 347

haemostasis 77 - 8,78-81
external 154
lungs in fetal / perina ta l death 185-6

primalY 78-80
his tological, for conduction
prenatal dental , age estimation via

measurements 82
abnormalities 162
441

secondary 80-1
internal 155-6
subdural haem ato m a 153

hair samples (toxicology) 263-6, 272-3

muscle see myocardium
in sudden death 162-3

applications 272-3
surgery, death 370
SUm/SIDS 210-11,218

fetal exposure determination

weight 484
history-taking

273-4
fetal 471
abuse 9-10, 25-6

environmental contamination 268,

neonatal and infant 481-3
bleeding child 83-4

268-9
heart block 232
death (perinatal /infant/you ng child)

hanging by ligatu re 332

heart disea se 226-35
146

head
congenital see congenital SUDr see subheading below
circumference (He)
malformations
sum 200,201-2

charts 475-80
drowning an d 354, 356
adverse historical factors 202-3

fetal autopsy 471-2

neurogeni c 249
HIV (human immunodeficiency virus)

occipitofrontal 150
non-congenital/in general, sudden
infection 42

imaging see neuroim ag ing

death due to 226-35
human bites/scratches and risk of

moulding, excessive 188

instantaneous infant death 205-7
428

head injury 60-7, 282 - 317

heat
homicide (inc!. murder)

accidental 283, 293

injUly or death due to see thermal
bath tub drowning 355

cycling, and its prevention 338-9,

injUly or dea th
by fire 321

392-3
loss with burns 325
in hospital 373-5

mechanisms of brain injury 304

height
scene of death 142

coagulation abnorm alities and

child's (stature/length), assessment
see also filicide; infanticide;

101
abuse 12-13
n eonaticide; serial killing
death, scene 141
charts 475-80
homicide-s uicide see dyadic deaths
definition 282
post-mortem 150, 444
homocysti n uria, thromboembolism
differential diagnosiS 69-70
falls from 338 , 339- 40
235

epidemiology 283-94
see also crown-heel length ; horse-riding 341

imaging see radiology

crown-rump length ; foot length; hospita l, sudden death 362-84

non-accidental 63-7, 283-4,

umbilical cord exa mination causes 363-75

294-300
helmets see headwear
defin ition and frequency 362-3

dentists and 422-3

Henoch-Schon lein purpura 20, 84, 85,
investigation 375-7

diagnosis 296-8
94
hot fluid burns see scalds
investigations 299-300
heparin 86, 86-7, 98-9
house fires see fire fatalities
mechanisms of brain injury 305
hepatic injury/problems see liver
HPV 41 ,42,404
severity, classification 283
heredity see entries under genetic
human immunodeficiency virus see
see also brain injury; neurolo gical
Hermansky-Pudlak syndrome 81, 82,
HIV
lesions
94-5
human papillomavirus (HPV) 41,42,
head-on collisions 393
herniations, brain 3 11
404

headspace technique (in toxicology)

heroin, coercive administration
humeral fracture 55, 57

260, 262
272
hyaloid, posterior 129

headwear/helmets
high-molecular weight kininogens
hydroc ephalus 237

cycling 338-9,392-3
(HMWKs) 80
shun ts 369

winter sports 341

deficiency 97
hydrogen cyanide (fires) 322, 323

494 IIr Index

3-hydroxyacyl CoA dehydrogen ase immers ion into hot water 326 infections 247-8
deficiency, long chain 244 see also drowning and near­ ante-moliem \09, 110
hymen exa mination 26 drowning bite injuries 428
measurements 35-6 immune thrombocytopenia 96 brain, young children 177
normal appearance/variants 29 -31, neonatal 98 fetal see congenital infec tions
32-5 immunoassays for drugs 259-60 fundus haemorrhages seconda ry to
configuration 32 false-positives 268 132
ga pi ng/ narro w i ng/ a tten ua tion oral fluids 266 sexually-transmitted, tests for 41-2,
34-5 imp act injuries 404
notches/ c1efts/ tra nsectio ns to head 302 sudden death 247-8

33-4 non-accidental, signs of 294 airw ay infections 241-2

ridges/bumps/tags/bands 32-3 road traffic accident 388, 390, 391, in hospital 372

signs of abuse 39 392, 393 neonates 209, 372

hyp ernatraemic dehydration III
hypersensitivity/allergy 342
see also anaphyl ax is
hypeliension, pulmonary arterial
242
hype li hermia, malignant 371
hypertonic vitreou s humour III
hyp ertrophic cardiomyopathy 228-9
familial 228
inborn errors of met a bo lism see
s udd en death 156, 228-9
metabo lic disease
infant 206-7,228
hypofibIinogenemiaseefibrinogen
hypo g lycaemia 112, 113
from road traffic accident involvees
neonatal/infant 177
388-9
hy pohidrotic ectodermal dys plasia ,
X- linked 235
hyp otensio n, matern a l, due to injUly
192
hypo thermia
speed of impact see speed
see also shaken imp act syndrome
impac tion fractures 53
impetigo, bullous 20
impress ions (w ith bite marks)
bite 431
suspect 431
in vitro fertilization mix-up 4\0
146
post-mortem exc hange of (between
indu ced illness see fabricated illness
infant(s) and babies (up to I year)
death
adequacy of cause of, in SUDI
200-1
bath tub drownings 354
bucket/pa il drownings 353
insta ntaneo us, path ology 205-7
post-moliem examination,
see also specific pathogens/diseases
inflamm ation
ante-m ortem \09
SUD! 2 10-11
inflammatory mediators, burns 325
informatio n
from abused child 10-11
death (peI"inatal/infant/young child)
professionals) 163-4
recording see documentation
see also histOly-taking; in terv iew ing
informed consent see consent
inhalation (aspiration)
foreign body, asphyxial death
330-1

immers ion 349 babies/infants/young c hildren of gastIic contents see sto mac h

induced, head injUly 292 145-65 sm oke 318,320-3

hypotonic vitreous humour 111 rapid (in recognised illn ess), solvents see vo latile substances

