Proceedings of

The Lancet Conference on Healthcare-Associated Infections

London, UK, 11—12 December 2008
Guest Editor: Professor Kevin G. Kerr
The Official Journal of the Hospital Infection Society
Editor
S. J. Dancer (Glasgow)
Assistant Editors
S. P. Barrett (London)
J. Bates (Worthing)
A. Berrington (Sunderland)
J. Child (Worthing)
R. Cooke (Liverpool)
A. Guyot (Guildford)
P. N. Hoffman (London)
H. Humphreys (Dublin)
T. Inkster (Glasgow)
K. Jacobson (Bristol)
P. Jumaa (Birmingham)
K. G. Kerr (Harrowgate)
G. McDonnell (Basingstoke)
A. J. Mifsud (London)
S. W. B. Newsom (Cambridge)
B. Oppenheim (Birmingham)
S. Schelenz (Norwich)
P. Shears (Wirral)
R. C. Spencer (Bristol)
J. Stephenson (Surrey)
J. Stockley (Worcester)
E. L. Teare (Chelmsford)
A. M. Walker (Pentraeth)
A. P. R. Wilson (London)
Reviews
I. M. Gould (Aberdeen)
Correspondence
G. D. Corcoran (Cork)
Editorial Coordinator
N. Atherton
Editorial Board
P. Astagneau (France)
T. S. J. Elliott (Birmingham)
A. P. Fraise (Birmingham)
C. M. Fry (London)
G. Gopal Rao (London)
K. Gould (Newcastle)
J.-J. Haxhe (Belgium)
A. Holmes (London)
P. P. Mortimer (London)
R. N. Olmsted (USA)
J. Philpott-Howard (London)
G. L. Ridgway (London)
M. Rotter (Austria)
E. T. M. Smyth (Belfast)
Hospital Infection Society
Office
162 Kings Cross Road,
London WC1X 9DH
Tel: 020 7713 0273
Editorial Advisers
Chief Executive
G. Ayliffe (UK) Dr T. R. Logan
F. Daschner (Germany) President
G. L. French (London) Dr G. L. Ridgway
162 King’s Cross Road,
O. B. Jepsen (Denmark) London WC1X 9DH
H. Kobayashi (Japan)
Chairman
S. Mehtar (South Africa) Dr A. P. Fraise
J. W. Pearman (Australia) Department of Clinical
Microbiology,
G. Reybrouck (Belgium) University Hospital Birmingham
H.-G. Sonntag (Germany) NHS Foundation Trust,
Queen Elizabeth Medical
A. Voss (Netherlands) Centre,
Edgbaston,
Birmingham B15 2TJ
Secretary
Ms C. Fry
Department of Health,
Wellington House,
133–155 Waterloo Road,
London SE1 8UG
Scientific Secretary
Dr T. C. Boswell
Department of Microbiology,
Queens Campus,
Nottingham University Hospitals
NHS Trust, Derby Road,
Nottingham NG7 2UH
Treasurer
Dr W. A. Telfer Brunton
Department of Clinical
Microbiology,
Royal Cornwall Hospitals NHS
Trust,
Penventinnie Lane, Treliske,
Truro TR1 3LQ
VOLUME 73 ISSUE 4 DECEMBER 2009

Contents
Introduction
Healthcare-associated infection: moving behind headlines to clinical
solutions
D. Pittet 293
The Hospital Infection Society

Reviews
Historical and changing epidemiology of healthcare-associated infections
A. Pearson 296
The authors, editors, owners and
publishers do not accept any
Role of hand hygiene in healthcare-associated infection prevention
responsibility for any loss or damage
arising from actions or decisions B. Allegranzi and D. Pittet 305
arising from information contained
in this publication; ultimate respon- Preventing surgical site infection. Where now?
sibility for the treatment of patients
H. Humphreys 316
lies with the medical practitioner.
The opinions expressed are those of
the authors and the inclusion in this Intravascular catheter infections
publication of material relating to J. Edgeworth 323
a particular product, method or
technique does not amount to
an endorsement of its value or Noroviruses in healthcare settings: a challenging problem
quality, or of the claims made by its M. Koopmans 331
manufacturers.

Abstracted/Indexed by: Pseudomonas aeruginosa: a formidable and ever-present adversary

Current Contents, ASCA, Science K.G. Kerr and A.M. Snelling 338
Citation Index, Infomed, Index
Medicus, Medline, EMBASE/Excerpta
Extended-spectrum ␤-lactamase-producing organisms
Medica, Abstracts of Hygiene,
Communicable Disease and Tropical M.E. Falagas and D.E. Karageorgopoulos 345
Diseases Bulletin and Cumulative
Index to Nursing and Allied Health Acinetobacter: an old friend, but a new enemy
Literature.
K.J. Towner 355

Community-associated meticillin-resistant Staphylococcus aureus as a

cause of hospital-acquired infections
R.L. Skov and K.S. Jensen 364

Screening and isolation for infection control

E. Tacconelli 371

The role of environmental cleaning in the control of hospital-acquired

infection
S.J. Dancer 378

Controversies in infection: infection control or antibiotic stewardship to

control healthcare-acquired infection?
I.M. Gould 386

Where does infection control fit into a hospital management structure?

E.T. Brannigan, E. Murray and A. Holmes 392
Hand hygiene and infection in hospitals: what do the public know; what
should the public know?
M. Fletcher 397

What are the drivers of the UK media coverage of meticillin-resistant

Staphylococcus aureus, the inter-relationships and relative influences?
T. Boyce, E. Murray and A. Holmes 400

Are national targets the right way to improve infection control practice?
M. Millar 408

Responsibility for managing healthcare-associated infections: where does

the buck stop?
B.I. Duerden 414

Letter to the Editor

Review of Lancet conference on healthcare-associated infections
M. Meda 418
Journal of Hospital Infection (2009) 73, 293e295
Available online at www.sciencedirect.com
INTRODUCTION
Healthcare-associated infection: moving behind
headlines to clinical solutions
D. Pittet*
Infection Control Programme, University of Geneva Hospitals and Faculty of Medicine, 4 Rue Gabrielle-
Perret Gentil, 1211 Geneva 14, Switzerland
Available online 12 October 2009
Introduction
Healthcare-associated infection (HCAI) is universal
and complicates patient care both in developed
and developing countries. Worldwide, it is esti-
mated that as many as 1.4 million patients acquire
infections each day in hospitals alone.1 Recognis-
ing that HCAI prevention is of paramount impor-
tance on the current patient safety agenda, The
Lancet organised an international conference on
HCAI in London in December 2008.
What did we learn?
Our old friendsdmeticillin-resistant Staphylococ-
cus aureus (MRSA), vancomycin-resistant
enterococci (VRE), Pseudomonas aeruginosa,
Acinetobacter spp., and Clostridium difficiled
continue to challenge healthcare settings world-
wide. An MRSA pandemic is ongoing and strains
are capable of acquiring new virulence factors
in addition to multiple resistance. Rapid diagnosis
seems useful, but only in conjunction with
preventive actions. The screening of target popu-
lations deserves further investigation, whereas
universal screening appears to be neither effec-
tive nor cost-effective. The role of antibiotic
* Tel.: þ41 22 372 9828; fax: þ41 22 372 3987.
E-mail address: didier.pittet@hcuge.ch
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.08.004
www.elsevierhealth.com/journals/jhin
pressure on the MRSA reservoir deserves con-
trolled interventions. Importantly, MRSA control
is feasible and many examples are available at
hospital, regional, or even national levels.
VRE spread continues to represent a paradox
on both sides of the Atlantic Ocean. VRE is
a coloniser of the gut like Gram-negatives, of
the skin like MRSA, and of the environment like
C. difficile. Thus, VRE control mandates a com-
plete, multimodal infection prevention strategy,
but is clearly feasible. P. aeruginosa remains
a key pathogen in hospitals and causes multiple
infections. However, the jury is still out on the
role of tap-to-patient and patient-to-tap cross-
transmission as the reservoir/source. Should
P. aeruginosa be considered now as an occupa-
tional hazard? Acinetobacter spp. are acquiring
pan-resistance to antimicrobials and causing
outbreaks, particularly among specific patient
populations. In particular, their capacity for
long-term survival in the environment compli-
cates control. Should we propose a zero toler-
ance for P. aeruginosa and Acinetobacter spp.?
Extended-spectrum b-lactamase-producing en-
terobacteriaceae (ESBLs) are now becoming
universal with multiple, complex resistance
mechanisms that differ worldwide, and studies
are needed to better delineate their epidemiol-
ogy and control. The emergence of a new C. dif-
ficile strain over the past years has highlighted
major failures in patient safety. Large
294
differences in the disease epidemiology and con-
trol are observed among hospitals and countries,
but control is feasibledeven at country level.
Our old friends are going public: community
pathogens are now entering the hospital. Examples
include ESBLs and community-acquired MRSA.
Norovirus outbreaks originating in the community
are affecting institutionalised patient populations,
sometimes with devastating consequences. And
finally, C. difficile is now responsible for severe
community infections among young adults and
non-immunocompromised populations in contrast
to its more traditional role previously restricted
to healthcare settings.
Are there clinical solutions?
Targeted HCAI prevention is clearly effective.
Prevent intravascular line infection? e ‘Yes, we
can!’ How? By using multimodal intervention strat-
egies, sometimes also called ‘bundles’. Prevent
surgical site infection? e ‘Yes, we possibly can.’
Again, by decreasing several risk factors, in
particular, by the optimal and timely use of
perioperative antibiotic prophylaxis. Preventing
ventilator-associated pneumonia remains difficult
and complex, but is certainly possible using mul-
timodal interventions whose components still need
to be defined. In particular, there is a pressing
need to revisit the content of the currently
proposed bundle strategies.
The critical role of setting targets for practice
improvement needs to be reviewed, especially at
national level. First, healthcare systems should
decide what and how to measure. Other ques-
tions that follow are: what are the best in-
terventions to implement? What are the targets
for improvement? Targets should be set locally
considering feasibility and cost-effectiveness.
Risk taking is an integral part of healthcare,
similar to leadership. In the UK, mandatory
surveillance, public reporting, and the setting of
performance targets have clearly provided effec-
tive leverage.
Towards integrated solutions
Hand hygiene remains the key measure for HCAI
prevention. As illustrated by the successful imple-
mentation of the ‘‘Clean Care is Safer Care’’ multi-
modal strategy proposed by the World Health
Organization First Global Patient Safety Challenge
and launched in 2005, promotion is feasible
and effective worldwide (http://www.who.int/
gpsc/en).2 Although environmental control must
improve in most healthcare settings with special
D. Pittet
emphasis on hand-touch sites, it is not the sole
responsibility of the infection control team.
Cleanliness is everyone’s business, from
housekeeping to executive staff and board members.
There is a clear link between antimicrobial use
and resistance at all levels; ward, hospital, re-
gional, country, and even continental. As shown
recently through successful nationwide campaigns,
the impact of antibiotic control is easier to
demonstrate in the community. Infection control
teams in hospitals have traditionally worked as
fire-fighters and now need to move on to more
prevention-oriented stewardship. When well-
organised and effective, antimicrobial stewardship
is largely cost-effective.
Infection control must be integrated into the
hospital management structure at all levels, in-
cluding the decision-making process. It should use
a system-based approach to ensure sustainability,
be viewed as a corporate priority, and be based on
corporate accountability and institutional leader-
ship. Performance management tools related to
infection control metrics are available and HCAI
rates can be used as a proxy indicator for differ-
ences in system management. The best healthcare
model needs to deliver effective, reliable and
resilient infection prevention.
Should we go public?
HCAI is a topic of high public and media interest.
However, this increased openness and transpar-
ency could also result in a rising concern about the
infection risk. As a consequence of the ‘public
right to know’, public reporting and performance
benchmarking is increasingly well-established in
several countries. But does it make a difference to
the HCAI risk in hospitals and other healthcare
settings?
Martin Fletcher reviews the evidence on
publication reporting and proposes actions for
boards and senior leaders to improve the
institutional safety culture.3 He suggests hand
hygiene promotion as an exemplar model for in-
creasing patient safety at institutional, regional,
or even national level, and the necessity to
allow patients to participate in this continuous
and complex challenge. It follows, therefore,
that communication at all levels must be opti-
mised in the field of infection prevention; it
must consider the public right to know,
allow patients and families to participate,
and work in synergy with the media in multidis-
ciplinary forums that include patients, health-
care workers, scientists, journalists and
politicians.
Healthcare-associated infection
Infection control is moving behind the headlines
to guarantee safer healthcare by applying effec-
tive clinical solutions and by conducting research
where needed. Infection control is no longer
optional. For the sake of every patient in our
care, the healthcare community across the globe
must make every effort to ensure that it is fully
integrated into each institution’s management
structure.
295
References
1. Lynch P, Pittet D, Borg MA, Mehtar S. Infection control in
countries with limited resources. J Hosp Infect 2007;
65(Suppl. 2):148e150.
2. Pittet D, Donaldson L. Clean Care is Safer Care: a worldwide
priority. Lancet 2006;366:1246e1247.
3. Fletcher M. Hand hygiene and infection in hospitals: what do
the public know; what should the public know? J Hosp Infect
2009;73: 397e399.
Journal of Hospital Infection (2009) 73, 296e304
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Historical and changing epidemiology of

healthcare-associated infections
A. Pearson a,b,*
a
Department of HCAI & AMR, Centre for Infections, London, UK
b
Department of Advanced Computational Biology, University of Maryland, College Park, Maryland, USA

Available online 3 November 2009

KEYWORDS Summary This review compares the historical perspectives on health-

Clostridium difficile; care-associated infections (HCAIs) with the current changing epidemio-
Healthcare-associated logical picture as it relates to these infections. Evidence in support of
infection;
these changes is given using trends in mortality for Clostridium difficile
Medical history;
MRSA;
and meticillin-resistant Staphylococcus aureus bacteraemia in England as
Surveillance examples. The impact of current intervention programmes which target
these, and other HCAIs, is also considered and knowledge gaps and options
for changes in public health strategy required to achieve further reductions
in HCAIs in the National Health Service in England are identified.
ª 2009 Published by Elsevier Ltd on behalf of The Hospital Infection Society.

Historical perspectives strictly attributable to him. Subsequently the pro-

In the fourth century BC, Hippocrates, the father of
western medicine, is generally attributed as
having introduced the need to ‘use treatments
for the benefit of the ill in accordance with my
ability and my judgment, but from what is to their
harm and injustice I will keep them.’ This is one of
a series of statements that were subsequently
referred to as the Hippocratic Oath. The oath
was an ancient protocol in medicine at the time
of Hippocrates although it may not have been
* Corresponding address: Department of HCAI & AMR, Centre
for Infections, 61 Colindale Avenue, London NW9 5EQ, UK.
E-mail address: dandrewpearson@googlemail.com
tocol has been adopted universally by medical
schools and accreditation bodies including the
General Medical Council in the UK.1
By the twenty-first century AD the Institute for
Health Improvement in its ‘5 Million Lives’ cam-
paign had enhanced the definition of medical
harm as:
Unintended physical injury resulting from or contrib-
uted to by medical care, including the absence of
indicated medical treatment, that requires additional
monitoring, treatment or hospitalisation, or that
results in death. Such injury is considered harm
whether or not it is considered preventable, whether
or not it resulted from a medical error or it occurred
within a hospital.2