hypovol ae mic shock with burns 325 path o logy 206, 207-8 water 332

hypoxaemia, head injury 290 sudden unexpected see sudde n inhelitance see entries under genetic
hypoxa nthine, time of death infant death syndrome; sudden injury/ traum a
estim ation 111-1 2 natural death; sudd en unexpected ch ild
hypoxi a (inc!. brain) \09 death in infancy accidental see ac cidental injury
drowning 349, 350 swimmin g pool drownin gs 356 asphyxia due to 330
hypoxic-ischaemic injury 167-8, 172 see also in fa nticide birth see birth trauma
non- accidenta l causes 297 genetic metabol ic disea se 114, 115 non-accidental see physical abuse
in shaken baby syndrome 307 growth chalis relative rankin g order of deaths due
boys 475
to 346
ICD see International Classificat ion of girls 478
sce ne of death from 141
Diseases neurological abno rmaliti es see
maternal , fet a l death due to 190-3
ice, spOliS injuIies 340-1 neurological les ions
see also specifIc sites and types of
identification of body/body remains newborn see neonates
injury
411-13
organ weigh ts 481-3
inqu ests and inquiri es 143
dental see dental identifica tion
infanticide 352-3 instrumental delivery
DNA techniques 411-13
definition
intracranial ha emo rrhage 174
fire fatality 320,4 12
England and Wales 181
sk ull fractures 30 1
missing persons 411-13
USA/elsewhere 181
insulin, levels 244
road traffic fatality 389
drowning 351,352-3,355
insulin-dependent diabetes melli tus
imaging see rad iology scene 141
244, 245
 Index I 495

intensive care unit

jaws, radiography see radiography
DNA samples 403

head injury 290-4

jellyfish venom and drowning 358
fetal/perinatal death 181-2

road traffic victim dying in 390

joint bleeding 88, 89
Leigh's syndrome 177

intentional acts see non-accidental joint working in abuse cases,

length see height

acts paediatricians- forensic specialists

lens damage 126

interagency dimensions see 6-7, 8-9, 11 , 458-69

leucodystrophies 238

multi-agency and see also mUlti-agency and

leucoencephalopathy

multidisciplinary dimensions mlJ1tidisciplinary dimensions

multicystic 171-2

International Classification of Diseases telencephalic 171

(ICD)
karyotype 396
leucomalacia, periventricular 169, 171,

head injury 283

Kawasaki's disease 231-2
211

sudden death - cause unknown

instantaneous infant death 205, 206
leukaemia 84, 99

198
keratin matrix, drug isolation from
fundus haemorrhages 132

interviewing at scene of road traffic

265
lichen sclerosus et atrophicus 20, 40

accident 388
ketamine 272
life support systems, infant deaths

intestine (bowel)
ketones 113
following over 12 hours on
enquiry on protection examination
kidney
199-200

form 461
disease and failure
ligature, hanging by 332

injury 68
ante-mortem 113-14
lighting, scene of sudden death 139

sudden death relating to 239

bleeding in 83, 100-1
lightning 327

intracranial haematoma/haemorrhage injury 68

likelihood ratios, DNA evidence 402

in abuse 61-3,296,297,306-7
SIDS 211
linea vestibularis 41

chronic 296, 306

weight 484
lipid (fat)

fundus haemorrhage and 130

fetal 471
accumulation in liver, SIDS 211-12

birth-related 71, 174-5, 189-90, 302

neonatal and infant 481-3
depletion in adrenals, SIDS 212

imaging 71,289
kininogens see high-molecular weight
overload in intravenous feeding

differential diagnosis of cause 70

kininogens
364

maternal abdominal trauma causing

see also fat embolism

191-2
labia minora and majora, examination lipid storage disorders 100, 177

perinatal 172-3
26
liquid chromatography

birth-related see subheading above

normal anatomy 29,31
high-performance see high­

death due to 189-90, 193

signs of abuse 39
performance liquid

in sudden death 236-7, 240

labial frenulum, injury 151,213,423,
chromatography

estimation of blood loss 152

424
ultraperformance 261

in sudden death, examination for

laboratory tests
liquid-liquid extraction (in toxicology)

156
biochemical tests on post-mortem
260

infant 218
specimens see biochemical tests
liver

neonatal 160, 161

in haematological disorders see
failure 100-1

traumatic 293-4
haematological tests
injury

birth-related see subheading above

microbiological, sudden death
abuse 68

non-accidental see abuse

(infants/early childhood) 148-9
at birth 162

(subheading above) samples/specimens for see samples

siderophages 217

intracranial pathology, sudden death

in sexual abuse, interpreting findings
in sudden death, assessment and

237-8
42-3
findings 154

intracranial pressure, raised 130

labour, prolonged or difficult 187-8
SIDS 211-12,217

head injury 291-2, 307

lakes, drowning 357-8
tests, fatty acid oxidation defects

intradural haemorrhage 174, 307

larynx
116-17

birth-related 174
cyst 207
weight 484

intrauterine damage and death see

malformations 193
fetal 471-2

fetus
spasm with immersion 350
neonatal and infant 481-3

ischaemic stroke in mild head injury

laser, portable, road traffic fatality
local anaesthestics in dentistry 371

290
389
long-bone fractures, radiology 52-6

see also hypoxic-ischaemic injury

legal issues (inc], legislation)
long chain acyl-CoA deficiency 244

islets, pancreatic, sum 218

abuse 1-2, 4, 7-8, 22
long chain 3-hydroxyacyl CoA

isotope bone scan, abuse 51

dentists'role 421
dehydrogenase deficiency 244

Italy, child car restraints 387

cycle helmets 393
long QT syndrome 207, 232, 365

496 II: Index

low copy number (LCN) analysis 400

maternal issues see mother
normal variants 70

sex ual assault 404

mattresses 203, 204
methadone 273

lumba r puncture 368-9

May-Hegglin anomaly 81,95
methylamphetamine, prenatal

lun gs
Meadow, Professor Roy 16, 199, 202,
exposure 274

in drowning, pathology 350,351

274, 455
methylphenidate 257-8,272-3

in Feta l/perinata l death,

meconium
micellar electrokinetic capillalY

examination/assessment 185-6
staining 159, 187-8, 193
chromatography (lVlECC) 261

maturity 187
in stomach 185
microbiological samples, sudden death

haemorrhage see haemorrhage

toxicology 267
(infants/ea rly childhood) 148-9

oedema, postoperative 371

median perineal raphe, congenital
micro dialysis catheter 293

in sudden death 215-16,240

abnormalities 41
microscopic appearance see histo logy

examination 154, 155, 156

medical assessment, abuse see child
midbrain shearing injury 307

histology 162
abuse
mineralization, brain 168

inFant, pathology 210

medical conditions/illnesses/ diseases
missing persons, DNA tests on body

weight 484
death due to 225-55
411-13

Fetal 471 - 2
bath tub drownings 356
mission statement of expert 448-9

neonatal and inFant 481-3

infant see subheading below
mitochondrial DNA 397, 407-8

lupus anticoagulant 87, 89

infant death due to 225-55
analysis 407-8

lymphadenopathy, SUDI/SlDS 210

rapid 207-8
abandoned body (baby/fetus ) 413

lymphoblastiC leukaemia, fundus

simulating/mim icking abuse 20,
paternity tests 411

haemorrhages 132
69-70, 71-3,76-105, 13 2
diseases of 244-7

lymphohistiocytosis, haemophagocytic
haematological 20, 76-105, 132
babies 177

84-5
sexual abuse 20,40- 1
cardiomyopathy 230

see also sudden natural death

sudden death 244-7

maceration
medical devices/procedures 368-71
molecular methods, post-mortem 150

Fetal, measuring 471-2

medical history
see also DNA tests

neonatal 158-9, 184

family 10
molecularly imprinted polymers (in

macrocyto pathies 95
perinatal/inFant/young child death
toxicology) 260

macrophage response,
146
monitoring (physiological)

hypox ic-ischaemic injury 167-8

SUDI 201-2
death to failure to monitor 365-6

magnetic resonance imag in g (MRi) of

medium chai n acyl-CoA deficiency
in head injury 290-4

head 12,50-1 , 61-7

116, 117, 239, 245-6
monozygotic twins 396

accidental injury 293

sudden death 239, 245-6
Monstrad's dental age estimation

birth injury 289

infants 208
method 443

non-accidental injury 12, 61-7,

meningitis
mosaicism 410

299-300
coning 369
mother

malabsorption syndromes 101

menin gococcal see meningococcal
DNA tests with abandoned baby/fetal
male children see boys
infection
remains 413

malignancy, bone marrow infiltration

meningococcal (N. meningitidis)
filicide 353

100
infection 247
infanticide 352

malignant hyperthermia 371

meningitis 247
pregnant

malnutrition, post-mortem assessment

fundus ha emorrhages 132
complete separation from,
108
metabolic disease 177
definition 181

manslaughter, drug/anaesthetic errors

of bone 71,94
drug abuse 257, 267, 273-4

366, 367
genetic (inborn errors of metabolism)
injury causing fetal death 190-3

Marfan's syndrome 235

fundus haemorrhages 132,177, neonaticide by drowning 352

Marshall CT score 293, 294

204-7
motor control in abuse

mass disaster
Reye-like 239
assessmen t 14, 17

dental identification 439

sudden unexpected death see
ocular 128

fire 321
sudden unexpected death motor vehicle accidents see road traffic
mass spectrophotometry
post-mortem investigations 150
accidents; vehicles

gas chromatography and (GC-MS)

acute metabolic autopsy 119-20
mouth see oral cavity

261-2
inFants/yo un g children 150
mucocutaneo us bleeding 88

liquid chromatography and (LC-MS)

metaphyses multi-agency and multidisci plinary

262
Fractures 53-6
dimen s ions

time-oF-flight 262
at birth 71
abuse 5-6

mast cell tryptase 109, 110, 239

in scurvy 71
sudden death

 Index I 497

infan t (SUD1) 201

post-m0l1em examination 158-62
see also pregnancy
and post-mortem information
dental age estimation 442
ocular problems see eye
163-4
drug tests on hair 273-4
oedema (swelling)
see also joint working
organ weight 481
bra in/cerebral 66-7,309-10,364-5

multichannel EEG, head injulY 292

neonaticide 181 , 352
ea rly life 167

multicystic leucoencephalopathy
drowning 351,352
in hospital 364- 5

171-2
in hospital 372
imaging 66-7,310

mummified remains 183

neoplasms see tumours immersion v ictims 351

Mun chause n's syndrome by proxy see

nerve fibre layer of retina, traumatic causes 66, 309-10

fabricated illn ess

h aemorrha ge into 128
pulmonary, postoperative 371

murder see homicide

nerve injury, bi!1h-related 302
Offences Against the Person Act (1861)

muscle, cardiac see myocardium

nerve root injury, birth-related 176
181

mutations 410-11
neu roblastoma 84
ongoing health care, abuse 21

myelodysplasia 96, 99 - 100

neurogenic heart disease 249
ophthalmology see eye

myocarditis 110, 227-8

neuroimaging (imaging of head/brain)
opiate tests, false-positives 268

myocardium (ca rdi ac muscl e)

60-7, 3 10
opinion in suspected abuse

examination in sudden death 155,

in abuse 12,60-7,299-300
expert wit nesses 450

156 CT 12,50, 61-2, 299-300

formulating 21

infarction with cardiomegaly 233

MRI 12,50-1,61-7,299-300
optic nerve lesions 127,1 28

neurogenic disease 249

ultrasound (infant) 50
oral cavity (mouth)

tumours 233-5
accidental head injLllY 289-90, 293
examination in sudden death 150-1

myristic acid oxidation ass ay

birth injury 289
fluids in toxicology 266

117, 208
neuroin tensive care, head injulY
injuries in abuse, dentists' role

myxoma, cardiac 234

290-4
423-4

neurological lesions/damage 166-79

surgery, deaths 371

napkin dermatitis 217

electricity-induced 327
organ(s) see viscera

nasal decongestants a ffe cting

infants/early life 166-79
organic acid disorders 115

amphetamine tests 268

at bil1h see birth trauma sudden death 114, 117

nasal haemorrhage see epistaxis

clinical manifestations 167
orogastric tu bes 368

nasogastric tubes 368

traumatic see brain inju ry
osteochondrodysplasia 236

natural death see sudden natural dea th

see also brain; periph eral nerve
osteogenesis imperfecta (brittle bone

nea r miss events see app arent life-

injury; spinal cord disease) 20, 71- 3

threatening events neurones

expert witnesses and 451

neck
death 167
types I-IV 72

dissection 154
migration disorders 170, 217
osteomyelitis 71

injury
newborns see neonates
ostial ste nosis 230

muscles at birth 161

nicotine, neonatal hair testing 273-4
overheating, sleeping infant 203, 204,

see also cervical spi ne

9/11 disaster (2001), identitying
205

necropsy see post-mortem

remains 412
overlaying 203, 2 15, 33 1

ex amination
non - a cci denta 1/ in te n ti 0 n a 1/ deli bera te
~ - ox id at ion see fatty acid ~ -o x idation