0195-6701/$ - see front matter ª 2009 Published by Elsevier Ltd on behalf of The Hospital Infection Society.
doi:10.1016/j.jhin.2009.08.016
Historical and changing epidemiology of healthcare-associated infections
Through the following centuries the Egyptians
and Europeans clearly enunciated their interest in
medical oversight and culpable negligence. In
Egypt, Moses ben Maimon practised and taught
medicine insisting on cleanliness as the physician’s
best friend: ‘Never forget to wash your hands after
having touched a sick person’ and ‘I dismount from
my animal, wash my hands, go forth to my
patients.’
In Europe by the nineteenth century several
physicians developed further the realisation of the
importance of hand washing. Dr Ignaz Semmelweis,
the Hungarian-born physician, generally regarded
as the father of infection control practice, was
responsible in 1847 for the maternity service of the
Allgemeine Krankenhaus teaching hospital in
Vienna. There he observed that women delivered
by physicians and medical students had a much
higher rate of post-delivery mortality (13e18%) than
women delivered by midwives or trainee midwives
(2%). Earlier, less well-evidenced, statements are
attributed to Oliver Wendell Holmes:
Let the men who mould opinions here look to it; if
there is any interested oversight, any culpable neg-
ligence and the facts shall reach the public ear, the
pestilence-carrier of the lying-in chamber must look
to God for pardon, for man will never forgive him.
The work of Florence Nightingale in October
1854 and April 1855 made a meticulous analysis of
the mortality data and identified poor sanitation as
the underlying cause as well as creating new
statistical diagrams that persuaded government
to reform health. This development of a modern-
day analytical approach to patient outcome was
the evidence base for her Notes on hospitals. This
had a profound impact on the design and manage-
ment of hospitals as it tackled problems of over-
crowding in civil hospitals, poor ventilation and
lack of cleanliness and reported deaths from pre-
ventable diseases, deaths from wounds and deaths
from all other causes.
HCAI in the twenty-first century
The contribution of deaths from HCAI to the total
lives lost per year has been estimated by Amalberti
who compared healthcare with ultrasafe activities
such as flying with scheduled airlines, the nuclear
power industry, road traffic deaths, and with
dangerous activities such as bungee jumping and
mountain climbing.2
The risk from healthcare is further exemplified
by reference to the published mortality reports for
C. difficile and meticillin-resistant Staphylococcus
297
aureus (MRSA). Figures 1 and 2 respectively show
the UK trends for increasing numbers of death
certificates reporting as a cause of death or men-
tioning C. difficile (1993e2007) and MRSA (1999e
2007). These trends are subject to ascertainment
bias as clinicians may change their threshold for
notifying deaths due to HCAI.
National surveillance and evidence for
the changing epidemiology of C. difficile
infection
Case surveillance based on laboratory reporting is
accepted as the most reliable routinely available
information by which to measure changes in
incidence. In England there are two surveillance
systems for measuring S. aureus (including MRSA)
and C. difficile infection: voluntary laboratory
reporting (LabBase electronic reporting system)
and the Department of Health-initiated mandatory
MRSA bacteraemia and C. difficile surveillance
programmes. The increase in laboratory-confirmed
cases of C. difficile infection (CDI) reported to the
voluntary national laboratory reporting system
between 1990 and 2007 is shown in Figure 3. These
English National Health Service (NHS)-based lab-
oratories receive specimens from both hospital
and community patients across the whole NHS.
Ascertainment during this time varied from around
45% in 1990 to 70% in 2007. Notwithstanding this
increased reporting there was a steady increase
from 1990 to 2001, after which there was an accel-
erating rate of increase recognised in retrospect as
related to the emergence of newly recognised hy-
per-toxin-producing strains of C. difficile. Figure 3
also indicates the timelines for the introduction of
mandatory surveillance of laboratory-confirmed
cases of CDI, mandatory strain surveillance and
mandatory enhanced surveillance to monitor a per-
formance target to reduce CDI. In addition the UK
Treasury’s Public Service Delivery Agreement, pub-
lished in 2007, introduced a target for a 30% reduc-
tion in the number of cases of CDI reported by
English NHS hospitals in 2010e11 compared with
an agreed baseline in 2007e8.
The impact of this mandatory surveillance pro-
gramme and the setting of a 30% reduction target
is seen in Figure 4. Data from the voluntary surveil-
lance shows ascertainment for patients 65 years
as 77e85% of the mandatory scheme for each quar-
ter between April 2007 and March 2008. So manda-
tory surveillance, public reporting and the setting
of a performance target was highly effective at in-
creasing ascertainment. This intervention might
298 A. Pearson

9000
Underlying cause Mentions
8000

7000

6000
No. of deaths

5000

4000

3000

2000

1000

0
1999 2000 2001 2002 2003 2004 2005 2006 2007
Year

Figure 1 Number of deaths reported with C. difficile as underlying cause or mentioned 1993e2007. Data from the
UK Office of National Statistics.

itself be expected to influence physicians’ behav- guide physicians and clinical teams in the preven-
iour. During 2006 and 2007 the Department of tion of CDI.3 The overall impact of introducing
Health (DoH) launched the Saving Lives campaign mandatory surveillance, the Saving Lives campaign
with a High Impact Intervention care bundle to and the follow-up of NHS trusts against the

1800

Underlying cause Mentions

1600

1400

1200
No. of deaths

1000

800

600

400

200

0
1993 1995 1997 1999 2001 2003 2005 2007
Year

Figure 2 Number of deaths reported with MRSA as underlying cause or mentioned 1999e2007. Data from the UK
Office of National Statistics.
Historical and changing epidemiology of healthcare-associated infections 299

45 000

40 000

35 000

30 000
C. difficile counts

25 000

20 000

15 000

10 000

5000

0
90

91

92

93

94

95

96

97

98

99

00

01

02

03

04

05

06

07

08
19

19

19

19

19

19

19

19

19

19

20

20

20

20

20

20

20

20

20
Year

Figure 3 Voluntary NHS laboratory C. difficile surveillance. Timeline for introduction of mandatory surveillance:
2004, mandatory surveillance introduced; 2005, mandatory strain sampling introduced; 2007, mandatory enhanced
surveillance introduced. Source: Health Protection Agency LabBase Voluntary Reports; Department of Health Manda-
tory Surveillance Reports.

performance target by a DoH-based improvement National surveillance and evidence for

team has had a remarkable effect within the first
full year of the programme. Figure 5 compares
the pattern of changes in rates of CDI per 1000
bed-days. Marked variation across the English
NHS underlines both the variability and opportu-
nity for further health gain that exists across the
NHS.
The enhanced surveillance programme has also
begun to provide an insight as to the incidence of
CDI in patients aged <65 years as well as in
the general population diagnosed with CDI in
non-acute care settings. Measurement and the
interpretation of enhanced surveillance data
from the mandatory scheme is complex but initial
analysis indicates that an average of 29% of CDI
cases would fit a definition of community-acquired
infection based on three categories of specimen:
on-presentation cases, cases developing less than
2 days after admission, and non-acute specimens.
Aggregate analysis may hide important local vari-
ation. This is the case with CDI in the community.
The proportion of community CDI cases in local
NHS trusts varies markedly from 0 to >55% in
a small number of localities. These findings are
newly recognised and under further investigation
by both epidemiologists and local microbiologists.
Early indications are that, as in the past, a pro-
portion of CDI cases do not have either exposure to
antibiotics or healthcare as risk factors for their
infection.
the reduction in MRSA bacteraemia
The laboratory-based voluntary system showed
a steep rise in the reported number of MRSA bac-
teraemia cases between 1990 and 2001 (Figure 6).
In 2001, the Chief Medical Officer noted that ‘the
prevalence of methicillin-resistant S. aureus has
increased markedly in the last decade, primarily
associated with hospital acquired infection.’4 Man-
datory reporting of MRSA and meticillin-suscep-
tible S. aureus (MSSA) bacteraemia was
introduced for all NHS acute trusts in England in
April 2001. Enhanced mandatory surveillance was
initiated in 2005 which allows cases to be entered
in real time and collects more detailed information
regarding the incidence and risks of infection.
The voluntary reporting system continues to
operate alongside the mandatory surveillance
scheme. Voluntary surveillance will underestimate
the true number of MRSA bacteraemia cases (data
not shown). Ascertainment is affected by the
number of NHS trusts contributing voluntary data
and the completeness of these data, both of which
are likely to have varied over time. In total, 30 946
reports of MRSA bacteraemia were made via
LabBase between 2002 and 2008. The introduction
of mandatory reporting to the web increased this
number to 44 344. Mandatory reporting had the
greatest impact in the first year when reporting
increased by 49% from 4879 to 7274 cases.
 300

18 000

16 000

14 000

12 000

10 000

8000

C. difficile counts
6000

4000

2000

Jul-Sep
Jul-Sep
Jul-Sep
Jul-Sep
Jul-Sep
Jul-Sep
Jul-Sep
Jul-Sep

Apr-Jun
Apr-Jun
Apr-Jun
Apr-Jun
Apr-Jun
Apr-Jun
Apr-Jun
Apr-Jun
Apr-Jun

Jan-Mar
Jan-Mar
Jan-Mar
Jan-Mar
Jan-Mar
Jan-Mar
Jan-Mar
Jan-Mar
Jan-Mar

Oct-Dec
Oct-Dec
Oct-Dec
Oct-Dec
Oct-Dec
Oct-Dec
Oct-Dec
Oct-Dec

2000 2001 2002 2003 2004 2005 2006 2007 2008

Quarter

Figure 4 Comparison of Mandatory and Voluntary NHS laboratory C. difficile surveillance. Solid black line: mandatory 65 years; solid grey line: voluntary 65
years; dashed line: mandatory 2e64 years; dotted line: voluntary 2e64 years. Source: Health Protection Agency LabBase Voluntary Surveillance Reports.
A. Pearson
 3.50

3.00

2.50

2.00

1.50

C. difficile rate per 1000 bed-days

1.00

0.50
Historical and changing epidemiology of healthcare-associated infections

0
2004 - 2007 2004 - 2007 2004 - 2007 2004 - 2007 2004 - 2007 2004 - 2007 2004 - 2007 2004 - 2007 2004 - 2007
and nds n ast est ast st s ber
ndo hE hE We nd
E ngl i dla Lo r t r t hW u t u t h i dla Hum
f st M No No So So st M &
st o Ea We re
Ea k shi
r
Yo
HPA Region
301

Figure 5 Regional C. difficile rate per 1000 bed-days. HPA, Health Protection Agency. Source: Department of Health Mandatory MRSA bacteraemia surveillance.
302 A. Pearson

16 000

14 000

12 000
MRSA
S. aureus bacteraemia counts

10 000 MSSA

No susceptibility
8000

6000

4000

2000

0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Year

Figure 6 Staphylococcus aureus bacteraemia voluntary laboratory reporting 1990e2006: meticillin sensitivity
reported by English NHS acute trusts.

Comparison of the reported numbers to 2008
showed that mandatory reporting has increased
ascertainment by between approximately 900 and
2300 cases per year to 39%.
The impact of introducing mandatory surveil-
lance, the Saving Lives campaign and the use of
improvement teams by the DoH along with a na-
tional ‘cleanyourhands’ campaign has effected the
required national reduction of 50% by June 2008
(Figure 7). Again the introduction of mandatory
enhanced surveillance of MRSA bacteraemia cases
gave a new insight to the epidemiology as it dem-
onstrated that only 65% of cases were acquired
during the current hospital admission whereas
34% were reported from patients who were recent
admissions from the community or non-acute care
facilities. Voluntary submitted risk factor data on
3495/12 483 (28%) cases reported to the mandatory
surveillance system between May 2006 and Decem-
ber 2008 indicated key information on the sources
of MRSA bacteraemia and highlighted the impor-
tance of vascular lines, skin and soft tissue infec-
tions as well as other localised infections and
urinary tract infections as potentially preventable
sources of infection. This is invaluable information
on which to base local strategies to further reduce
MRSA bacteraemia.
So, with the introduction of mandatory surveil-
lance, HCAI has been made a visible and unambig-
uous indicator of quality and safety of patient care
and the Health Protection Agency has provided the
NHS with high quality information for the public,
patients and clinical teams so that the risks
associated with the performance of certain pro-
cedures are transparent.
Worldwide data on measurement of HCAI is
remarkable in its inadequacy. An editorial from
leading US senators compared measurement in
healthcare and baseball and cited that:
Remarkably, a doctor today can get more data on the
starting third baseman on his baseball team than on the
effectiveness of life-and-death medical procedures.
Studies have shown that most health care is not based
on clinical studies of what works best. Instead, most
care is based on informed opinion, personal observa-
tion or tradition .
To deliver better health care, we should learn from the
successful teams that have adopted baseball’s new
evidence-based methods. The best way to start im-
proving quality and lowering costs is to study the stats .
In the past decade, baseball has experienced a data-
driven information revolution. Numbers-crunchers now
routinely use statistics to put better teams on the field
for less money. Our overpriced, underperforming
health care system needs a similar revolution .
To deliver better health care, we should learn from
the successful teams that have adopted baseball’s
Historical and changing epidemiology of healthcare-associated infections 303

2000

1500
MRSA bacteraemias reported

1000

500

0
Apr-Jun
Jul-Sep
Oct-Dec

Apr-Jun
Jul-Sep
Oct-Dec

Apr-Jun
Jul-Sep
Oct-Dec

Apr-Jun
Jul-Sep
Oct-Dec

Apr-Jun
Jul-Sep
Oct-Dec

Apr-Jun
Jul-Sep
Oct-Dec

Apr-Jun
Jul-Sep
Oct-Dec

Apr-Jun
Jan-Mar

Jan-Mar

Jan-Mar

Jan-Mar

Jan-Mar

Jan-Mar

Jan-Mar
2001 2002 2003 2004 2005 2006 2007 2008
Quarter

Figure 7 Reduction of mandatory MRSA bacteraemia surveillance 2005e2008. Sources: NHS Mandatory and Volun-
tary MRSA bacteraemia surveillance.

new evidence-based methods. The best way to start tracking will assist infection control teams in pre-
improving quality and lowering costs is to study the venting as well as controlling clusters of infection
stats.5 in hospitals at the local level but rapid strain char-
acterisation will be invaluable in recognising
The introduction into the English NHS of a web-
changes not only in local but also in regional and
enabled reporting and user access networked
national epidemiology. Similarly, real-time elec-
mandatory surveillance system has addressed this
tronic prescribing systems will allow enhanced an-
deficiency in measurement for a limited number
tibiotic audit and better analysis of antibiotic mis-
of infections currently targeted by mandatory
prescribing which, in turn, will help in efforts to
surveillance and performance management.
control CDI and antimicrobial resistance.
Selective use of improved evidence-based
Knowledge gaps performance management interventions such as
reduction targets, ‘never events’ and ‘zero toler-
Encouraging though these trends are, current gaps ance’ will also be beneficial. Finally development
in our knowledge relating to healthcare-associated and refinement of tools which permit assessment
infection need to be addressed if further reduc- of new interventions to identify how they can be
tions in these infections are to be achieved. These deployed in a cost-effective manner remains a key
include introduction of robust surveillance systems priority.
to monitor infections other than those related to
MRSA bacteraemia and C. difficile. This should
include systematic population-based epidemiolog-
ical surveillance. Technologies which permit real- Conflict of interest statement
time strain characterisation and real-time patient None declared.
304 A. Pearson

Funding sources 2. Amalberti R. The paradoxes of almost totally safe transpor-

None. tation systems. Saf Sci 2001;37:109e126.
3. Department of Health. Saving Lives: a delivery programme
to reduce healthcare associated infection including MRSA.
London: DoH; 2006.
References 4. Department of Health. CMO’s update 30: surveillance of
healthcare associated infections. London: DoH; 2001.
1. General Medical Council. Good medical practice: the duties 5. Beane W, Gingrich N, Kerry J. How to take American health
of a doctor registered with the General Medical Council. care from worst to first. New York Times Online, 24 October
London: GMC; 2006. 2008.
Journal of Hospital Infection (2009) 73, 305e315
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Role of hand hygiene in healthcare-associated

infection prevention
B. Allegranzi a,*, D. Pittet a,b
a
World Alliance for Patient Safety, World Health Organization, Geneva, Switzerland
b
Infection Control Programme, University of Geneva Hospitals and Faculty of Medicine,
Geneva, Switzerland

Available online 31 August 2009

KEYWORDS
Alcohol-based hand
rub; Hand hygiene;
Healthcare-associated
infection;
Intervention; Patient
safety; Promotion;
World Health
Organization
Summary Healthcare workers’ hands are the most common vehicle for
the transmission of healthcare-associated pathogens from patient to
patient and within the healthcare environment. Hand hygiene is the leading
measure for preventing the spread of antimicrobial resistance
and reducing healthcare-associated infections (HCAIs), but healthcare
worker compliance with optimal practices remains low in most settings.
This paper reviews factors influencing hand hygiene compliance, the
impact of hand hygiene promotion on healthcare-associated pathogen
cross-transmission and infection rates, and challenging issues related to
the universal adoption of alcohol-based hand rub as a critical system
change for successful promotion. Available evidence highlights the fact
that multimodal intervention strategies lead to improved hand hygiene
and a reduction in HCAI. However, further research is needed to evaluate
the relative efficacy of each strategy component and to identify the most
successful interventions, particularly in settings with limited resources.
The main objective of the First Global Patient Safety Challenge, launched
by the World Health Organization (WHO), is to achieve an improvement in
hand hygiene practices worldwide with the ultimate goal of promoting
a strong patient safety culture. We also report considerations and
solutions resulting from the implementation of the multimodal strategy
proposed in the WHO Guidelines on Hand Hygiene in Health Care.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

* Corresponding author. Address: First Global Patient Safety Challenge, World Alliance for Patient Safety, IER/PSP, Room L319,
L Building, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland. Tel.: þ41 22 791 2689; fax: þ41 22 791 1388.
E-mail address: allegranzib@who.int