necrosis, brai n cells (in early life) 167

acts
defects
necrotizing enterocolitis 212
biological speci men alteration 269-70
oxidative phosphorylation (electron

necrotizing fasciitis 248

drownin g 354-6
transport ch ai n) defects 116, 244,

needles, multipl e, penetration by 69

injury see physical abuse
246-7

neglect
pOisoning by health professionals
oxygenation

assessment 16, 458-69

274
brain, inadequate 310

dental 422, 424

non-steroidal anti-inflammatory drugs
monitoring status in head injury 29 0

Neisseria meningitidis see

99

meningococcal infection
noxious gases in fires 322
pacemaker

neonates/newborns
nucleic acid amplification tech niques
failure 369

bleeding disorders 97-8

(incl. PCR) 400, 403, 407
post-mortem assessment 377

dea th/sudd en de ath 249, 371-2

sexually-transmitted infections 404
paediatricians/ pa ediatric doctors in

abandoned body see ab andoned nutritional status of deceased 107-8

abuse 22

bod y
joint working with forensic

causes 209
observing the scene of sudden death
specialists 6-7,8-9,11,458-69

definition 180
139-40
paedophilia-related de aths 142

in hospital 371-2
obstetric events, death related to 248
pail, drowning in 353

498 I Index

pancreas
J t scene of sudden dea t h 138
mother, ca usin g fetal death 191

histology in sudden death 244

in tervi ewin g 388
see also specific parts of body and
injury 68
petechiae 84
types of injury e.g. bruises;
islets, SUDl 218
in fetal/perin ata l dea th
frac ture
pan creatitis, acute, en ce phalopathy cutan eo us 188
physical exa min ati on in ab us e 11- 12

associated with 248-9

pulmonary 185
on child protection exa mination

paracetamol poison in g 149

sh owers 18
form 462-7

paren ts/i m med ia te family

in sudden de at h
sex ual a buse 25

in sudden death, considera tio ns 142-3

in asphyx ia 329-30
te chn iq ue 26-7

ro ad tra ffic acciden t 389

neonates 159
see a/so externa l exam inat ion

in suspected a buse, presenting

observed at scen e 140
physiological monitoring see

history from 10
in sudd en death in infan ts
monitoring

see also carers; mother

(SUDI/SIDS)
pigmentation, hair, dntg concentration

parotitis, cytomegalovirus 250

co njuncti va l, possibility of non ­
an d 264, 265

pa rtial -thic kness burns 324

accidenta l injury 213-14
pillows 203. 204

partial thromboplastin time with

thymic/ pulmon aJy/coronary artelY
place nta

kaolin see activated partial

210
abruption 192

th rombo pl astin time

ph alangeal injuries 59
DNA tests with aband oned

particulates (sm oke) in a irway 320,

pha rm acok in etics 258, 269
ba by/fetus 4 13

32 1
Pha rmChek'" 266-7
exa min at ion 160-1

patch testing, sweat 266-7

ph aryn gea l perforation 68- 9
we ight 473

paternity testing 408- 10, 412

phospholipid an tibody (lupus
plas mi no ge n ac ti vator inhibito r-l
mtDNA 411
anticoagu lan t) 87,89 deficiency 93

mutations affecting 410

pho togra phic evidence plastic bag asphyxia 331

Y chromosome 411
bite injury
platelets 78-9

pedestrians (in car acc idents) 386,

marks 430- 1
acti vation 78

38 7,388,390,391.392
susp ec t 431
adhesio n 78

run over by reversi ng car 330

on child protection examination aggregation 78

veh icle inspectio n 388

forms 468
measurement 82

pelvic fra ctu re 59

dental iden tificat ion 438
diso rders 78. 93-4

penetrating injuries/ tra um a

fun du s hae morrh age \ 33 -4
drug-induced 99

head 294. 302

ro ad tra ffi c accident 38 7
dysfunctional 82 , 99

ma te rn al abd omen . causing feta l

sex ual abuse 28
morphological 8 1, 94-6

death 19 2
sudd en death 147-8
neonatal 97-8

oral cavity. in abuse 425

infants/youn g children 147-8
numerical 90. 94-6, 97-8

v iscera 68-9
scene of 138
sto ra ge po ol 92

penile injury 15
physical abuse/violence (non­ enhancem en t of coagul ation 79

perfusion. cerebral see cerebral acc ide nta l/in te ntional inj ury ; secretion/ rel ease 78-9

perfus ion ; cerebral perfusion battered child syndrome) measureme nt 82

pressure child 47-76, 125- 36, 420- 34

pl ayground/recreation al area injuries

perianal area see anus and perianal

airway obs truction 332 -3
339,340

a re a
assessing pattern s of injury 17-l[l
fa lls 340

perinatal death 180-97

deaths per 100 000 337
poisoning see toxic substances;

p ost-mortem exa mination 145-6 5

dentist a nd see dentist
veno ms

perin eal raphe. median. co ngen ital

differen tial diagnOSis 69 -73
polari ty of drugs a nd hair sam ples

abnormalities 41
drownin g 354, 354-5
264

perios tea l reaction 51-2

expert witnesses 451
polycystic kidney disease, autosomal

in frac ture repair 50

head injury see head injury
dominant, subarachnoid

physiological 51-2. 70
medica l co nditions simulating see
hae morrha ge 236

peripheral nerve inju ry. birth-related

medical conditi ons
p olymerase chain reaction (PCR) see
302
ocul ar invo lvement see eye
nucleic ac id a mplificatio n

periventricular leu comalacia 169, 171 .