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.04.019
306
Introduction
Numerous studies document the pivotal role of
healthcare workers’ (HCWs) hands in the propaga-
tion of micro-organisms within the healthcare
environment and ultimately to patients.1 As
recently described, patient-to-patient transmis-
sion of pathogens via HCWs’ hands involves five
sequential steps.2 Patients’ skin can be colonised
by transient pathogens that are subsequently
shed onto surfaces in the immediate patient
surroundings, thus leading to environmental
contamination.2 As a consequence, HCWs contam-
inate their hands by touching the environment or
patients’ skin during routine care activities,
sometimes even despite glove use.2 It has been
shown that organisms are capable of surviving
on HCWs’ hands for at least several minutes fol-
lowing contamination.2 Thus, if hand hygiene
practices are suboptimal, microbial colonisation
is more easily established and/or direct transmis-
sion to patients or a fomite in direct contact with
the patient may occur.2
Based on this evidence and the demonstration
of its effectiveness, optimal hand hygiene behaviour
is considered the cornerstone of healthcare-
associated infection (HCAI) prevention.2e4 Fur-
thermore, not only is it a key element of standard
and isolation precautions, but its importance is
emphasised also in the most modern ‘bundle’
approaches for the prevention of specific site in-
fections such as catheter-related bloodstream
infection (CRBSI), catheter-related urinary tract
infection (CRUTI), surgical site infection (SSI),
and ventilator-associated pneumonia (VAP).5e9
Together with other specific prevention
measures, environmental cleaning is another
essential measure to prevent the spread of
some pathogens, particularly Clostridium diffi-
cile, vancomycin-resistant enterococci (VRE),
norovirus, Acinetobacter spp. and meticillin-
resistant Staphylococcus aureus (MRSA), and
should not be neglected.10e12
Over the past few years, scientific evidence to
support the role of hand hygiene in the improve-
ment of patient safety has increased consider-
ably, but some key controversial issues still
challenge care practitioners and researchers.
This review summarises the key themes on
the role of hand hygiene in preventing HCAI.
Interpretations and solutions based on the
evidence and experience available through the
work of the First Global Patient Safety Challenge
of the WHO World Alliance for Patient Safety are
suggested.
B. Allegranzi, D. Pittet
Factors influencing hand hygiene
compliance
It has been known for many years that HCWs
encounter difficulties in complying with hand
hygiene indications at different levels.4 Insuffi-
cient or very low compliance rates have been
reported from both developed and developing
countries.1,4 Reasons which explain suboptimal
practices are multiple and may vary according to
the setting and the resources available. For exam-
ple, the lack of appropriate infrastructure and
equipment to enable hand hygiene performance,
the cultural background, and even religious beliefs
can play an important role in hindering good prac-
tices.13e15 The most frequently observed factors
determining poor hand hygiene compliance are:
(i) belonging to a certain professional category
(i.e. doctor, nursing assistant, physiotherapist,
technician); (ii) working in specific care areas
(i.e. intensive care, surgery, anaesthesiology,
emergency medicine); (iii) understaffing and over-
crowding; and (iv) wearing gowns and/or gloves.1
Unfortunately, hand hygiene indications at higher
risk of being neglected are the ones that prevent
pathogen transmission to the patient (i.e. before
patient contact and clean/aseptic procedures).1
This is also in concordance with the fact that
care activities with a higher risk of cross-transmis-
sion lead to a higher risk of poor compliance.1
Individual factors such as social cognitive de-
terminants may provide additional insight into
hand hygiene behaviour.3,16e18 Many factors play
a role in eventually determining either a hand
hygiene action or lack of compliance: perception
and knowledge of the transmission risk and of the im-
pact of HCAI; social pressure; HCWs’ conviction of
their self-efficacy; the evaluation of perceived ben-
efits against the existing barriers; the intention to
perform the hand hygiene action. For instance, in-
tention to wash hands did not predict observed hand-
washing behaviour in one study, whereas it did in
another.19,20 Hence, hand hygiene behaviour ap-
pears not to be homogeneous and can be classified
into at least two types of practice.21 Inherent hand
hygiene practice, which drives most community
and HCW hand hygiene actions, occurs when hands
are visibly soiled, sticky or gritty. On the other
hand, elective hand hygiene practice represents
those opportunities for hand cleansing not encom-
passed in the inherent category. Among HCWs, this
component of hand hygiene behaviour is similar to
many common social interactions, such as shaking
hands. During healthcare, it would include touching
a patient (e.g. taking a pulse or blood pressure) or
Hand hygiene and HCAI prevention
having contact with an inanimate object in the pa-
tient’s surroundings. As they recall a common social
behaviour, these contacts do not necessarily trigger
an intrinsic need to cleanse hands, although they
do involve the risk of cross-transmission. According
to behavioural theories, this is the component of
hand hygiene most likely to be omitted by busy
HCWs and it has been repeatedly confirmed by field
observations.
Impact of hand hygiene promotion on
HCAI
Given the complexity of hand hygiene behaviour
and the influence of numerous external factors,
promotion of good practices is complex and its
potential for success depends on the delicate
balance between evaluation of benefits and exis-
tent barriers. Demonstration of the effectiveness of
recommendations and strategies to improve hand
hygiene on the ultimate outcome, i.e. the HCAI
rate, is crucial in both motivating HCWs’ behav-
ioural change and securing an investment in this
preventive measure by policy-makers and health-
care managers. However, research in this field
represents a very challenging activity since meth-
odological and ethical concerns make it difficult to
conduct randomised controlled trials with appro-
priate sample sizes that could establish the relative
importance of hand hygiene in the prevention of
HCAI. In addition, HCAI surveillance is a very
resource- and time-consuming activity requiring
rigorous and standardised methods, and therefore
is seldom available on a regular and reliable basis.
Nevertheless, there is convincing evidence that
improved hand hygiene can reduce infection rates.
More than 20 hospital-based studies of the impact of
hand hygiene on the risk of HCAI have been pub-
lished between 1977 and 2008 (Table I).22e45 Of
these, some were conducted hospital-wide and
report long-term follow-up to demonstrate sustain-
ability.29,30,38,42 Despite study limitations, almost
all reports showed a temporal association between
improved hand hygiene practices and reduced
infection and cross-transmission rates. Most investi-
gations were conducted in adult or neonatal inten-
sive care units (ICUs) and the large majority
introduced the use of alcohol-based hand rubs in
association with other promotional components in
a multimodal implementation strategy (Table I).
Three studies failed to show HCAI reduction follow-
ing hand hygiene promotion.24,41,44 In one study,
the intervention did not succeed in significantly in-
creasing hand hygiene compliance.24 In another,
the methods and definitions used to detect HCAI
were not described and therefore the data
307
reliability cannot be assessed.44 In a prospective,
controlled, cross-over trial, Rupp and colleagues ob-
served no substantial change in device-associated
infection rates and infections due to multidrug-
resistant pathogens, despite a significant and sus-
tained improvement in hand hygiene adherence.41
Nevertheless, although the study was well designed,
it was criticised for lack of screening for cross-trans-
mission, lack of statistical power, and use of an
alcohol-based hand rub that failed to meet the EN
1500 standards for antimicrobial efficacy.46e48
In many countries, the evidence from studies on
hand hygiene effectiveness has been convincing
enough to motivate governments to invest re-
sources in hand hygiene national and subnational
campaigns.49 However, this evidence mainly
reflects findings from interventions implemented
in healthcare settings in developed countries.
Further research is needed to evaluate the relative
efficacy of each key element of multimodal strat-
egies, to assess their implementation feasibility
in settings with limited resources, and to gather
information on successful solutions allowing adap-
tation. Among its main objectives, the First Global
Patient Safety Challenge, launched by the WHO
World Alliance for Patient Safety, intends to
make available implementation tools for field use
and to assess their validation and adoption in
countries at different income levels.49
Another controversial issue is how significant
should be the hand hygiene compliance increase
following the intervention in order to be considered
satisfactory. No data are available yet to answer
this question. Among all the above-mentioned
studies, increased compliance rates at follow-up
did not exceed 81% (Table I). One study with a fol-
low-up of eight years showed a sustained compli-
ance increase of up to a maximum of 66% and
succeeded in parallel to maintain the achieved
reduction in HCAI rates of <10%.29,30 To achieve
100% compliance is not strictly necessary to deter-
mine improvement of patient safety at the bedside.
On the other hand, the goal of sustained 100% com-
pliance appears unlikely to be achieved because of
the complex range of factors influencing HCWs’ be-
haviour related to hand hygiene performance.
Thus, there is a need for careful consideration be-
fore setting a goal of zero tolerance to hand hygiene
non-compliance to avoid failure and frustration.
Challenging issues related to the
adoption of alcohol-based hand rubs
The adoption of alcohol-based hand rubs is con-
sidered the gold standard for hand hygiene in most
Table I Most relevant studies assessing the impact of hand hygiene promotion on HCAI (1977e2008)
308

Year Hospital Intervention Impact on hand Impact on HCAI Duration of Reference

setting hygiene compliance follow-up
1977 Adult ICU Promotion of hand washing NA Significant reduction (P < 0.001) in 2 years 22
with a chlorhexidine hand the percentage of patients
cleanser colonised/infected by Klebsiella
spp.
1989 Adult ICU Education on hand Compliance increase from 14% to Significant reduction (P ¼ 0.02) in 6 years 23
washing, hand hygiene 73% (before patient contact) and HCAI rates (from 33% to 12% and
observation, performance from 28% to 81% (after patient from 33% to 10%, respectively,
feedback contact) after two intervention periods 4
years apart)
1990 Adult ICU Hand-washing promotion Compliance increase from 22% to No impact on HCAI rates 11 months 24
29.9%
1992 Adult ICUs Prospective multiple cross- NA Significant reduction (P < 0.02) in 8 months 25
over trial on hand hygiene HCAI rates using hand washing
with either chlorhexidine with chlorhexidine soap
soap or 60% isopropyl
alcohol with optional hand
washing with plain soap
1994 NICU Introduction of hand NA Elimination of MRSA, when 9 months 26
washing with triclosan 1% combined with multiple other
(w/v) infection control measures.
Significant reduction (P < 0.02) in
nosocomial bacteraemia (from
2.6% to 1.1%) using triclosan
compared with chlorhexidine for
hand washing
1995 Newborn Introduction of HCWs’ NA Control of MRSA outbreak 3.5 years 27
nursery hand washing and
neonates’ bathing with
triclosan 0.3% (w/v)
2000 MICU/NICU Organisational climate NA Significant (85%) relative 8 months 28
intervention reduction (P ¼ 0.02) in VRE rate in
the intervention hospital;
statistically not significant (44%)
relative reduction in control
hospital; no significant change in
MRSA
B. Allegranzi, D. Pittet
2000 Hospital-wide Alcohol-based hand rub Significant increase in Significant reduction (P ¼ 0.04 and 8 years 29,30
introduction, hand compliance from 48% to 66% P < 0.001) in the annual overall
hygiene observation, HCAI prevalence (42%) and MRSA
training, performance cross-transmission rates (87%).
feedback, posters Active surveillance cultures and
contact precautions implemented
during same period. A follow-up
study showed continuous
increase in hand rub use,
stable HCAI rates and cost
savings.
2003 Orthopaedic Alcohol-based hand rub NA 36% decrease (P value, NA) in 10 months 31
surgical unit introduction, posters, HCAI (mainly urinary tract
Hand hygiene and HCAI prevention

feedback on HCAI rates, infection and SSI) rates (from 8.2%

patient education and to 5.3%)
involvement
2004 Hospital-wide Alcohol-based hand rub No significant increase in Significant reduction (P ¼ 0.03) in 1 year 32
introduction, hand compliance before and after hospital-acquired MRSA cases
hygiene observation, patient contact (from 1.9% to 0.9%)
posters, performance
feedback, informal
discussions
2004 Adult Hand hygiene electronic Compliance increase from 19.1% Reduction in HCAI rates (not 2.5 months 33
intermediate monitoring at exit from to 27.3% by electronic statistically significant, P value,
care unit patient rooms, direct monitoring NA)
observation and voice
prompts
2004 NICU Alcohol-based hand rub Compliance increase from 40% to Reduction (P ¼ 0.14) in HCAI rates 6 months 34
introduction, hand 53% (before patient contact) and (from 11.3 to 6.2 per 1000
hygiene observation, from 39% to 59% (after patient patient-days)
training, hand-hygiene contact)
protocols, posters
2004 NICU Education, written Compliance increase from 43% to Significant reduction (P ¼ 0.003) 2 years 35
instructions, hand hygiene 80% in HCAI rates (from 15.1 to 10.7
observation, posters, per 1000 patient-days), in
performance feedback, particular for respiratory
financial incentives infections
(continued on next page)
309
Table I (continued)
310

Year Hospital Intervention Impact on hand Impact on HCAI Duration of Reference

setting hygiene compliance follow-up
2005 Hospital-wide Alcohol-based hand rub Compliance increase from Significant reduction (P ¼ 0.01) in 4 years 36
introduction, hand 62% to 81% hospital-associated rotavirus
hygiene observation, infections
training, posters
2005 Adult ICUs Hand-washing Compliance increase from Significant reduction (P < 0.001) 21 months 37
observation, training, 23.1% to 64.5% in HCAI rates (from 47.5 to 27.9
guideline dissemination, per 1000 patient-days)
posters, performance
feedback
2005 Hospital-wide Alcohol-based hand rub Compliance increase from Significant reduction (57%, 36 months 38
introduction, hand 21% to 42% P ¼ 0.01) in MRSA bacteraemia
hygiene observation,
training, posters,
promotional gadgets
2007 Neurosurgery Alcohol-based hand rub NA Reduction (54%, P ¼ 0.09) in 2 years 39
introduction, training, overall incidence of SSI.
posters Significant reduction (100%,
P ¼ 0.007) in superficial SSI rates
2007 Neonatal unit Posters, focus groups, Compliance increase from Reduction (P value, NA) in overall 27 months 40
hand hygiene observation, 42% to 55% HCAI rates (from 11 to 8.2
HCWs’ perception infections per 1000 patient-days)
assessment, feedback on and 60% decrease (P value NA) in
performance, perception risk of HCAI in very low birth
and HCAI rates weight neonates (from 15.5 to 8.8
episodes per 1000 patient-days)
2008 ICU Prospective, controlled, Compliance increase from No impact on device-associated 2 years 41
cross-over trial in two 38e37% to 68e69% infection and infections
units with education, due to multidrug-resistant
posters and alcohol- pathogens
based hand rub
introduction
B. Allegranzi, D. Pittet
2008 (1) Six pilot Alcohol-based hand rub (1) Compliance increase from (1) Significant reduction (1) 2 years 42
hospitals introduction, hand 21% to 48% (P ¼ 0.035) in MRSA bacteraemia
hygiene observation, (from 0.05 to 0.02 per 100 patient
training, posters, discharges per month) and of
promotional gadgets clinical MRSA isolates (P ¼ 0.003)
(2) All public (2) Compliance increase from (2) Reduction in MRSA (2) 1 year
hospitals in 20% to 53% bacteraemia (from 0.03 to 0.01
Victoria per 100 patient discharges per
(Australia) month, P ¼ 0.09) and of clinical
MRSA isolates (P ¼ 0.043)
2008 Urology Unit Alcohol-based hand rub Compliance increase from 0% Significant reduction (P < 0.001) 6 months 43
introduction, hand (estimation) to 28.2% in HCAI rates from 13.1% to 2.1%
Hand hygiene and HCAI prevention

hygiene observation,
training, posters, patient
education
2008 NICU Alcohol-based hand rub NA Significant reduction (P ¼ 0.009) 18 months 44
introduction, training, in HCAI incidence (4.1 vs 1.2 per
posters 1000 patient-days)
2008 NICU Alcohol-based hand rub Compliance increase from 6.3% No impact on HCAI rates (9.7 vs 7 months 45
introduction, hand to 81.2% 13.5 per 1000 patient-days) (P-
hygiene observation, value NA)
training, posters,
performance feedback,
focus groups
HCAI, healthcare-associated infection; ICU, intensive care unit; NICU, neonatal intensive care unit; MICU, medical intensive care unit; VRE, vancomycin-resistant enterococcus; MRSA,
meticillin-resistant Staphylococcus aureus; SSI, surgical site infection; NA, not available.
311
312 B. Allegranzi, D. Pittet

clinical situations. This recommendation, pro- Controversial issues related to the