radiolo gy see radiology
p olymicro gy ria 170-1 , 177

171 ,2 11
SID S and p ath ol ogy raiSing
polymorphisms 399

perso nnel
co ncerns 212-17
restri ction fragment length 398-9

health see carers a nd health

spinal injury 312
short tandem repeat 399

professionals
ther ma l injuries see th erm al injury
Sin g le nucl eotide (SNP) 405, 41 3

 Index I 499

polystyrene-filled bean-bags 204

girls 478
abuse 12,47-76

polythene bag asphyxia 331

bronchopulmonary dysplasia 210
differential diagnosis 69-73

porencephaly 170
sudden death 210, 248
head see neuroimaging

posterior seg ment (eye) injury 126-7

procalcitonin 109
radiologist's role 47

post-mortem
professional witness 449-50
head injury see neuroim aging

CT scan 50
proforma 11, 17,458-69
immersion victims 347

drug changes 270-1

prone sleeping position 203
mod a lities 48-51

skeletal survey 48-9, 146

protein , post-mortem evaluation 108
see also specific modalities
specimens, biochemica l tests
protein C deficiency 132
road traffic fatal ity 389

106-26
prothrombin deficiency 89, 97
sudden death 146-7

post-moliem examination (autopsy;

prothrombin time (PT) 81
neonatal 158

necropsy)
prolonged 89-90
radiolluclid e bone scan, abuse 51

babies/infants/young children 145-65

psychological features of sexual abuse
random match probability 401-2

SUDI 200, 205-18

25
raphe, median perineal , congenital

birth trauma 161-2, 188-90

psychological maltre atment (emotional
abnormalities 41

dental identification and 438-9

abuse), assessment 15-16
rear-end collisions 393

fetal/perinatal death 183-90,194-5

psychological stress and emotional
records see documentation

injuries 188-90
trauma
recovery room, death in 371

live birth determination 183-6

immediate family in sudden death
recreational a rea see playground

standard measurements 471-2

142- 3
recreational drugs see drugs

hospital death 375-9

road traffic collision witnesses 388
rectal perforation 68

intraoral. in abuse 424

psychometric tests in abuse 14
rectal tempera ture 146

road traffic accident 389-90

public swimming pool drownings 357
referral in abus e, initial 5

technical considerations at time of

pugilistic attitude of body (fires) 319
reflex anal dilatation 36-7,40

117-20
pulmonary arterial hypertension 242
reflex pulmonary vasc ular

post-mortem interva l 111-12

pulmonary embolism 235, 364
vas oconstriction 350

post-nata l depression 352

pulmonary non-vascular tissue see
renal injuly/c1 isord ers see kidney

potassium levels, post-mortem 110,

lung reperfusion injury 212

111
pulmonary vascular vasoco nstriction, reporting of abuse

time of death estimation 111-12

reflex 350
dentists 420-1

prealbumin, post-mortem evaluation

pulmonalY veno -occlusi ve disease 242
written report 16-17

108
pulp, dental, DNA from 444
see also docum e ntation

pre-excitation (Wolff-Parkinson­
pul se oximetry, head injury 290
respiratory chain (electron transport

White) syndrome 232, 236

purpura 83
chain; oxidative phosphorylation)

pregnancy
Henoch-Schonlein 20 , 84, 85, 94
defects 116, 244, 246-7

concealed 181. 182,352

idi opathic thrombocytopenic 94, 96
respiratory depression as adverse
drug abuse in 257, 267, 273-4
showers of 18
sedation event 367

termination see a bortion

Purtscher retinopathy 129
respiratory syncytial virus 242

see also fetus; mother; obstetric

pyrexia in brain trauma 292
res piratory trac t

events
pyruvate dehydrogenase deficiency
inh a lation into see inh alat ion

prekaJikrein 80, 81
177
in sudden death 240-2 , 243

deficiency 88, 97
SUDI and respiratory symp toms

premature babies see pre-term babies

QT interval, prolonged (long QT
202

prenatal child see fetus

syndrome) 207, 232, 365
see also airways; lung

pre-retinal hae morrhages 128-9

restraints, child (in cars) 386-7,388,

pre-school children see toddlers and

radicul ar (nerve roots) injury, birth­
391

pre-school children
related 176
restricted-access material s (RAMs) 260

presentation (fetal) 188

radiography
restriction fragment len gth

presenting history/clin ical presentation

dental Ua ws and teeth) 438-9
pol ymorph isms 398-9

9-10,460
prenatal age estimation 441
resuscitation (cardiopulmonary; CPR)

bleeding child 83-4

skeletal survey see skeleton, survey
changes induced by 212

from parent 10
see also digital dental radiographic
retinal haemorrhages 129, 132

pre- term/premature babies 248

identificat ion
rib fractures 58,212

body weight/height and head

radioimmunoassay for drugs (RIA)
immersion victims 351

circumference charts
259
road traffic acc ident, unsuccessful

boys 475
radiology 47-76
388

~ ~ ------~---------~ ----- ""'"

500 I In dex

retinal detachme nt 126

samples/sp ecim ens
screening tests

retinal haemorrhages (and over/under

collection and processing (for
coagulation 8 1

retina)
laboratory tests) 12
drugs 259-61

cardiopulmonary res uscitation

for biochemical tests in sudden
doping agents 271-2

ca using 129, 132

death 149
pitfalls and limi tations 267-71

differentia l diagnosis 131

for DNA evidence see DNA tests
ScUlVY 71 , 94

exami nation with 133-4

for drug detection 260, 262-7
sea/salt-water immersion 348, 358

mech an isms 129

for haematological tests, pitfalls 86
seats and seat belts (carl, child

non- accidenta l 296

hospi tal deaths 376, 377
386 - 7

shaking injury see shaking

for microbiological tests in sudden
Sebastian platelet syndrome 95

retinal haemorrhages (and over/ under

death 148 - 9
second-degree burns 323

retina) 128-9 13 2, 133

sa li va traces w ith bites 429
security, scene of sudden death 13 9

retroph aryngea l abscess 240-1

sexual abuse 86
sedation, adverse o utcomes 367

retro placental haemorrhage 192

for toxicological tests in sudden
seizures (epileptic)