moted by the CDC and WHO and embraced by use of alcohol-based hand rubs and
many national hand hygiene guidelines, is based on
Clostridium difficile spread
the evidence of better microbiological efficacy,
less time required to achieve the desired effect,
Following the widespread use of alcohol-based
point of patient care accessibility and a better skin
hand rubs as the gold standard for hand hygiene in
tolerance profile.1,29,50e56
healthcare, concern has been raised about their
The WHO Guidelines on Hand Hygiene in Health
lack of efficacy against spore-forming pathogens.
Care have been conceived to catalyse hand
Indeed, apart from iodophors, albeit at a concen-
hygiene improvement in any setting regardless
tration remarkably higher than the one used in
of the resources available and the cultural back-
antiseptics, no hand hygiene agent (including
ground.1,49,57 Since there is a strong emphasis in
alcohols, chlorhexidine, hexachlorophene,
the Guidelines and in their implementation tools
chloroxylenol, and triclosan) is reliably sporicidal
on the availability of alcohol-based hand rubs as
against Clostridium or Bacillus spp.1,58 Mechani-
a key factor for hand hygiene improvement, the
cal friction while washing hands with soap and
issue of the procurement and cost of these
water may help physically remove spores from
products, especially in developing countries, chal-
the surface of contaminated hands.59e61 As a con-
lenges the recommendation feasibility. Indeed,
sequence, contact precautions are highly recom-
global sales of commercially produced, alcohol-
mended during C. difficile-associated outbreaks,
based hand rubs in 2007 were as high as US $3
in particular, glove use and hand washing with
billion, corresponding to 295 million L in volume,
a non-antimicrobial or antimicrobial soap and
with an overall 16.3% increase compared with
water following glove removal after caring for
2003 (WHO, unpublished data), mostly observed
patients with diarrhoea.5
in Europe and North America (27% and 23% in-
The widespread use of alcohol-based hand rubs
crease, respectively). Looking at procurement
in healthcare settings has been blamed repeatedly
opportunities, these products are available only
for the increase in C. difficile-associated disease
in South Africa in the African continent and in
rates, although this has not been demonstrated
China, India, and Japan in the AsiaePacific region
by any study to date.62,63 On the contrary, the
(WHO unpublished data). The most important is-
observed increase in C. difficile-associated disease
sue curbing the purchasing power in these regions
began in the USA long before the wide use of alco-
is the high cost of these products. Market prices
hol-based hand rubs.64,65 Furthermore, one large
vary from US $2.50 to 8.40 per 100 mL dispenser
outbreak with the epidemic strain REA-group B1
and are clearly unaffordable for many developing
(equivalent to ribotype 027) was managed success-
countries. The WHO multimodal hand hygiene im-
fully by introducing alcohol-based hand rub for all
provement strategy offers a possible solution to
patients other than those with C. difficile-
this obstacle: the local production of either of
associated disease.66 In addition, several studies
two WHO-recommended hand rub formulations.1
recently demonstrated a lack of association be-
The implementation toolkit accompanying the
tween the consumption of alcohol-based hand
WHO Guidelines on Hand Hygiene in Health Care
rubs and the incidence of clinical isolates of
includes a Guide to Local Production to manufac-
C. difficile.67e69 In conclusion, discouraging the
ture alcohol-based hand rubs in hospital pharma-
widespread use of alcohol-based hand rubs for
cies or other facilities for local use.1 Two
the care of patients other than those with C. diffi-
formulations are proposed: one based on ethanol
cile-associated disease will only jeopardise overall
80% v/v, and one based on isopropyl alcohol 75%
patient safety in the long term.
v/v; both include hydrogen peroxide 0.125% v/v
and glycerol 1.45% v/v. Local production has
been carried out in many healthcare settings Discussion
worldwide and was carefully monitored and eval-
uated by WHO in several sites (WHO unpublished From the available evidence it appears that multi-
data). No major procurement, production, and modal interventions are the most suitable strategy
storage obstacles were encountered and long- to determine behavioural change leading to im-
term stability at tropical temperatures was shown proved hand hygiene compliance and reduction in
(up to 19 months). The final products complied HCAI rates. Introduction of alcohol-based hand rubs
with quality control standards and had good skin and continuous educational programmes are key
tolerability at very low cost (less than US $0.50 factors to overcome infrastructure barriers and to
per 100 mL). build solid knowledge improvement. Support by
Hand hygiene and HCAI prevention
healthcare administrators and commitment by
national and local governments are essential to
make hand hygiene an institutional and national
priority for patient safety and to ensure long-term
sustainability of promotional programmes. Higher
priority should also be given to hand hygiene as a
research topic, through good-quality, randomised,
controlled trials to determine definitively its impact
on HCAI and the relative effectiveness of the
different components of multimodal strategies.
Acknowledgements
We wish to thank all members of the Infection
Control Programme, University of Geneva Hospi-
tals and members of the WHO First Global Patient
Safety Challenge ‘Clean Care is Safer Care’ core
group (lead, D. Pittet): J. Boyce, B. Cookson, N.
Damani, D. Goldmann, L. Grayson, E. Larson, G.
Mehta, Z. Memish, H. Richet, M. Rotter, S. Sattar,
H. Sax, W.H. Seto, A. Voss, A. Widmer.
Conflict of interest statement
WHO takes no responsibility for the information
provided or the views expressed in this paper.
Funding sources
None.
References
1. WHO guidelines for hand hygiene in health care (Advanced
draft). Geneva: World Health Organization; 2006.
2. Pittet D, Allegranzi B, Sax H, et al. Evidence-based model
for hand transmission during patient care and the role of
improved practices. Lancet Infect Dis 2006;6:641e652.
3. Kretzer EK, Larson EL. Behavioral interventions to improve
infection control practices. Am J Infect Control 1998;26:
245e253.
4. Pittet D, Boyce J. Hand hygiene during patient care: pursuing
the Semmelweis legacy. Lancet Infect Dis 2001; April:9e20.
5. Siegel JD, Rhinehart E, Jackson M, Chiarello L. Guideline for
isolation precautions: preventing transmission of infectious
agents in health care settings. Am J Infect Control 2007;
35(Suppl. 2):S65eS164.
6. Eggimann P, Harbarth S, Constantin MN, Touveneau S,
Chevrolet JC, Pittet D. Impact of a prevention strategy
targeted at vascular-access care on incidence of infections
acquired in intensive care. Lancet 2000;355:1864e1868.
7. O’Grady NP, Alexander M, Dellinger EP, et al. Guidelines for
the prevention of intravascular catheter-related infections.
MMWR Recomm Rep 2002;51:1e29.
8. Pronovost P, Needham D, Berenholtz S, Sinopoli D, Chu H,
Cosgrove S. An intervention to decrease catheter-related
bloodstream infections in the ICU. N Engl J Med 2006;335:
2725e2732.
9. Mangram AL, Horan TC, Pearson ML, Silver LC, Jarvis WR,
Hospital Infection Control Practices Advisory Committee.
313
Guideline for prevention of surgical site infection, 1999.
Infect Control Hosp Epidemiol 1999;20:247e278.
10. Rampling A, Wiseman S, Davis L, et al. Evidence that hospi-
tal hygiene is important in the control of methicillin-
resistant Staphylococcus aureus. J Hosp Infect 2001;49:
109e116.
11. Dancer SJ. Importance of the environment in meticillin-
resistant Staphylococcus aureus acquisition: the case for
hospital cleaning. Lancet Infect Dis 2008;8:101e113.
12. Goodman ER, Platt R, Bass R, Onderdonk AB, Yokoe DS,
Huang SS. Impact of an environmental cleaning intervention
on the presence of methicillin-resistant Staphylococcus
aureus and vancomycin-resistant enterococci on surfaces
in intensive care unit rooms. Infect Control Hosp Epidemiol
2008;29:593e599.
13. Ahmed QA, Memish ZA, Allegranzi B, Pittet D. Muslim
health-care workers and alcohol-based handrubs. Lancet
2006;367:1025e1027.
14. Duerink DO, Farida H, Nagelkerke NJD, et al. Preventing
nosocomial infections: improving compliance with standard
precautions in an Indonesian teaching hospital. J Hosp
Infect 2006;64:36e43.
15. Allegranzi B, Memish ZA, Donaldson L, Pittet D. Religion and
culture: potential undercurrents influencing hand hygiene
promotion in healthcare. Am J Infect Control 2009;37:28e34.
16. Whitby M, PessoaSilva CL, McLaws ML, et al. Behavioural
considerations for hand hygiene practices: the basic build-
ing blocks. J Hosp Infect 2007;65:1e8.
17. Seto WH. Staff compliance with infection control practices:
application of behavioural sciences. J Hosp Infect 1995;
30(Suppl.):107e115.
18. Pittet D. The Lowbury lecture: behaviour in infection
control. J Hosp Infect 2004;58:1e13.
19. O’Boyle CA, Henly SJ, Larson E. Understanding adherence
to hand hygiene recommendations: the theory of planned
behavior. Am J Infect Control 2001;29:352e360.
20. Jenner EA, Watson PWB, Miller L, Jones F, Scott GM. Ex-
plaining hand hygiene practice: an extended application
of the theory of planned behaviour. Psychol Health Med
2002;7:311e326.
21. Whitby M, McLaws M-L, Ross RW. Why healthcare workers
don’t wash their hands: a behavioral explanation. Infect
Control Hosp Epidemiol 2006;27:484e492.
22. Casewell M, Phillips I. Hands as route of transmission for
Klebsiella species. Br Med J 1977;2:1315e1317.
23. Conly JM, Hill S, Ross J, et al. Handwashing practices in an
intensive care unit: the effects of an educational program
and its relationship to infection rates. Am J Infect Control
1989;17:330e339.
24. Simmons B, Bryand J, Neiman K, Spencer L, Arheart K. The
role of handwashing in prevention of endemic intensive
care unit infections. Infect Control Hosp Epidemiol 1990;
11:589e594.
25. Doebbeling BN, Stanley GL, Sheetz CT, et al. Comparative
efficacy of alternative hand-washing agents in reducing nos-
ocomial infections in intensive care units. N Engl J Med
1992;327:88e93.
26. Webster J, Faoagali JL, Cartwright D. Elimination of meth-
icillin-resistant Staphylococcus aureus from a neonatal
intensive care unit after hand washing with triclosan.
J Paediatr Child Health 1994;30:59e64.
27. Zafar AB, Butler RC, Reese DJ, Gaydos LA, Mennonna PA.
Use of 0.3% triclosan (Bacti-Stat) to eradicate an outbreak
of methicillin-resistant Staphylococcus aureus in a neonatal
nursery. Am J Infect Control 1995;23:200e208.
28. Larson EL, Early E, Cloonan P, Sugrue S, Parides M. An orga-
nizational climate intervention associated with increased
314
handwashing and decreased nosocomial infections. Behav
Med 2000;26:14e22.
29. Pittet D, Hugonnet S, Harbarth S, et al. Effectiveness of
a hospital-wide programme to improve compliance with
hand hygiene. Lancet 2000;356:1307e1312.
30. Pittet D, Sax H, Hugonnet S, Harbarth S. Cost implications
of successful hand hygiene promotion. Infect Control Hosp
Epidemiol 2004;25:264e266.
31. Hilburn J, Hammond BS, Fendler EJ, Groziak PA. Use of
alcohol hand sanitizer as an infection control strategy in an
acute care facility. Am J Infect Control 2003;31:109e116.
32. MacDonald A, Dinah F, MacKenzie D, Wilson A. Performance
feedback of hand hygiene, using alcohol gel as the skin
decontaminant, reduces the number of inpatients newly
affected by MRSA and antibiotic costs. J Hosp Infect 2004;
56:56e63.
33. Swoboda SM, Earsing K, Strauss K, Lane S, Lipsett PA. Elec-
tronic monitoring and voice prompts improve hand hygiene
and decrease nosocomial infections in an intermediate care
unit. Crit Care Med 2004;32:358e363.
34. Lam BC, Lee J, Lau YL. Hand hygiene practices in a neonatal
intensive care unit: a multimodal intervention and impact
on nosocomial infection. Pediatrics 2004;114:e565ee571.
35. Won SP, Chou HC, Hsieh WS, et al. Handwashing program for
the prevention of nosocomial infections in a neonatal intensive
care unit. Infect Control Hosp Epidemiol 2004;25:742e746.
36. Zerr DM, Allpress AL, Heath J, et al. Decreasing hospital-
associated rotavirus infection: a multidisciplinary hand
hygiene campaign in a children’s hospital. Pediatr Infect
Dis J 2005;24:397e403.
37. Rosenthal VD, Guzman S, Safdar N. Reduction in nosocomial
infection with improved hand hygiene in intensive care
units of a tertiary care hospital in Argentina. Am J Infect
Control 2005;33:392e397.
38. Johnson PD, Martin R, Burrell LJ, et al. Efficacy of an alco-
hol/chlorhexidine hand hygiene program in a hospital with
high rates of nosocomial methicillin-resistant Staphylococ-
cus aureus (MRSA) infection. Med J Aust 2005;183:509e514.
39. Thi Anh Thu L, Dibley MJ, Nho VV, Archibald L, Jarvis WR,
Sohn AH. Reduction in surgical site infections in neurosurgi-
cal patients associated with a bedside hand hygiene pro-
gram in Vietnam. Infect Control Hosp Epidemiol 2007;28:
583e588.
40. Pessoa-Silva CL, Hugonnet S, Pfister R, et al. Reduction of
health care associated infection risk in neonates by success-
ful hand hygiene promotion. Pediatrics 2007;120:
e382ee390.
41. Rupp ME, Fitzgerald T, Puumala S, et al. Prospective, con-
trolled, cross-over trial of alcohol-based hand gel in critical
care units. Infect Control Hosp Epidemiol 2008;29:8e15.
42. Grayson ML, Jarvie LJ, Martin R, et al. Significant reductions
in methicillin-resistant Staphylococcus aureus bacteraemia
and clinical isolates associated with a multisite, hand hy-
giene culture-change program and subsequent successful
statewide roll-out. Med J Aust 2008;188:633e640.
43. Nguyen KV, Nguyen PT, Jones SL. Effectiveness of an alco-
hol-based hand hygiene programme in reducing nosocomial
infections in the urology ward of Binh Dan Hospital, Viet-
nam. Trop Med Int Health 2008;13:1297e1302.
44. Capretti MG, Sandri F, Tridapalli E, Galletti S, Petracci E,
Faldella G. Impact of a standardized hand hygiene program
on the incidence of nosocomial infection in very low birth
weight infants. Am J Infect Control 2008;36:430e435.
45. Picheansathian W, Pearson A, Suchaxaya P. The effective-
ness of a promotion programme on hand hygiene compli-
ance and nosocomial infections in a neonatal intensive
care unit. Int J Nurs Pract 2008;14:315e321.
B. Allegranzi, D. Pittet
46. Mermel LA, Boyce JM, Voss A, Allegranzi B, Pittet D. Trial of
alcohol-based hand gel in critical care units. Infect Control
Hosp Epidemiol 2008;29:577e579; author reply 580e582.
47. McGuckin M, Waterman R. ‘‘Cannot detect a change’’ is not
the same as ‘‘there is not a change’’. Infect Control Hosp
Epidemiol 2008;29:576e577; author reply 580e582.
48. Widmer AF, Rotter M. Effectiveness of alcohol-based hand
hygiene gels in reducing nosocomial infection rates. Infect
Control Hosp Epidemiol 2008;29:576; author reply 580e582.
49. Allegranzi B, Pittet D. Healthcare-associated infection in
developing countries: simple solutions to meet complex
challenges. Infect Control Hosp Epidemiol 2007;28:
1323e1327.
50. Boyce JM, Pittet D. Guideline for hand hygiene in
health-care settings. Recommendations of the Healthcare
Infection Control Practices Advisory Committee and the
HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. MMWR
Recomm Rep 2002;51:1e45.
51. Larson EL, Eke PI, Laughon BE. Efficacy of alcohol-based
hand rinses under frequent-use conditions. Antimicrob
Agents Chemother 1986;30:542e544.
52. Larson EL, Aiello AE, Bastyr J, et al. Assessment of two hand
hygiene regimens for intensive care unit personnel. Crit
Care Med 2001;29:944e951.
53. Picheansathian W. A systematic review on the effectiveness
of alcohol-based solutions for hand hygiene. Int J Nurs Pract
2004;10:3e9.
54. Widmer AF. Replace hand washing with use of a waterless
alcohol hand rub? Clin Infect Dis 2000;31:136e143.
55. Boyce JM. Scientific basis for handwashing with alcohol and
other waterless antiseptic agents. In: Rutala WA, editor.
Disinfection, sterilization and antisepsis: principles and
practices in healthcare facilities. Washington, DC: Associa-
tion for Professionals in Infection Control and Epidemiology;
2001. p. 140e151.
56. Graham M. Frequency and duration of handwashing in an in-
tensive care unit. Am J Infect Control 1990;18:77e81.
57. Pittet D, Allegranzi B, Storr J, et al. Infection control as
a major WHO priority for developing countries. J Hosp
Infect 2008;68:285e292.
58. Rotter ML. Hand washing and hand disinfection. In:
Mayhall G, editor. Hospital epidemiology and infection
control. Baltimore: Williams & Wilkins; 1996. p. 1052e1068.
59. McFarland LV, Mulligan ME, Kwok RY, Stamm WE. Nosoco-
mial acquisition of Clostridium difficile infection. N Engl J
Med 1989;320:204e210.
60. Bettin K, Clabots C, Mathie P, Willard K, Gerding DN. Effec-
tiveness of liquid soap vs chlorhexidine gluconate for the re-
moval of Clostridium difficile from bare hands and gloved
hands. Infect Control Hosp Epidemiol 1994;15:697e702.
61. Hubner NO, Kampf G, Löffler H, Kramer A. Effect of a 1 min
hand wash on the bactericidal efficacy of consecutive surgi-
cal hand disinfection with standard alcohols and on skin
hydration. Int J Hyg Environ Health 2006;209:285e291.
62. Clabots CR, Gerding SJ, Olson MM, Peterson LR, Gerding DN.
Detection of asymptomatic Clostridium difficile carriage by
an alcohol shock procedure. J Clin Microbiol 1989;27:
2386e2387.
63. Wullt M, Odenholt I, Walder M. Activity of three disinfec-
tants and acidified nitrite against Clostridium difficile
spores. Infect Control Hosp Epidemiol 2003;24:765e768.
64. McDonald LC, Owings M, Jernigan DB. Clostridium difficile
infection in patients discharged from US short-stay hospi-
tals, 1996e2003. Emerg Infect Dis 2006;12:409e415.
65. Archibald LK, Banerjee SN, Jarvis WR. Secular trends in hos-
pital-acquired Clostridium difficile disease in the United
States, 1987e2001. J Infect Dis 2004;189:1585e1589.
Hand hygiene and HCAI prevention
66. Muto CA, Pokrwka M, Shutt K, et al. A large outbreak of
Clostridium difficile-associated disease with an unexpected
proportion of deaths and colectomies at a teaching hospital
following increased fluoroquinolone use. Infect Control
Hosp Epidemiol 2005;26:273e280.
67. Boyce JM, Ligi C, Kohan C, Dumigan D, Havill NL. Lack of
association between the increased incidence of Clostridium
difficile-associated disease and the increasing use of alco-
hol-based hand rubs. Infect Control Hosp Epidemiol 2006;
27:479e483.
315
68. Vernaz N, Sax H, Pittet D, Bonnabry P, Schrenzel J,
Harbarth S. Temporal effects of antibiotic use and hand
rub consumption on the incidence of MRSA and Clostridium
difficile. J Antimicrob Chemother 2008;62:601e617.
69. Kaier K, Hagist C, Frank U, Conrad A, Meyer E. Two time-
series analyses of the impact of antibiotic consumption
and alcohol-based hand disinfection on the incidences of
nosocomial methicillin-resistant Staphylococcus aureus in-
fection and Clostridium difficile infection. Infect Control
Hosp Epidemiol 2008;29:593e599.
Journal of Hospital Infection (2009) 73, 316e322
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Preventing surgical site infection. Where now?

H. Humphreys*

Department of Clinical Microbiology, RCSI Education and Research Centre, Smurfit Building, Beaumont
Hospital, PO Box 9063, Dublin 9, Ireland

Available online 22 August 2009

KEYWORDS Summary Surgical site infection (SSI) is increasingly recognised as a mea-

Healthcare bundles; sure of the quality of patient care by surgeons, infection control practi-
Laparoscopic surgery; tioners, health planners and the public. There is increasing pressure to
Meticillin-resistant
compare SSI rates between surgeons, institutions and countries. For this
Staphylococcus
aureus;
to be meaningful, data must be standardised and must include post-
Performance discharge surveillance (PDS) as many superficial SSIs do not present to
indicators; the original institution. Further work is required to determine the best
Post-discharge method of conducting PDS. In 2008 two important documents on SSI were
surveillance; published from the Society for Healthcare Epidemiology of America/The
Surgical site infection Infectious Disease Society of America and the National Institute for Health
and Clincal Excellence, UK. Both emphasise key aspects during the preop-
erative, operative and postoperative phases of patient care. In addition to
effective interventions known to be important for some time, e.g. not
shaving the surgical site until the day of the procedure, there is increasing
emphasis on physiological parameters, e.g. blood glucose concentrations,
oxygen tensions and body temperature. Laparoscopic procedures are in-
creasingly associated with reduced SSI rates, and the screening and
decontamination of meticillin-resistant Staphylococcus aureus carriers is
effective for certain surgical procedures but has to be balanced by cost
and the risk of mupirocin resistance. Finally, there is a need to convert
theory into practice by the rigorous application of SSI healthcare bundles.
Recent studies suggest that, with a multidisciplinary approach, simple
measures can be effective in reducing SSI rates.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