reverse suspens ion 333

death 149
retin al haemorrhage 132

Rey e's and Reye- like syndrome

po st-mortem, biochemical tests see
sudden death 242-3

238-9
biochem ica l tests
by drowning 354, 35 5, 35 7

Reyn eJ Developmental La nguage

skin, with bites 431
SUDr and history of epilepsy 203

Scale 14
scalds (hot fluids/ liquids) 324, 355
self-inflicted bite marks 19, 427

rhabd omyoma, card iac 233 - 4

non-accidental 235-7, 355
self-strangul ation 331-2

rheumatic fever, acute 228

scalp
semen

rib fractures 56-8

injury
evidence of 404, 405

at birth 71, 161-2,215

birth-related 173-4
samples 28, 403

CPR-related 58 , 212
non-accidental 294
septic shock 363-4

post-mortem examination for 154,

neonatal post-mortem examination septicaemia

155, 161-2
159
H. injluenzae 248

radio logy 56-8

scapu lar fract ure 59
meningococcal 247

sudden death (infants/early

scene
serial killing, carers/health

childhood) 146
fire fatalities 141,319-20 profeSSionals 274, 373 - 5

SlDS and 212,215

sudden unexpected death 13 7-44
sewers, drow nings 358

rickets 71 , 215
asphyxia l deaths 330
sex

ri ding injuries 341

babies/young children 145-6
denta l determination 443-4

ri vers, drowning 357-8

definition 140
haemophilia and 83

road traffic (motor vehic le) accidents

good look around 139-40
sex chromosomes see X chromosome;

338, 385-94
management 137-8
Y chromosome

asphyxia l death 330

road traffic collision 38 7
sexual abuse 24- 46

co llision (MYC) 338, 38 7-9 1

sens itivity and stress of
definiti on 24

fetal injuries and dea th 192- 3,

investigation 142-3
diagnosis/investigation 7, 24-46

193,393
sequence of events 139
conditions simulating 20, 40-1

conseq uences 338

schizencephaly 170-1
consisten t vocabulary 29

fatal 385-90
school-age children
DNA tests 403 - 7

asp hyx ial death 330

growth charts
interpreting findings 42-3

epidemiology and causes 385 -7

boys 477
'jigsaw'in 4

investigation 38 7-90
girls 480
skills and experience for 28-9

pedestrians in see pedestrians

see also adol escents
drug-facilitated 273

root (tooth) pattern 437

organ weights 484
epidemiology 2, 24- 5

Roth spots 128

Schour and Massier charts 442-3
expe rt witnesses 43, 451

route of drug adm inistratio n, wrong

scintigraphy, abuse 51
forensic ev idence 27 - 8

366
Scotland
sex ually -transm itted in fectio ns , tests

RSV 242
abuse
41-2, 404

Russell's viper venom time, dilute 87

bite injury 431
's haken baby ' syndrome 283, 297

legal dimensions 2, 3, 7
death scen e 141

safety (seat) belts 386, 387

concealment of birth 181
diagnosis 297

sa liva traces with bites, evidence of

expert witnesses 454- 5
encephalop athy 295, 295-6

429
road traffi c casualties and fatalities
hypoxic ischaemi a in 307

sa lt/sea water immersion 348, 358

390
shaken impact syndrome 283, 296

. ..
_...l.~_
--­

 ~,:'- _I .

Index I 501

shaking 303-6
skin
speed (vehicle)
cervical spine injury see
burns see burns
of impact

cervicomedullary syndrome
disorders 20
deaths related to 386

head injury 303-6

excision of samples with bites 431
determining 387

biomechanical determinants 303

see also external exami nation
limit, casualties by 392

mechanism 303
skull (cranium)
sperm/spermatozoa 395

predisposing factors 303

birth-related injury 17 3-4
evidence of 404, 405

pathophysiology 306
fractures 60-1, 294, 298
samples 28

retinal haemorrhage 296, 29 7

biomechanics 298
spinal cord
arousal shaking 132
birth-related 174, 189,301
injury 311-12

unilateral 134
non-accidental 60-1 , 294-5
birth-related 176, 302, 312

shearing injury, brain 65-6, 307

short falls as cause 298
cervical see cervical spine and

imaging 65-6
in sudden death (infants/early
spinal cord trauma

shock
childhood) 146
wit hout radiographic abnormalities

haemorrhagic shock encephalopathy

normal variants 70
(SCIWORA) 311

syndrome 238
post-mo rtem exa mination, neon atal
post-mortem examination 157-8

hypovolaemic, with burns 325

159
neonatal 160

septi c 363-4
birth-related 161
spin al trauma 58-9

spinal 311-12
post-mortem opening 156
cervical see cervical spine

toxic shock syndrome 110, 248, 364

neonatal 159-60
perinatal 193

short chain acyl-CoA defici ency 244

slap, pa tte rn of injury 18
spiral fracture of tibia 53

short tandem repeats (STRs) 399-402

sleeping environment (SUDI/SIDS and)
spleen

identification of body remains 411,

203-5
injUlY 68

413
asphyxial deaths 204, 205, 331
weight 484

interpreting data 401-2

in hospital 372
fet a l 471-2

mutations in 410
small bowel, sudden dea th relating to
neonatal and infant 471-3

paternity testing 408

239
sports

sexual assault 404

smoke (from fire)
dopin g agents see doping agents

technique 399-401
detectors/alarms 318-19
win ter, injuries 340-1

shunts in hydrocephalus 369

inhalation 318,320-3
sta ture see height

sickle cell disease 240

poisoning 322
steroids, anabolic 258, 2 71-2

side collisions 390-1, 393

smoking
still birth, defin itio n 180-1

side s leeping position 203

house fires related to 318
stomach

siderophages
passive
aspiration (of foreign material) fro m

hepatic, SlDS 217

neonatal hair testing 2 73-4
anaesthesia-related 370

pulmonalY, SlDS 2 15-16

SIDS and 203-4
SIDS an d 2 16,2 17-18

significant harm 4
smothering
live birth determination, examination
signs of life, definition 181
accid ental (=overlaying) 203 , 215,
185