Introduction

* Tel.: þ353 1 8093708/3710; fax: þ353 1 8093709. Recent years have seen an increasing emphasis
E-mail address: hhumphreys@rcsi.ie on the importance of surgical site infection (SSI)
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.03.028
Preventing surgical site infection
from the perspective of the surgeon, infection
prevention and control practitioners, health ad-
ministrators, managers and patients. While some
healthcare-associated infections (HCAIs) such as
bloodstream infections are more likely to be
associated with life-threatening illness or even
death, SSI remains among the commonest infec-
tions occurring in acute hospitals.
In a survey of 75 694 patients in England, Wales,
Northern Ireland and the Republic of Ireland, the
overall prevalence of HCAI was 7.59%, with SSI
accounting for 14.5% of these infections.1,2 The
surveillance of some or all SSIs is conducted in
many hospitals as part of routine surveillance and
as a measure of patient safety. For a period of
15 years from 1982 to 1997, 59 335 surgical sites
were surveyed in a 600-bed hospital with an over-
all SSI rate of 4.5%, and for clean surgery 2.4%,
with little change in rates during that period.3
The significance of SSI and its prevention to
surgeons carrying out procedures has increased as
they realise that this is a measure of their effec-
tiveness. Many surgeons rightly consider that the
most critical factor in the prevention of SSI,
although difficult to scientifically quantify, is the
sound judgement and proper technique of the
surgeon and his/her team, as well as the general
health of the patient.4 It is important to ensure that
the actual surgical procedure is successful, e.g.
complete removal of a tumour, and that preventible
adverse consequences arising from not adequately
addressing infection prevention are addressed.
Quality iniatives taken by surgeons in collabora-
tion with others are bearing fruit in highlighting
the issues and possible interventions. As part of
the American College of Surgeons Standard
National Surgical Quality Improvement Programme
(NSQIP), a survey of 117 private sector hosptials
was undertaken to identify low and high outliers in
terms of SSI prevention.5 High outliers (increased
infection rates) were assoicated with higher num-
bers of trainees, longer operations, and low hae-
moglobin levels in operated patients; low outliers
had a policy of minimising operating-room traffic.5
Apart from the obvious morbidity and some-
times mortality that follows SSI, it is increasingly
recognised that there is a significant cost to the
healthcare system such as prolonged hospital stay
as well as the cost to society in disability payments
while out of work and loss of income tax revenue.
An example of the healthcare costs is described in
a caseecontrol study of SSI following coronary
artery bypass surgery in Australia between 1996
and 1998 in which 125 patients developed SSI; the
mean excess healthcare cost attributable to SSI
was $AU12,419 per case.6
317
SSI rates are increasingly seen as an example of
a performance indicator of the quality of health-
care. Changes are taking place in some countries
leading to the public reporting of HCAI, including
SSI rates.7,8 Although there are some logistical
difficulties in ensuring that rates are truly
comparable, it is likely that the future will see
patients having access to the SSI rate of their lo-
cal hospital before admission for elective surgery.
In North America, this aspect of what could be
described as ‘healthcare consumerism’ has led
to major initiatives involving healthcare profes-
sionals, patient groups and others, to try to signif-
icantly improve the quality of care by focusing on
those aspects that are amenable to prevention,
thus aiming to reduce infection rates and produce
better patient care.9 In this review, some aspects
of surveillance and recent issues in SSI prevention
are discussed.
Measuring the SSI rate
The details of techniques used for the surveillance
of SSI are beyond the scope of this review,
nevertheless it is essential that standardised def-
initions are used, e.g. classifying infections into
superficial, incisional, deep incisional and organ/
space, and that allowance is made for risk factors,
especially when infection rates are likely to be
compared between one surgeon or one institution
and another.10,11 This must be done accurately and
allow for changes in healthcare such as those re-
sulting in shorter acute hospital stay, leading to
many infections being diagnosed either in the out-
patients or in the community. This emphasises the
importance of post-discharge surveillance (PDS).
Although there are a number of options for PDS,
existing research does not appear to have identi-
fied a valid, reliable, practical and universally
agreed method for PDS.12
Whitby and colleagues in Queensland, Australia
attempted to improve the reliability of PDS
through patient self-determination and a pro-
gramme of educating patients on SSI.13 The posi-
tive predictive value for a correct diagnosis was
65.2% for the educated group compared with
83.3% for the non-educated patient group, and
the authors concluded that pre-discharge educa-
tion resulted in patients overdiagnosing SSI.13
PDS is particularly important for superficial
incisional SSI, which may not be serious enough
to warrant re-admission to hospital. For example,
in a survey in Finland of orthopaedic SSI, 86% of
SSIs detected after hospital discharge were super-
ficial.14 Data derived from 12 885 operations in
318
Scotland yielded a statistically higher SSI rate
when PDS was performed (6.34% compared with
2.61%) and the authors concluded that a proced-
ure-specific approach was required with direct
observation of patients.15 In The Netherlands, a na-
tional nosocomial surveillance network has been
collecting data on HCAI for some years. SSI rates
were determined according to different PDS
methods for surgical procedures performed in
62 hospitals.16 Recommended PDS methods, e.g.
direct observation by the surgeon, were compared
with passive surveillance, e.g. patients re-present-
ing to the hospital. Recommended compared with
passive surveillance methods detected a higher SSI
rate, 3.7% compared with 3.1%.16
It is universally acknowledged that PDS is
essential in accurately determining the SSI rate,
but there is a need for consensus as to how best
this can be achieved while taking into account
limited resources.
Preventing SSI
Approaches to the prevention and control of SSI
have evolved over many years and traditionally
have been classified into those interventions
before surgery, during surgery and after surgery.
However, prevention must be underpinned by
a knowledge and understanding of the microbial
pathogenesis, and the importance of surveillance.
Much of the advice from the Hospital Infection
Control Practices Advisory Committee (HICPAC)
from North America on SSI prevention is derived
from practice and custom, rather than rigorous
scientific studies such as randomised controlled
trials, but practice, custom and even ritual,
continue to influence approaches to prevention.17
In 2008 two important documents were pub-
lished that synthesised and presented clearly many
of the important issues pertaining to the pre-
vention of SSI. The Society for Healthcare Epi-
demiology of America and the Infectious Diseases
Society of America produced recommendations on
HCAI including SSI, addressing surveillance
methods, interventions to actively prevent SSI
and approaches to monitoring the implementation
of those strategies.11 In the UK, the National Insti-
tute for Health and Clinical Excellence published
a document on the prevention and treatment of
SSI which emerged after an extensive review of
the literature and a wide consultation exercise.10
Both documents repesent an impressive body of
literature on SSI but combine this with practical
and feasible advice about prevention; both em-
phasise the importance of education and the iden-
tification of risk factors for SSI. Some of the issues
H. Humphreys
from both documents are outlined in Table I. Some
are discussed below in terms of the evidence and
what should be incorporated into current practice.
Hand hygiene and the surgical scrub
Hand hygiene is regarded as one of the key
components in any infection prevention strategy.
For many years, the traditional surgical scrub
where the surgeon ensures that hands, nails and
parts of the forearm are lathered and scrubbed has
been standard practice. However, surgeons them-
selves accept that their practice, both in the
operating theatre and outside, has often been
suboptimal; 90% compliance is not sufficient.18 Re-
cent evidence suggests that the traditional ritual
of scrubbing can be replaced by a less lengthy
and less ritualistic procedure. In a study involving
six surgical services and 4387 consecutive patients
who underwent clean and clean-contaminated sur-
gery earlier this decade, there was no difference
in the SSI rate, whether the surgeon used the
hand-rubbing protocol with 75% aqueous alcohol
or a hand-scrubbing protocol with povidone-iodine
or chlorhexidine.19 Furthermore, the alcohol solu-
tion was better tolerated and associated with
improved compliance.
Physiological parameters
Tissue hypoxia leads to necrosis and is often
followed by infection. High blood glucose levels,
such as occur in patients with diabetes mellitus for
example, are also associated with increased risks
of infection. It is logical to assume that normalising
these will assist in the prevention of SSI, and
recent studies have supported this. Gottrup has
emphasised that the decisive period for the de-
velopment of SSI is during surgery and the first few
hours thereafter; consequently, the physiological
status of the patient at that time is crucial.20 A
higher fraction of inspired oxygen (80% compared
with 30%) was associated with a 39% lower risk
of SSI in a randomised controlled trial of 300
patients.21 It is believed that the mechanism of
action is by more effective neutrophil killing of
potential pathogens.22
In a case-controlled study of 260 patients un-
dergoing mastectomy, variables associated with
SSI included high blood glucose levels.23 Similarly,
in an assessment of risk factors for SSI after surgery
for hepato-biliary pancreatic cancer, poor blood
glucose levels were associated with an odds ratio
of 6.6 for the development of SSI.24 Physiological
homeostasis is therefore important in maintaing
body defences, and inspired oxygen, controlled
Preventing surgical site infection
Table I Major headings and some factors in preventing surgical site infection from recent UK and North American
guidelines
NICEa
Preoperative phase
Patient showering and hair removal
Patient and staff theatre wear
Movement to and from theatre area
Nasal decontamination (do not use mupirocin routinely)
Mechanical bowel preparation (not routine)
Patient and staff jewellery
Antibiotic prophylaxis (which patients,
when and number of doses)
Intraoperative phase
Hand decontamination
Incise drapes
Gowns and gloves
Antiseptic skin preparation and diathermy
Patient homeostasis (oxygenation, normothermia, etc.)
Wound irrigation and dressings
Antiseptics before closure
Postoperative phase
Dressings
Postoperative cleansing of surgical site
Topical agents (not indicated)
Antibiotic treatment and debridement for SSI
Specialist wound care services
MRSA, meticillin-resistant Staphylococcus aureus; SSI, surgical site infection.
a
Adapted from reference 10.
b
Society for Healthcare Epidemiology of America/Infectious Diseases Society of America (adapted from reference 11).
blood glucose concentrations and normothermia,
among other things, are important.
Surgical practice
Appropriate use of antimicrobial prophylaxis, i.e.
during induction to ensure adequate tissue levels
at the time of the first incision, is associated with
reduced SSI rates, particularly with surgical pro-
cedures at high risk of infection, such as those
involving the gastrointestinal tract. However, it
has taken some time before best practice has been
incorporated into routine protocols by surgeons
and others. Recently, a statement for urological
surgery incorporated important principles of ap-
propriate surgical prophylaxis and offered sensible
options in terms of the choice of antibiotics for
a variety of urological procedures.25 Indications for
prophylaxis are likely to increase with the com-
plexity of surgical procedures performed, and
with the increasing longevity of the patient popu-
lation. For example, antibiotic prophylaxis is not
usually recommended for elective clean surgical
319
SHEA/IDSAb
Surgical Care Improvement Project
Proper hair removal
Controlling blood glucose
Maintain normothermia
Infrastructure
Trained personnel
Education
Computer-assisted decision support and automated
reminders
Antimicrobial prophylaxis
Measure and provide feedback on process measures,
e.g. hair removal
Accountability
Chief executive responsible for support
Senior management ensures adequate
personnel and perform job responsibilities
Healthcare workers responsible for their practices
Non-routine approaches
Vancomycin not routine for antimicrobial prophylaxis
Don’t delay surgery for parenteral nutrition
Unresolved issues
Preoperative bathing with chlorhexidine
Positive screening for, and decolonisation of, MRSA
Supplemental oxygenation for colorectal
procedures
Maintaining normothermia after colorectal surgery
procedures such as those involving the breast.
However, prophylaxis has been recommended for
patients undergoing breast cancer surgery if they
have had previous chemotherapy, or if the surgery
involves reconstruction.26 Therefore surgeons
need to liaise with others, e.g. anaesthetists, mi-
crobiologists, infectious disease physicians and
others, to ensure that the correct agent is given
at the right time and for the appropriate duration.
One of the major advances in surgical practice in
recent decades has been the development of
laparoscopic or minimally invasive surgery. This
offers the potential for reducing infection as the
incision site is much smaller, but it is unclear
whether all other practices and the environment
setting associated with open surgery should be
replicated. In a survey of 150 hospitals in the UK
and Ireland, 3.9e5.6% of minimally invasive surgical
procedures associated with urology, e.g. trans-
urethral prosthetic surgery, uretheric stent inser-
tion, etc. took place in a non-ventilated room or in
a treatment room with some form of ventilation,
but outside a conventionally ventilated theatre.27
320
Meta-analysis of laparoscopic compared with
open repair of a perforated peptic ulcer suggests
that the risk of postoperative SSI is reduced when
carried out laparoscopically, and similar findings
have been associated with colon surgery.28e30 Most
of these procedures currently take place in a full
operating theatre, but it may be that some proced-
ures may not require the conventional ventilated
theatre with appropriate air changes, air pressures
and filters. Surgical factors, e.g. the necessity to
convert from a laparascopic to an open procedure
if the procedure is complicated, may demand that
all laparscopic procedures are carried out in oper-
ating theatres, yet the benefits for patients from
laparoscopic surgery e irrespective of where it
takes place e in terms of reduced morbidity are
increasingly clear.
Surgical site infection and Staphylococcus
aureus
Staphylococcus aureus is the commonest cause of
SSI and increasingly meticillin-resistant S. aureus
(MRSA) accounts for a greater proportion of infec-
tions in many hospitals throughout the world. Rec-
ommended measures to prevent MRSA will also
assist in the reduction of SSI and these include
screening, decontamination and glycopeptide
prophylaxis.31
The role of pre-emptive mupirocin prophylaxis
to the anterior nares to reduce MRSA carriage and
its impact on postoperative infection has been
studied in detail in recent years. This has to be
balanced by the cost of mupirocin, the risk of
resistance e particularly with repeated courses e
and the impact on postoperative SSI. However, two
recent systematic reviews and meta-analyses
strongly suggest that prophylactic intranasal mu-
pirocin significantly reduces the rate of postoper-
ative infections including that caused by MRSA and
meticillin-susceptible S. aureus (MSSA).32,33 None-
theless, there is a reluctance to use mupirocin in
all patients but to identify those patients most at
risk, i.e. those carrying S. aureus or those under-
going certain types of surgery such as orthopaedic
and cardiovascular surgery, where the con-
sequences of postoperative infection are great.
The rapid detection of S. aureus might poten-
tially assist in achieving this, as only patients
carrying S. aureus would receive mupirocin pro-
phylaxis. However, a prospective interventional
cohort study on twelve surgical ward patients using
PCR to detect MRSA did not lead to a reduction
overall in HCAI;34 the failure of this study to
demonstrate a benefit may be partly related to
the local setting, where MRSA was not endemic.
H. Humphreys
Alternatively, there may have been a failure of
one intervention, i.e. rapid MRSA detection, to
impact on what is a multifactorial process.
Converting theory to practice
Despite the scientific literature, and the publi-
cation of recommended best practice by many
professional bodies, compliance with preventive
measures is often suboptimal. A survey of almost
590 surgeons in Canada revealed that 63% were
not in compliance with recommended guidelines
on preoperative bathing, hair removal, anti-
microbial prophylaxis or intraoperative skin
preparation.35 The absence of full compliance
indicates that infection prevention and its impor-
tance is not yet embedded in all routine surgical
practice. However, studies increasingly show the
potential impact of collective efforts to minimise
SSI.
Enhancing compliance with best practice re-
quires a multidisciplinary approach and ownership
from all concerned. In a survey of antimicrobial
prophylaxis, each surgical service was internally
monitored for 30 consecutive surgical procedures
with considerable educational and communication
efforts to improve practice.36 Compliance on en-
suring that prophylactic antibiotics were started
within 60 min of surgical incision was increased
from 72.25% in the first six-month period to
83.78% during the conduct of the survey.36 None-
theless, the authors were disappointed that they
did not achieve their objective of 90% compliance.
Fifty-six hospitals participated in a project to
implement a quality improvement approach to
decrease SSI that involved appropriate prophy-
lactic antibiotic practice, preventing hypergly-
caemia, maintaining normothermia, optimising
oxygen tension and avoiding the shaving of the
surgical site.37 Improvements in process measure-
ments ranged from 3% to 27% and, during the first
three months of the programme, SSI rates fell from
2.28% to 1.65%.37 Improving compliance with best
practice is also possible locally.38
The introduction of healthcare bundles, incor-
porating validated methods and measures to pre-
vent infection, which are rigorously policed and
audited, is the way forward to decreasing SSI and
other HCAIs. Such bundles can be combined with
other measures to ensure the quality and safety of
patient care. Eight hospitals in eight cities around
the world implemented a surgical safety list that
reduced mortality and inpatient complications.39
There is likely to be increasing pressure in the
years ahead from patients/consumers and health-
care management/funding agencies to comply
Preventing surgical site infection
with best practice and to demonstrate improve-
ments in outcomes.
Discussion
SSIs are one of the commonest HCAIs and are
increasingly recognised as a measure of the quality
of healthcare. When benchmarking one surgeon
with another or between institutions, it is impor-
tant to ensure that data collected are accurate,
adjusted for risk, and incorporate PDS. There is
now a wealth of scientific literature and consensus
recommendations regarding those measures that
are important before, during and after surgery.
When these are implemented in the form of
healthcare bundles and involve multidisciplinary
input, significant improvements can be achieved.
Such improvements will be expected by patients
and members of the public as part of the in-
creasing consumerisation of healthcare. The years
ahead will see better surveillance of SSI, particu-
larly PDS. New products or devices used in patient
care, e.g. application of shock wave therapy to the
surgical site, are likely to be widely evaluated and
some may be incorporated into routine practice.40
Ultimately, while there has been some progress to
date, more can be done to prevent SSI by the sys-
tematic application of best practice before, during
and after all surgical procedures.
Conflict of interest statement/funding sources
The author was not in receipt of financial
support during the drafting of this manuscript.
However, he has received support in the last
two years for research from Steris Corporation,
3M, Cepheid and Inova8 Science. He has also
received fees for lectures or consultancy work
from 3M, Pfizer and Astellas.
References
1. Smyth ETM, McIllvenny G, Enstone JE, et al. Four country
healthcare associated infection prevalence survey 2006:
overview of the results. J Hosp Infect 2008;69:230e248.
2. Humphreys H, Newcombe RG, Enstone J, et al. Four country
healthcare associated infection prevalence survey 2006:
risk factor analysis. J Hosp Infect 2008;69:249e257.
3. Creamer E, Cunney RJ, Humpreys H, Smyth EG. Sixteen
year’s surveillance of surgical sites in an Irish acute-care
hospital. Infect Control Hosp Epidemiol 2002;23:36e40.
4. Nichols RL. Preventing surgical site infections: a surgeon’s
perspective. Emerg Infect Dis 2001;7:220e224.
5. Campbell DA, Henderson WG, Englesbe MJ, et al. Surgical
site infection prevention: the importance of operative
321
duration and blood transfusion e results of the first Ameri-
can College of SurgeonseNational Surgical Quality Improv-
ment Program Best Practices Initiative. J Am Coll Surg
2008;207:810e820.
6. Jenney AWJ, Harrington GA, Russo PL, Spelman DW. Cost
of surgical site infections following coronary artery bypass
surgery. Aust N Z J Surg 2001;71:662e664.
7. Humphreys H, Cunney R. Performance indicators and the
public reporting of healthcare-associated infection rates.
Clin Microbiol Infect 2008;14:892e894.
8. McKibben L, Horan T, Tokars JI, et al. Guidance on public
reporting of healthcare-associated infections: recommen-
dations of the Healthcare Infection Control Practices Advi-
sory Committee. Am J Infect Control 2005;33:217e226.
9. Bratzler DW, Hunt DR. The surgical infection prevention and
surgical care improvement projects: national initiatives to
improve outcomes for patients having surgery. Clin Infect
Dis 2006;43:322e330.
10. National Institute for Health and Clinical Excellence. Surgi-
cal site infection, prevention and treatment of surgical site
infection. National Collaborating Centre for Women’s and
Children’s Health; 2008.
11. Anderson DJ, Kaye KS, Classen D, et al. Strategies to pre-
vent surgical site infections in acute care hospitals. Infect
Control Hosp Epidemiol 2008;29(Suppl. 1):S51eS61.
12. Petherick ES, Dalton JE, Moore PJ, Cullum N. Methods for
identifying surgical wound infection after discharge from
hospital: a systematic review. BMC Infect Dis 2006;6:170.
13. Whitby M, McLaws M-L, Doidge S, Collopy B. Post-discharge
surgical site surveillance: does patient education improve
reliability of diagnosis? J Hosp Infect 2007;66:237e242.
14. Huotari K, Lyytikäinen O. Impact of postdischarge surveil-
lance on the rate of surgical site infection after orthopaedic
surgery. Infect Control Hosp Epidemiol 2006;27:
1324e1329.
15. Reilly J, Allardice G, Bruce J, Hill R, McCoubrey J. Proced-
ure-specific surgical site infection rates and postdischarge
surveillance in Scotland. Infect Control Hosp Epidemiol
2006;27:1318e1323.
16. Manniën J, Wille JC, Snoeren RLLM, van den Hof S. Impact
of postdischarge surveillance on surgical site infection rates
for several surgical procedures: results from the nosocomial
surveillance network in The Netherlands. Infect Control
Hosp Epidemiol 2006;27:809e816.
17. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR.
Guideline for prevention of surgical site infection, 1999.
Infect Control Hosp Epidemiol 1999;20:247e280.
18. Gawande A. On washing hands. N Engl J Med 2004;350:
1283e1286.
19. Parienti JJ, Thibon P, Heller R, et al. Hand-rubbing with
an aqueous alcoholic solution vs traditional surgical
hand-scrubbing and 30-day surgical site infection rates. A
randomised equivalence study. J Am Med Assoc 2002;288:
722e727.
20. Gottrup F. Prevention of surgical-wound infections. N Engl J
Med 2000;342:202e204.
21. Belda FJ, Aguilera L, Garcı́a de la Asunción J, et al. Supple-
mental perioperative oxygen and the risk of surgical wound
infection. A randomised controlled trial. J Am Med Assoc
2005;294:2035e2042.
22. Alan DB, Maguire JJ, Mahdavian M, et al. Wound hypoxia
and acidosis limit neutrophil bacteria killing mechanisms.
Arch Surg 1997;132:991e996.
23. Vilar-Compte D, de Iturbe I, Martı́n-Onraet A, et al. Hyper-
glycemia as a risk factor for surgical site infections in pa-
tients undergoing mastectomy. Am J Infect Control 2008;
36:192e198.
322
24. Ambiru S, Kato A, Kimura F, et al. Poor postoperative blood
glucose control increases surgical site infections after sur-
gery for hepato-biliary-pancreatic cancer: a prospective
study in a high volume institute in Japan. J Hosp Infect
2008;68:230e233.
25. Wolf JS, Bennett CJ, Dmochowski R, Hollenbeck BK,
Pearle MS, Schaeffer AJ. Best practice policy statement
on urologic surgery antimicrobial prophylaxis. J Urol 2008;
179:1379e1390.
26. Nicolas P, Yazdanpanah Y, Marie-Pierre C, et al. Prevention
of surgical site infection after breast cancer surgery by tar-
geted prophylaxis antibiotic in patients at high risk of surgi-
cal site infection. J Surg Oncol 2007;96:124e129.
27. Smyth ETM, Humphreys H, Stacey A, Taylor EW, Hoffman P,
Bannister G. Survey of operating theatre ventilation facili-
ties for minimally invasive surgery in Great Britain and
Northern Ireland: current practice and considerations for
the future. J Hosp Infect 2005;61:112e122.
28. Lunevicius R, Morkevicius M. Systematic review comparing
laparoscopic and open repair for perforated peptic ulcer.
Br J Surg 2005;92:1195e1207.
29. Romy S, Eisenbring M-C, Bettschart V, Petignat C,
Francioli P, Troillet N. Laparoscope use and surgical site in-
fections in digestive surgery. Ann Surg 2008;247:627e632.
30. Poon JT, Law W-L, Wong IW, et al. Impact of laparoscopic
colorectal resection on surgical site infection. Ann Surg
2009;249:77e81.
31. Rao GG, Osman M, Johnson L, Ramsey D, Jones S, Fidler H.
Prevention of percutaneous endoscopic gastrostomy site
infections caused by methicillin-resistant Staphylococcus
aureus. J Hosp Infect 2004;58:81e83.
32. Kallen AJ, Wilson CT, Larson RJ. Perioperative intranasal
mupirocin for the prevention of surgical-site infections:
H. Humphreys
systematic review of the literature and meta-analysis. In-
fect Control Hosp Epidemiol 2005;26:916e922.
33. van Rijen M, Bonten M, Wenzel RP, Kluytmans JAJW. Intra-
nasal mupirocin for reduction of Staphylococcus aureus
infections in surgical patients with nasal carriage:
a systematic review. J Antimicrob Chemother 2008;61:
254e261.
34. Harbath S, Fankhauser C, Schrenzel J, et al. Universal
screening for methicllin-resistant Staphylococcal aureus
at hospital admission and nosocomial infection in surgical
patients. J Am Med Assoc 2008;299:1149e1157.
35. Davis PJB, Spady D, deGara C, Forgie SED. Practices and at-
titudes of surgeons toward the prevention of surgical site
infections: a provincial survey in Alberta, Canada. Infect
Control Hosp Epidemiol 2008;29:1164e1166.
36. McCahill LE, Ahern JW, Gruppi LA, et al. Enhancing compli-
ance with medicare guidelines for surgical infection preven-
tion. Arch Surg 2007;142:355e361.
37. Dellinger EP, Hausmann SM, Bratzler DW, et al. Hospitals
collaborate to decrease surgical site infections. Am J Surg
2005;190:9e15.
38. Trussell J, Gerkin R, Coates B, et al. Impact of a patient
care pathway protocol on surgical site infection rates in
cardiothoracic surgery patients. Am J Surg 2008;196:
883e889.
39. Haynes AB, Weiser TG, Berry WR, et al. A surgical safety
checklist to reduce morbidity and mortality in a global pop-
ulation. N Engl J Med 2009;360:491e499.
40. Dumfarth J, Zimpfer D, Vo }gele-Kadletz M, et al. Prophylac-
tic low-energy shock wave therapy improves wound healing
after vein harvesting for coronary artery bypass graft sur-
gery: a prospective, randomised trial. Ann Thorac Surg
2008;86:1901e1913.
Journal of Hospital Infection (2009) 73, 323e330
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Intravascular catheter infections

J. Edgeworth*

Directorate of Infection, Guy’s and St Thomas’ Hospital, London SE1 7EH, UK

Available online 22 August 2009

KEYWORDS Summary Formerly an under-appreciated iatrogenic infection, catheter-

Catheter-related
bloodstream infection;
Costebenefit analysis;
Infection control
related bloodstream infections (CRBSIs) are now the focus of considerable
preventive strategies. Although robust clinical definitions remain elusive
due to the difficulty in identifying the focus of infection in hospitalised
patients, surveillance definitions are proving useful to monitor and com-
pare institutional rates of CRBSI and to target infection control resources.
New catheter-sparing diagnostic techniques have been developed, that are
probably most applicable to assessment of infection in stable ambulatory
patients with single long-term tunnelled catheters rather than acutely un-
well hospitalised patients. There is an impressive body of evidence that can
be used to support implementation, surveillance and audit of basic infec-
tion control practices that should help institutions reduce CRBSI rates. The
introduction of preventive antimicrobial strategies at the catheter site has
been recommended by international guidelines, yet there remains
justifiable concern about long-term selection of resistant organisms. This
has not been adequately addressed in current studies. Economic analyses
require data on the clinical effect of CRBSIs to adequately assess the
benefit; such data are scarce, owing to the difficulty in assessing the con-
tribution from comorbidities, with consequential conflicting results. Over-
all, institutions can justifiably first assess the benefit of a sustained
programme of re-enforcing basic infection control practice on CRBSI before
assessing whether the introduction of additional preventive antimicrobial
strategies are likely to have any benefit.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

Introduction

Intravascular catheters are indispensable for

modern healthcare.1 They include short-term pe-
* Tel.: þ44 207188 3107; fax: þ44 7261 9816. ripheral venous and arterial catheters, short-term
E-mail address: Jonathan.Edgeworth@gstt.nhs.uk central venous catheters (CVCs), long-term tunnelled
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.05.008
324
and cuffed CVCs and long-term peripherally in-
serted CVCs. Venous catheters are predominantly
used for administration of fluids, drugs, total paren-
teral nutrition and blood products and for renal re-
placement therapy. Arterial catheters are
predominantly used for haemodynamic monitoring
and collection of blood samples in the critical care
setting. The main complication of catheter use is
the development of an infection which can be
either localised, within the bloodstream or distal.2,3
Such infections can have life-threatening conse-
quences, particularly distal infections at sites such
as heart valves and bones, which are most
frequently due to Staphylococcus aureus.4e6 There
are a number of new approaches to diagnosis and
prevention that have been considered in recent
comprehensive meta-analyses and reviews.1,7e10
This article provides an overview of the potential
benefit of these new strategies in the context of
current infection prevention and control practice.
Epidemiology
It is estimated that about 250 000 CVCs are used in
the UK each year, and in the USA nearly 300 million
catheters are used yearly of which about 3 million
are CVCs.7,8,11 Numbers of other catheters used,
particularly peripheral venous catheters on gen-
eral wards, are not known. Most epidemiological
studies on vascular catheters report on the use of
CVCs in intensive care units (ICUs), where there
are comparatively high rates of both catheter use
and infection within a limited, easily assessable,
geographical setting. It is important to remember,
however, that insertion of any vascular catheter
in any clinical setting can result in a catheter
infection. Most studies and surveillance schemes
focus on catheter-related bloodstream infections
(CRBSIs) as a measure of catheter infection and
use different denominators to compare rates
including number of patients, patient days, cathet-
ers and catheter days. A recent systematic review
of 200 prospective studies estimated that
incidence rates, expressed as CRBSIs/1000 catheter-
days, range from 0.5 for peripheral venous
catheters, to 1.6 for long-term CVCs, 1.7 for ar-
terial catheters, 2.1 for peripherally inserted
CVCs, and 2.7 for short-term CVCs.12
Clinical descriptions, microbiology and
pathogenesis
Local catheter infections occur at the exit site,
along the subcutaneous tract of tunnelled long-
J. Edgeworth
term catheters, and along the body and tip of
the catheter within the intravascular compart-
ment.3,13 Significant exit-site and tunnel infections
are commonly due to pyogenic organisms, particu-
larly S. aureus, resulting in a combination of a hot,
red catheter exit site, cellulitis along the tract and
pus formation. The visual signs usually prompt
early removal of the catheter, although laboratory
confirmation of infection is only made in a small
proportion of cases, with mechanical and chemical
irritation being alternative causes of local inflam-
mation. Catheters lacking any overt signs of local
infection are also frequently removed from
patients suspected of having a healthcare-associ-
ated infection, and between 10% and 25% are
found to grow significant numbers of organisms
without an attendant bloodstream infection.3,10
This has been referred to as either catheter coloni-
sation or infection. All organisms commonly
cultured from hospitalised patients can also be
isolated from catheter tips including skin staphylo-
cocci, Candida spp., enterococci and Gram-negative
bacteria. The extent to which these organisms
cultured from catheter tips contribute to a sys-
temic inflammatory response in the absence of
a documented CRBSI remains unclear. It is an
area warranting further investigation.
There are two main pathways leading to
CRBSI.3,14,15 The first is contact between skin sur-
face organisms either at the time of insertion or
thereafter, leading to migration of organisms
down, and colonisation of, the external catheter
surface. It is thought to be the dominant mech-
anism associated with short-term catheters.15
The second involves transfer of organisms to the
catheter hubs from patient skin or healthcare
workers’ hands, usually leading to colonisation of
the internal catheter surface, and is more common
in long-term catheters.14,16 In both cases organisms
should be present on the catheter tip and in the
blood, although for the hub infections organisms
may be confined to the internal catheter surface.
Organisms colonise catheters through two main
mechanisms: adherence to the catheter surface,
facilitated by host proteins such as fibronectin and
fibrinogen adsorbed on to the catheter surface,
and formation of biofilm on the catheter surface,
which protects organisms in a vegetative state
where they resist both the innate immune system
and antibiotics.3,13,17,18 S. aureus and coagulase-
negative staphylococci (CoNS) both express
surface adhesins that permit binding to host extra-
cellular matrix proteins deposited on catheters
and CoNS, in particular, can produce factors
involved in formation of a biofilm.19e22 However,
no single factor or group of related molecules has
Intravascular catheter infections
been shown to be essential for catheter colonisa-
tion. Although staphylococci are the predominant
organisms isolated from catheters, other hospital
bacteria can be isolated from catheters suggesting
other mechanisms of colonisation. Indeed, the
dominance of staphylococci may be at least in part
explained by the opportunity they have for contact
with catheters due to their residence on skin.
Definitions of bloodstream infections
Robust definitions of bloodstream infections are
required for clinical diagnosis and management,
surveillance and investigational studies. Definitions
used depend upon these varied indications and also
any mandated requirements and local resources.
Healthcare-associated BSI have been divided into
primary (unknown), secondary to a specified focus
(such as skin and soft tissue, or the respiratory,
urinary and gastrointestinal tracts) and catheter-
related events. Clinical definitions of CRBSI gener-
ally require culture of an organism from blood
supported by one of a number of diagnostic tests
(to be discussed later) implicating the catheter, and
no apparent alternative focus for the infection.23
The difficulty with such clinical definitions is the
extent to which alternative foci are sought and
then deemed to be a ‘plausible’ alternative focus.
Identification of foci for infection causing sepsis in
ICU is notoriously difficult, and even when strict cri-
teria are used the identification retains a degree of
subjectivity in patients who are acutely unwell for
other reasons. A review of five representative
large ICU clinical studies shows a wide variation in
the proportion of hospital-acquired bloodstream
infection assigned to these three categories:
primary (22e61%), secondary (16e60%), CRBSI
(19e62%).24e28 Although this may represent hetero-
geneity in clinical practice and patient populations
it probably also reflects differences in local inter-
pretation, which complicates communication of
studies in this field. Given the location of colonised
catheters in the bloodstream, available evidence
would equally support definitions where alternative
foci for a bloodstream infection were considered
only after catheters had been excluded.
Surveillance definitions remove much of the
subjectivity surrounding clinical definitions. The
CDC definition of a catheter-associated infection is
a positive blood culture in a patient in whom
a central line has been in use <48 h previously.23
Surveillance definitions are expressed as per 1000
bed-days, per 1000 catheters or per 1000 cath-
eter-days. Although per 1000 catheter-days is
recognised to be the most informative measure
325
of risk, it requires the most resources, particularly
if applied across a whole hospital.12
Improving diagnosis of CRBSI for clinical
and economic benefit
A common approach to diagnosis of CRBSI in
a patient with suspected healthcare-associated
infection with a single short-term peripheral cath-
eter or CVC is to take a blood culture and remove
the catheter for semiquantitative analysis.29 The
catheter tip is ‘rolled’ across an agar plate and
CRBSI is diagnosed if the same organism is isolated
from blood culture and the tip culture, and the
quantity of organisms isolated from the tip is >15
colonies. This approach is simple and cheap, and
pooled data from 19 studies including short- and
long-term catheters demonstrate a sensitivity
and specificity of 85% and 82%, respectively.9
Alternatively a fixed length of catheter can be
immersed in broth, sonicated and vortexed to
release organisms into suspension that can then be
plated out and counted to give a quantitative mea-
sure of catheter load. A cut-off of >103 organisms
has been used. This technique has the advantage
of removing organisms from the internal catheter
surface. Pooled data from 14 studies indicate a sensi-
tivity and specificity of 83% and 87%, respectively.9
Diagnoses requiring catheter removal are prob-
lematic in settings such as the ICU where patients
have multiple short-term central venous and ar-
terial catheters, any or none of which may be the
cause of an infection. Replacement of all cathet-
ers is time-consuming, expensive and not without
clinical risk including pneumothorax and haemor-
rhage, particularly in patients with a bleeding
tendency. Similarly, in ambulatory patients who
usually have a single long-term tunnelled catheter,
immediate catheter removal often necessitates
hospital admission with surgical and or radiological
intervention and unnecessary replacement may
compromise precious vascular access. In both
settings, excluding a catheter as the infected
source while still in situ can preserve the catheter
life with potential clinical and economic benefit. A
number of catheter-sparing diagnostic approaches
have been developed, of which simultaneous,
quantitative culture and differential time to
positivity have shown the most promise.
Simultaneous quantitative blood cultures
Equal volumes of blood, usually 10 mL, are drawn
simultaneously through catheter(s) and a peripheral
vein. Blood is lysed and centrifuged and the blood is
plated out on agar. Although studies have reported
326
different cut-offs for a positive diagnosis, the Infec-
tious Diseases Society of America has recommended
a five-fold greater colony count from a catheter
culture to indicate it as the source of CRBSI. A
meta-analysis of diagnostic techniques has shown
this to be the most accurate test with a sensitivity
and specificity for short-term catheters of 75% and
97%, respectively, and for long-term catheters of
93% and 100%, respectively.9 This increases the
number of blood cultures processed by a laboratory,
however, and requires considerable additional work.
Differential time to positivity
Blood cultures are usually monitored using a real-
time continuous detection system which can re-
cord the time when organism growth is first
detected. Studies have shown that, when blood
cultures drawn through a catheter flag positive
more than 2 h before a peripheral blood culture,
the catheter can be implicated as the focus with
a sensitivity and specificity of 89% and 87%, respec-
tively, for short-term catheters and 90% and 72%,
respectively, for long-term catheters.9 There are
concerns that this method particularly may be
compromised by antibiotics received through the
catheter at or around the same time.
Institutions considering catheter-sparing tech-
niques should assess the impact on the laboratory
before implementation. Where a single blood
culture may have been taken from a patient with
a suspected hospital-acquired infection, this may
increase 3e5-fold in patients that have many
catheters in situ such as those on ICU. Laboratory
blood culture machines have a limited capacity.
Furthermore, the simultaneous collection of blood
through catheters can result in isolation of organ-
isms such as meticillin-resistant S. aureus (MRSA),
without confirmation of being a true CRBSI from
a peripherally obtained culture; in the UK this is
nonetheless recorded as an institutional MRSA
bacteraemia for surveillance purposes and could
potentially divert infection control resources and
lead to unnecessary clinical interventions. There-
fore, if catheter-sparing diagnostic techniques
are being considered it seems most appropriate
to introduce them in a setting such as renal dialysis
or oncology where single long-term catheters are
used and where catheter salvage probably has
the greatest clinical benefit.
Basic and enhanced strategies for
preventing CRBSI
Comprehensive national and international guide-
lines include a large number of recommendations
J. Edgeworth
on how to prevent CRBSI, supported by an impres-
sive body of evidence.23,30 Some guidelines
consider all catheters whereas most focus on
long-term CVCs which are associated with the
greatest infection burden.12 The challenge for
healthcare institutions is to integrate these com-
prehensive guidelines into different clinical set-
tings within available resources, supported by
a surveillance system to monitor adherence and
improvements.
Core components of a strategy for preventing
CRBSI include: hand hygiene; maximum sterile
barrier precautions for CVC insertions; 2% chlor-
hexidine for skin antisepsis; use of sterile semi-
permeable dressings to facilitate daily inspection;
documentation of exit site appearance; avoidance
of the femoral site for insertion for CVCs; removal
of the catheter as soon as it is no longer required;
routine removal of peripheral venous catheters
after 3 days; routine replacement of tubings and
administration sets; and use of a tunnelled dual
lumen catheter if placement is anticipated to be
more than 21 days.23,30
There is a considerable body of data to support
the majority of these interventions. One recent
comprehensive review stated that, ‘This is a well
researched area and few realistic research needs
were identified in developing these guidelines.’30
Nonetheless, some aspects of care still require fur-
ther investigation. For example, it has not been
determined whether 2% chlorhexidine in 70% alco-
hol, the current preferred solution for skin anti-
sepsis, is superior to 5% povidone iodine in 70%
alcohol, which has itself been demonstrated as
superior to 10% povidone iodine solution. It is
also not clear exactly how frequently dressings
and tubings should be changed.
An important focus recently has been to in-
tegrate these recommendations into a care pack-
age that can be implemented across an institution.
There have been a number of reports in the
literature, mostly in ICUs, documenting consider-
able success using a combination of these inter-
ventions.31,32 One notable study combined
improved hand hygiene, full barrier precautions
for CVC insertion, chlorhexidine for skin antisepsis,
avoidance of the femoral site and removal of un-
necessary lines.32 It reported a reduction in CRBSIs
from 7.7 per 1000 CVC-days to 2.3 per 1000 CVC-
days, which remained at low rates for 18 months
after the intervention period. Concerns have
been raised about the before-and-after study
design, compliance, potential for reversion to
mean, lack of data on relative importance of
individual components, definitions used and the
Hawthorne effect of being involved in the study.
Intravascular catheter infections
Nonetheless, it is consistent with previous studies
and has contributed to a body of data supporting
the UK national intervention programme, ‘Saving
Lives’, which includes care bundles consisting of
the core components described above to target
peripheral catheters, long- and short-term CVCs.33
It is recognised that the success of individual
institutional programmes will be dependent upon
introducing a comprehensive education and
training programme supported by pragmatic easy-
to-introduce audit, surveillance and feedback,
clear lines of accountability from ‘board to floor’
and clinical engagement and leadership.
Antimicrobial strategies and
antimicrobial lock
A number of enhanced strategies have also been
assessed that have as a common mechanism the
provision of sustained antimicrobials or antiseptics
at the catheter site.1,7,10 They include chlorhexi-
dine-impregnated sponges and dressings, cathe-
ters impregnated with antimicrobial agents, and
antimicrobial locks using antibiotics (including
vancomycin, gentamicin, cephalosporins, minocy-
cline and ciprofloxacin) or taurolidine.
Chlorhexidine-impregnated sponges
These are placed over the CVC insertion site and
covered with a transparent dressing. In theory they
should prevent growth and migration of organisms
along the external catheter surface but would
have no effect on hub colonisation and migration
on the internal surface. A number of randomised
studies in adult and paediatric groups have shown
variable reductions in CRBSI with use of these
sponges.1 A meta-analysis of five randomised
studies showed a significant reduction in catheter
tip colonisation and a non-significant trend to-
wards a reduction in CRBSI.34
Antimicrobial CVCs
Impregnation of CVCs with antimicrobial agents is
potentially an effective way of ensuring an appro-
priate concentration of antimicrobial along the
whole length of the catheter to prevent colonisa-
tion.7 First-generation antimicrobial catheters
have either chlorhexidine and silver sulphadiazine
(CSS), or silver alone impregnated on the external
catheter surface. Second-generation catheters are
treated on both the external surface and internal
lumen with either CSS, silver or benzalkonium
327
chloride, so as to prevent internal colonisation,
thought to be a more common mechanism after
the first couple of weeks of insertion. Catheters
impregnated with either minocycline and rifampi-
cin or miconazole and rifampicin have also been
developed.
There have been six systematic reviews of the
clinical effectiveness of antimicrobial CVCs on re-
ducing CRBSI, and a recent comprehensive Health
Technology Assessment, in which earlier reviews
were assessed and compared.8 The odds ratios (ORs)
of reducing CRBSIs for first-generation, second-
generation and antibiotic-coated CVCs all indicated
an effect compared with standard catheters,
although for first-generation catheters there was
a wider, non-significant confidence interval (CI)
(OR: 0.67, 95% CI: 0.43e1.06; 0.43, 0.26e0.70;
0.26, 0.15e0.46 respectively).8 It has been noted
that when studies including first-generation cath-
eters with short median insertion times (6 days;