simulated illness see fabricated illness

33 1
rupture 68

single nucleotide polymorphisms

non-accidental 203
in sudden death

(SNPs) 405, 413

snow, spol1S injuries 340-1
extern al examination 154

sinovenous thrombosis, cerebral 177

social histOlY 10
infant 2 10

skeletal dyspl asia 236

sodium levels, post-mortem 110-11,
internal examination 15 6

skeleton
150
stonefish 358

birth trauma 71
sodium val pro ate 99
stools, live birth determina tion

post-mol1em examination 161-2

sofa, s leeping on 205
185

injuries 51-60
soft-tissue injury 69
strangulation

at birth see subheading above

solid-phase ext raction and
acc idental 331-2

head injury and 298

microextraction (in toxicology) bruising patterns 18

non-accidental 51 -60, 294-5

260
Streptococcus group A

see also fractures

solvents see vo latile substances
0-hemolytic 247-8

normal variants 70
soot inhalation 320,321,322
invasive 248

survey (in abuse) 12, 48-51

spa pool drownings see bath tub
stress see psychological stress
follow-up 48
drown ings
stroke

post-mortem 48-9, 146

Spain, child car restraints 387
developing brain 176-7

skiing sports injuries 340- 1

specimens see samples
ischaemic, in mild head injury 290

#
- -------------------- ~- --------------­
502 I Index

subaponeurotic (subgaJeaJ)
in hospital 363, 363-5
technetium-99m scan, abuse 51

haemo rrhage J 61, 173-4, 30 I

infant 225-55
teenagers see adolescents

subarachnoid haemorrhage 63, 307

pathology 205-9
teeth see dental identification; dentist;

birth-rela te d J 76, 190, 302

'intermediate' pathology 249-50
dentistry
fundus haemorrhage and 130,
unexplained
tem pera ture
130-1
distinction from explained sudden
abnormal see hypelthermia;
post-mortem examination of neonate
death 226
hypothermia; pyrexia
for 160
older children 249
body

sudden death 236, 237

see also medical conditions
in head injury, monitoring 291-2

subconjunctival haemorrhage 126

sudden une xpected death 137-65
rectal, in sudden death 146

subdural haematoma/haemorrhage
diabetes mellitus 113
environmental, sleeping infant 203

in abuse 61-3,71,296,297,306
genetic metabolic disorders 114-17,
water, drowning and 348-9

chronic 296, 306

149, 244-7
see also heat; thermal injury

fundus haemorrhage and J30-1

infant 114, 208
tentorium, neonatal examination 160

at birth 71,175-6,189,218,289, neonate in hospital 372

Terso n's syndrome 130

302
in hospital see hospital
tet racyclin e and dental age estimatio n

location 175
nervou s system examination 166
443

natural history 175-6

older childre n 142,249
THO I (short tandem repeat) 388

differential diagnosis of cause 71

post-mortem examination
thalamus, hypoxic-ischaemic dam age

perinatal death due to 189, 19 3

145-65
172

in maternal abdominal trauma

scene of see scene
therapeutic misadventures, death 363,

191-2
sudden unexpected death in infancy
366-71

predisposing conditions 306

(SUD! - inc!. babies) 198-224
thermal (heat-induced) injury or death

in sudden death 156

clothing 139
318-28

histology 153
epidemiology 201-3
non-accidenta l 325-7

infant 218
genetic metabolic disease 114, 208
dentist's role 423

neonatal 160, 161

histolo gica I examin at ion 162
pathologist's role 319-27

traumatic 294
pathology 205-18
patterns 20

non-accidental see abuse

findings compatible with SIDS
see also burns

(subheading above)
209-12
th ird-d egree burns 323-4

subendocardial fibroelastosis 207

findings of questionable
thorax see chest

subgaleal haemorrhage 161 ,

significance 217-18
thrombasthenia, Glanz mann 's 78, 84,

173, 301
findings raising possibility of
93-4

subperiosteal haemorrhage 161

non-accidental injury 212-17
thrombin time (TT) 81

subretinal haemorrhages 128-9

terminology/definitions 198-203
long 89-90

substance misuse see drugs

see also sudden infant death
thrombocytopenia

sudden death - cause unknow n,

syndrome immune see immune

leo use of term 198

suicid e thrombocytop enia

sudden infant death syndrome (SIDS)

by perpetrator of homicide see neonatal 97-8

198-224
dyadic death thrombocytopenic purpura, idiopathic

epidemiology 201-3
in hospital 373
94, 96

expert witnesses 455

supin e sleeping position 203
thromboembolism 235,364

genetic metabo lic defects 114,

surfactant in water inhalation 350
thromboplastin see activated paltial

208
surgery, deaths assoc iated wi th
thromboplastin time

hypoxia 109
369-71
thrombosis, cerebral sinovenous 177

non-SIDS or 140
investigation 375
thromophilias, inherited 364

post-mo rtem findings compatible

sweat, toxicology 266-7
thymus

with conclusion of 209-12

swelling see oedema
in sudden death 154

predicting 201
swimming pool drownings 356- 7
infant 210,211

terminology/definitions 140,
syphilis, congenital 71
weight

198-203
systematic enquiry in abuse 460
fetal 471-2

tox icology and 274

systemic diseases, bleeding tendency
neonatal and infant 481-3

see aIso infants ; sudden unexpected

100-1
tibial fractures 53, 54, 55, 56, 57

death in pinfancy
systemic lupus erythematosus, AV
time (determination)

sudden natural death 225-55

block 232
of brain damage in early life 167

fetus 193
systemic respon se to burns 325
of death 111-12

- - :, ._ . .
Index I 503

of sudd en death 139

twins
VEGF 109

infant 199
monozygoti c 396
vehi cles

ti me-o f-fli g ht mass spec trophotometry

simultaneous SUD! 201
agricul tural 341

262
road see car; ro ad traffic acc idents

toddlers and pre-sc hool children (1-4

Uhl's anomaly 23 0
ve noms, sea creatures, a nd drowning

year olds)
ultrape rform ance liquid
358

bath tub drownings 354

chromatography 261
veno-o ccl usive di sease, pulmona ry

bucket/pail drow nings 3 53

ultrasound 49-50
242

dental age estimation 442- 3

see also transcranial Dopple r
venous ca theter-related deaths 368

growth cha rts

um bilical cord examination 159
venous embolism

boys 476
insertion/s tump 184- 5, J8 8
air 364

g irls 479
length 474
thrombotic 364

organ weights 484

'unasce rtained' or 'und eterm ined'
venous infa rction, brain parenchyma

swi mming pool d rowni ngs 356-7

ca use of dea th 219, 377
173

toilets, deli very into UlJ, 352

uncloth ing body
ve nous pressure, intraocul ar, raised

touch in g, non-consensua l 42 5
road traffic fatal ity in hospital,
129

tox ic shock syndro me 110, 248, 364

retrieval 388
ven ous throm bosis, cerebral 177

toxic sub sta nces (inc l. drugs and

scene of sudd en death 139
vent ilation, ass isted, cha nge due to

poison s) 256-81, 342

'undetermined' cause of death 219,
21 2

bleeding due to 98-9

377
ventouse delivery, intrac rani al

epidemiology 256-8, 337

United States, repo rting of ab use 421
haemorrhage 17 4

fi res 322
uraemi a, vitreous humour a nalysis III
ventricles (brain), 3rd, colloid

infants/early childh oo d deaths 149

urea (and urea nitrogen), post-mortem
cyst 23 7

sta tistics on death s from 337

111 , 114, 11 8, 150
ventricles (hea rt)

testing 259 -74

urea cyc le disorders 11 5
arrhythmogenic righ t ventri cu lar

ca rer-associated seria l killin g 374

sudd en death 117
cardiomyopathy 229-30

immersion victi ms 347

urethral prolapse 41
septal defect 227

pitfalls and limitations 267 -71

urin ary bladder, SJDS 2 10
v entriculoatrial and

sa mpl es 260, 262-7

urinary tract enquiry on protection
ventriculoperiton eal shunts 369

specific app lication s 271-4

exa mination form 461
vernix caseosa, dru g test in g 274

s pecific cases 274

urine sa mples
vertebral fra cture 58 - 9

techniqu es 259-62
collection 118- 19
vestibule, vaginal

see also drugs; venoms

tests
congenital ab normalities 4J

toxicity due to do sage errors 366-7

acute metabo lic a utop sy J] 9
normal findings 29-30

tox ins, bac terial, assays 110

dru gs 262, 262 -3, 272
video recordings

t race evidence, road traffic collision

in fatty acid oxidation defects 116
col poscopy 7, 8, 11 , 27

387
USA, rep0l1ing of ab use 42 1
sce ne of sudde n death 138

tracheal ab normal iti es 240

surveillan ce in suspected

tracheal tub es 368

vacuum delivery, intracranial
ab use 332

tra cheitis, bacterial 241

haemorrha ge 174
v iolence, scene o f death from 142

tracheostomy 368
vaginal examination 27
see also physical abuse

tra ctor-related injuries 34 1

norm a l an atomy and vari ants (incl.
vira l infection

transc ranial Doppler, head inju ry

vest ibule/posterior fo urchette)
ante-mortem 110

292 -3
29 - 30
bites 428

tran sp lantation, organ donation for

signs of abuse 39
bra in 177

390
see also vulvovag initis
neo natal, causi ng death 209

tran spo rt-related deaths 337

variant anatomy see a natomica l
virological samples, sudden death

traum a see injury

vari ants
148-9

tren ches, drownings 358

vas cular endothelial growt h factor 109
v isceralorga ns

tryptase, mast cell 109 , 110, 239

vasc ulature
donation for transplantation 390

tub ero us scl eros is, cardi ovascular

ca theter-related deat hs 368
injuries 68-9

pat holo gy 233, 236

electrical injury 327
post-mortem rete ntio n 163

tumours
oc ular, damage in non-accide nta l
road traffic acc ident 389- 90

cardiac 233-5
injury 129
SJDS 215

embol ism 364

vasocon strictio n, refl ex pulmonary
surgical traction on, causi ng ca rdiac

intracranial 23 7
vascular 350
arres t 369

504 I Index

viscera/organs (continued) vulval signs of abuse 39

Williams ' syndrome 227

weigh t (post-mortem) 150

vulvovaginitis 40
winter sports injuries 340-1

fetal 471-2
Wiskott-Aldrich syndrome 81,92,94

neonatal/infant 481-3
Wales, road traffic casualties and witness

visual pathways, non-accidental injury

fatalities 390
expert see expert witness

127-8
warfarin 89, 90
eye, road traffic collision 387

visual sequelae of ocular non­

water
professional 449-50

accidental injury. long-term 125,

hot see scalds
Wolff-Parkinson-White syndrome

134
loss with burns 324
232, 236

vitamin C deficiency and scurvy 71,

submersion in see drowning and
Working together 421

94
near-drowning; immersion
World Trade Center disaster (2001),

vitamin K
waterbeds 331
identifying remains 412

deficiency 98
watersiled damage 172, 311

malabsorption 101
wedging 204, 331
X chromosome 395

vitreous humour 117-18,271

weight assessment
Y chromosome regions homologous

acute metabolic autopsy 120

child
to 400-\

electrolytes 110-11
abuse 13
X-ray see radiography; skeleton, survey

glucose 112, 113, 150, 244

chalts 475-80

haemorrhage into 129

post-mortem 150
Y chromosome 395

vocal cords, SIDS 210

organs see viscera abandoned baby/fetal remains 413

volatile substances (incl. solvents)

placenta 473
paternity testing 411

analysis 260, 262

whiplash injury, road accidents 393
STR analysis 405-6

inhalation/abuse 262,333,341-2
whiplash syndrome (whiplash shaking
sexual abuse 404, 405-6

road traffic fatalities 389

injury) see cervicomedullary X chromosome regions homologous

von Willebrand disease 82, 90-2

syndrome to 400-1

family history 84
white matter
young children see infants; neonates;

neonatal 97
non-traumatic lesions 171-2
toddlers and pre-school children

types 1-3 90-2

shearing injuries 307

von Willebrand factor 79

see also leucodystrophies;

qualitative abnormalities 90, 91

periventricular leucomalacia
quantitative abnormalities (incl.
whole-body radiograph of baby see
deficiency) 88, 90, 90-2
babygram

- .
-~