range: 5.2e7.5) are assessed, a significant reduc-
tion in CRBSI is observed (0.48, 0.25e0.91).35
Indeed, when all antimicrobial catheters are
included, a significant reduction is only seen for
studies in which average catheter insertion was
5e12 days and not 13e20 days (0.4, 0.27e0.58 and
0.69, 0.42e1.14 respectively) although the
trend is in the same direction.8 Only one study
with a catheter insertion of >20 days has been
reported, demonstrating a significant benefit of
the antimicrobial catheter.36 Overall, the number
of antimicrobial catheters required to prevent one
CRBSI ranges from 13 to 221 depending upon the
rate in the control group.
Antimicrobial locks
This strategy involves infusion of 2e3 mL of an
antibiotic-containing solution, usually with an anti-
coagulant into a CVC at the end of a period of use.
Since patients in critical care usually have
frequently accessed multiple short-term catheters,
this strategy has predominantly been assessed with
long-term catheters in haemodialysis and oncology
patients. A recent review identified three studies
assessing vancomycin lock with heparin compared
with heparin alone in a neonatal ICU, paediatric
and adult cancer patients.10 Although each setting
is quite different, all three studies demonstrated
a significant reduction in CRBSI (0.37 vs 1.72 per
1000 catheter-days; 2.3 vs 17.8 per 1000 catheter-
days; 0 per 60 vs 4/57 CVCs; vancomycin vs controls
in each case).37e39 A few studies in haemodialysis
patients have assessed use of gentamicin and one
study using either a cephalosporin or neomycin
have all shown a similar reduction.10 Three small
328
studies have assessed the use of a 1.38% taurolidine
plus 3.8% citrate lock compared with heparin in
haemodialysis patients. Two reported a significant
reduction in either CRBSI (90 day bacteraemia
free survival rate 94 vs 47%; P ¼ 0.001) or an in-
crease in sepsis free catheter survival (0 vs 4 epi-
sodes; P ¼ 0.047), but one larger study reported
no significant effect on CRBSI (0.85 vs 0.81 per
1000 catheter days).40e42 Further studies are
warranted.
Clinical and economic evaluation of the
benefit of enhanced antimicrobial
strategies
The EPIC (Evidence-based Practice in Infection
Control) guidelines have recommended antimicro-
bial CVC for adult patients at risk of CRBSI who
require central venous access for 1e3 weeks.30
The Healthcare Infection Control Practices
Advisory Committee (HICPAC) of the Centers for
Disease Control and Prevention has recommended
the use of antimicrobial CVCs in adults whose cath-
eter is expected to remain in situ for >5 days and
where the CRBSI rate is above the goal set by the
institution despite implementation of a compre-
hensive strategy to reduce rates.23 There are
frequent methodological concerns in reported
studies, including the diagnostic criteria used,
inadequate blinding and in many cases poor ran-
domisation, so it is understandable that different
views have emerged on implementation of these
recommendations.7,10 It is useful to review the
clinical and economic evaluation of the benefit of
preventing CRBSIs that have underpinned these
recommendations.
There have been three formal cost-effective-
ness studies of antimicrobial catheters performed
in high risk or ICU settings.43e45 All included eco-
nomic models informed by a combination of
patient-level data from randomised studies and
outcome data from other sources. Models were
tested using multivariate sensitivity analysis across
a range of clinical and economic assumptions. All
concluded that antimicrobial catheters were
cost-effective. None was performed in the UK. An
estimate has been made of the cost-effectiveness
of using antimicrobial catheters to prevent CRBSIs
in the UK using a simple economic model.8 The
model used parameter estimates derived from
published studies with an incidence of CRBSI at
3% (range: 2e5%), relative risk reduction with anti-
microbial catheters of 54% (range: 38e66%), cost
of CRBSI of £9,148 (range: £2,500e£71,000) and
a price differential between normal and
J. Edgeworth
antimicrobial catheters of £10 (range: £2.50e
£25). All but one of the 81 potential scenarios
tested in their model using multivariate sensitivity
analysis demonstrated an economic benefit of an-
timicrobial catheters. Antimicrobial catheters
were shown not to be cost-effective only at the
lowest rate of CRBSI in control population, lowest
relative risk reduction, highest cost different for
the antimicrobial catheters, and lowest cost of
a CRBSI e a scenario the authors consider to be
very unlikely in the real world.
Many ICU studies have attempted to estimate
the excess length of stay (LOS) due to CRBSI as
a proxy for additional costs; these have reported
between 10 and 20 days.8 None has been per-
formed in the UK. Most of these studies have
used a caseecontrol design which is fraught with
difficulties in both selecting controls and adjusting
for potential confounding variables. Traditional
statistical approaches cannot overcome the prob-
lem that whereas a CRBSI can increase LOS, the
reverse may be equally true. Alternative statistical
approaches are being used that overcome these
inherent confounders, such as longitudinal model-
ling, where effects of variables on LOS can be
assessed daily and so a temporal relationship
between healthcare-acquired infection and in-
creased length of stay can be made. One study
using such an approach reported a non-significant
increase in LOS for CRBSI of 2.6 days.46
An important clinical consideration for use of
enhanced interventions to prevent CRBSI is the
potential to save lives. The attributable mortality
of CRBSI in caseecontrol studies, mostly recruiting
patients from ICUs, has been reported to vary from
0 to 35%.26,28,47e51 Studies have used a wide range
of matching criteria to select controls based on
admission parameters [specialty, diagnosis, Acute
Physiological Assessment and Chronic Health Eval-
uation (APACHE) II, Simplified Acute Physiology
Score (SAPS), Logistic Organ Dysfunction (LOD)
score] and on-unit parameters (3 day recalibrated
outcome score). Adjustments have also been made
for confounders occurring on the unit prior to the
CRBSI that impact on mortality using a day 3 or
day 4 LOD score. Studies that employed these so-
phisticated approaches have generally not been
able to show significant effects of CRBSI on mortal-
ity when all organisms are included.26,47,49,51 One
large study performed a subgroup analysis and
reported no attributable mortality even when
CoNS were excluded, either for S. aureus or for
Gram-negative CRBSI alone.47
At an institutional level a number of additional
questions may well be raised before introducing
antimicrobial catheters. For example, it is
Intravascular catheter infections
probable that meta-analyses of randomised stud-
ies have been able to show a significant benefit
only because more than half of CRBSIs are CoNS;
although a significant clinical burden in a neonatal
unit and perhaps ambulatory patients with long-
term catheters, CoNS are less significant in adult
critical care. There is also concern about resis-
tance to both antiseptics and antibiotics; qacA/B
and other efflux pumps conferring in vitro resis-
tance to cationic biocides including chlorhexidine
are carried by both CoNS and increasingly in
some strains of MRSA.52 Resistance to rifampicin
has been noted in CoNS at the catheter site in
one ICU study.53 Two studies have also noted an in-
creased risk of catheter colonisation with candida
[10/187 (5.2%) vs 2/180 (1.1%); relative risk: 4.9;
95% CI: 1.07e22.2].54,55 It may be tempting to
wait for a larger, more comprehensive study to
be performed, but in an era of reduced incidence
of CRBSI in institutions that have re-enforced basic
infection control measures e arguably the setting
where such studies should be conducted e it is
estimated that 10 000 patients would be required
in each study arm to address the methodological
concerns of previous investigations, and so such
a study is unlikely to take place.8
Conflict of interest statement
None declared.
Funding sources
None.
References
1. Raad I, Hanna H, Maki D. Intravascular catheter-related in-
fections: advances in diagnosis, prevention, and manage-
ment. Lancet Infect Dis 2007;7:645e657.
2. Jarvis WR, Edwards JR, Culver DH, et al. Nosocomial infec-
tion rates in adult and pediatric intensive care units in the
United States. National nosocomial infections surveillance
system. Am J Med 1991;91:185Se191S.
3. Raad II, Bodey GP. Infectious complications of indwelling
vascular catheters. Clin Infect Dis 1992;15:197e208.
4. Lamas C, Boia M, Eykyn SJ. Osteoarticular infections com-
plicating infective endocarditis: a study of 30 cases be-
tween 1969 and 2002 in a tertiary referral centre. Scand J
Infect Dis 2006;38:433e440.
5. Lamas CC, Eykyn SJ. Bicuspid aortic valve e a silent danger:
analysis of 50 cases of infective endocarditis. Clin Infect Dis
2000;30:336e341.
6. Fowler Jr VG, Justice A, Moore C, et al. Risk factors for
hematogenous complications of intravascular catheter-
associated Staphylococcus aureus bacteremia. Clin Infect
Dis 2005;40:695e703.
7. Casey AL, Mermel LA, Nightingale P, Elliott TS. Antimicro-
bial central venous catheters in adults: a systematic review
and meta-analysis. Lancet Infect Dis 2008;8:763e776.
329
8. Hockenhull JC, Dwan K, Boland A, et al. The clinical effec-
tiveness and cost-effectiveness of central venous catheters
treated with anti-infective agents in preventing blood-
stream infections: a systematic review and economic eval-
uation. Health Technol Assess 2008;12:iiieiv. xiexii, 1e154.
9. Safdar N, Fine JP, Maki DG. Meta-analysis: methods for diag-
nosing intravascular device-related bloodstream infection.
Ann Intern Med 2005;142:451e466.
10. Timsit JF. Diagnosis and prevention of catheter-related
infections. Curr Opin Crit Care 2007;13:563e571.
11. Crnich CJ, Maki DG. The role of intravascular devices in
sepsis. Curr Infect Dis Rep 2001;3:496e506.
12. Maki DG, Kluger DM, Crnich CJ. The risk of bloodstream
infection in adults with different intravascular devices:
a systematic review of 200 published prospective studies.
Mayo Clin Proc 2006;81:1159e1171.
13. Wenzel R, editor. Prevention and control of nosocomial
infections. Baltimore: Lippincott Williams & Wilkins; 2003.
14. Linares J, Sitges-Serra A, Garau J, Perez JL, Martin R.
Pathogenesis of catheter sepsis: a prospective study with
quantitative and semiquantitative cultures of catheter
hub and segments. J Clin Microbiol 1985;21:357e360.
15. Maki DG, Stolz SM, Wheeler S, Mermel LA. Prevention of
central venous catheter-related bloodstream infection by
use of an antiseptic-impregnated catheter. A randomized,
controlled trial. Ann Intern Med 1997;127:257e266.
16. Raad I, Costerton W, Sabharwal U, Sacilowski M, Anaissie E,
Bodey GP. Ultrastructural analysis of indwelling vascular
catheters: a quantitative relationship between luminal col-
onization and duration of placement. J Infect Dis 1993;168:
400e407.
17. Bisognano C, Vaudaux P, Rohner P, Lew DP, Hooper DC.
Induction of fibronectin-binding proteins and increased
adhesion of quinolone-resistant Staphylococcus aureus
by subinhibitory levels of ciprofloxacin. Antimicrob
Agents Chemother 2000;44:1428e1437.
18. Stewart PS, Rayner J, Roe F, Rees WM. Biofilm penetration
and disinfection efficacy of alkaline hypochlorite and chloro-
sulfamates. J Appl Microbiol 2001;91:525e532.
19. Francois P, Schrenzel J, Stoerman-Chopard C, et al. Identi-
fication of plasma proteins adsorbed on hemodialysis tubing
that promote Staphylococcus aureus adhesion. J Lab Clin
Med 2000;135:32e42.
20. Patti JM, Allen BL, McGavin MJ, Hook M. MSCRAMM-medi-
ated adherence of microorganisms to host tissues. Annu
Rev Microbiol 1994;48:585e617.
21. Gray ED, Peters G, Verstegen M, Regelmann WE. Effect of
extracellular slime substance from Staphylococcus epider-
midis on the human cellular immune response. Lancet
1984;1:365e367.
22. Rupp ME, Ulphani JS, Fey PD, Mack D. Characterization of
Staphylococcus epidermidis polysaccharide intercellular
adhesin/hemagglutinin in the pathogenesis of intravascular
catheter-associated infection in a rat model. Infect Immun
1999;67:2656e2659.
23. O’Grady NP, Alexander M, Dellinger EP, et al. Guidelines for
the prevention of intravascular catheter-related infections.
Infect Control Hosp Epidemiol 2002;23:759e769.
24. Edgeworth JD, Treacher DF, Eykyn SJ. A 25-year study of
nosocomial bacteremia in an adult intensive care unit. Crit
Care Med 1999;27:1421e1428.
25. Orsi GB, Di Stefano L, Noah N. Hospital-acquired, labora-
tory-confirmed bloodstream infection: increased hospital
stay and direct costs. Infect Control Hosp Epidemiol 2002;
23:190e197.
26. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream in-
fection in critically ill patients. Excess length of stay, extra
330
costs, and attributable mortality. J Am Med Assoc 1994;
271:1598e1601.
27. Rello J, Ricart M, Mirelis B, et al. Nosocomial bacteremia in
a medicalesurgical intensive care unit: epidemiologic char-
acteristics and factors influencing mortality in 111 episodes.
Intensive Care Med 1994;20:94e98.
28. Renaud B, Brun-Buisson C. Outcomes of primary and cath-
eter-related bacteremia. A cohort and caseecontrol study
in critically ill patients. Am J Respir Crit Care Med 2001;
163:1584e1590.
29. Maki DG, Weise CE, Sarafin HW. A semiquantitative culture
method for identifying intravenous-catheter-related infec-
tion. N Engl J Med 1977;296:1305e1309.
30. Pratt RJ, Pellowe CM, Wilson JA, et al. Epic2: National evi-
dence-based guidelines for preventing healthcare-associ-
ated infections in NHS hospitals in England. J Hosp Infect
2007;65(Suppl. 1):S1e64.
31. Gastmeier P, Geffers C. Prevention of catheter-related
bloodstream infections: analysis of studies published be-
tween 2002 and 2005. J Hosp Infect 2006;64:326e335.
32. Pronovost P, Needham D, Berenholtz S, et al. An interven-
tion to decrease catheter-related bloodstream infections
in the ICU. N Engl J Med 2006;355:2725e2732.
33. Department of Health UK. Going further faster: implement-
ing the saving lives delivery programme. Saving lives:
a delivery programme to reduce HCAI. London: DoH; 2006.
34. Ho KM, Litton E. Use of chlorhexidine-impregnated dressing
to prevent vascular and epidural catheter colonization and
infection: a meta-analysis. J Antimicrob Chemother 2006;
58:281e287.
35. Walder B, Pittet D, Tramer MR. Prevention of bloodstream
infections with central venous catheters treated with
anti-infective agents depends on catheter type and inser-
tion time: evidence from a meta-analysis. Infect Control
Hosp Epidemiol 2002;23:748e756.
36. Hanna H, Benjamin R, Chatzinikolaou I, et al. Long-term
silicone central venous catheters impregnated with mino-
cycline and rifampin decrease rates of catheter-related
bloodstream infection in cancer patients: a prospective ran-
domized clinical trial. J Clin Oncol 2004;22:3163e3171.
37. Carratala J, Niubo J, Fernandez-Sevilla A, et al. Random-
ized, double-blind trial of an antibiotic-lock technique for
prevention of gram-positive central venous catheter-
related infection in neutropenic patients with cancer.
Antimicrob Agents Chemother 1999;43:2200e2204.
38. Garland JS, Alex CP, Henrickson KJ, McAuliffe TL, Maki DG.
A vancomycineheparin lock solution for prevention of noso-
comial bloodstream infection in critically ill neonates with
peripherally inserted central venous catheters: a prospec-
tive, randomized trial. Pediatrics 2005;116:e198ee205.
39. Henrickson KJ, Axtell RA, Hoover SM, et al. Prevention of
central venous catheter-related infections and thrombotic
events in immunocompromised children by the use of van-
comycin/ciprofloxacin/heparin flush solution: a random-
ized, multicenter, double-blind trial. J Clin Oncol 2000;
18:1269e1278.
40. Allon M. Prophylaxis against dialysis catheter-related bac-
teremia with a novel antimicrobial lock solution. Clin Infect
Dis 2003;36:1539e1544.
J. Edgeworth
41. Betjes MG, van Agteren M. Prevention of dialysis catheter-
related sepsis with a citrateetaurolidine-containing lock
solution. Nephrol Dial Transplant 2004;19:1546e1551.
42. Quarello F, Forneris G. Prevention of hemodialysis catheter-
related bloodstream infection using an antimicrobial lock.
Blood Purif 2002;20:87e92.
43. Marciante KD, Veenstra DL, Lipsky BA, Saint S. Which anti-
microbial impregnated central venous catheter should
we use? Modeling the costs and outcomes of antimicrobial
catheter use. Am J Infect Control 2003;31:1e8.
44. Shorr AF, Humphreys CW, Helman DL. New choices for cen-
tral venous catheters: potential financial implications.
Chest 2003;124:275e284.
45. Veenstra DL, Saint S, Saha S, Lumley T, Sullivan SD. Efficacy
of antiseptic-impregnated central venous catheters in
preventing catheter-related bloodstream infection: a
meta-analysis. J Am Med Assoc 1999;281:261e267.
46. Beyersmann J, Gastmeier P, Grundmann H, et al. Use of
multistate models to assess prolongation of intensive care
unit stay due to nosocomial infection. Infect Control Hosp
Epidemiol 2006;27:493e499.
47. Blot SI, Depuydt P, Annemans L, et al. Clinical and economic
outcomes in critically ill patients with nosocomial catheter-
related bloodstream infections. Clin Infect Dis 2005;41:
1591e1598.
48. Digiovine B, Chenoweth C, Watts C, Higgins M. The attribut-
able mortality and costs of primary nosocomial bloodstream
infections in the intensive care unit. Am J Respir Crit Care
Med 1999;160:976e981.
49. Garrouste-Orgeas M, Timsit JF, Tafflet M, et al. Excess risk
of death from intensive care unit-acquired nosocomial
bloodstream infections: a reappraisal. Clin Infect Dis
2006;42:1118e1126.
50. Rello J. Impact of nosocomial infections on outcome: myths
and evidence. Infect Control Hosp Epidemiol 1999;20:
392e394.
51. Soufir L, Timsit JF, Mahe C, Carlet J, Regnier B, Chevret S.
Attributable morbidity and mortality of catheter-related
septicemia in critically ill patients: a matched, risk-
adjusted, cohort study. Infect Control Hosp Epidemiol
1999;20:396e401.
52. Smith K, Gemmell CG, Hunter IS. The association between
biocide tolerance and the presence or absence of qac genes
among hospital-acquired and community-acquired MRSA
isolates. J Antimicrob Chemother 2008;61:78e84.
53. Chatzinikolaou I, Hanna H, Graviss L, et al. Clinical experi-
ence with minocycline and rifampin-impregnated central
venous catheters in bone marrow transplantation recipi-
ents: efficacy and low risk of developing staphylococcal re-
sistance. Infect Control Hosp Epidemiol 2003;24:961e963.
54. Leon C, Ruiz-Santana S, Rello J, et al. Benefits of minocy-
cline and rifampin-impregnated central venous catheters.
A prospective, randomized, double-blind, controlled, multi-
center trial. Intensive Care Med 2004;30:1891e1899.
55. Sampath LA, Tambe SM, Modak SM. In vitro and in vivo
efficacy of catheters impregnated with antiseptics or
antibiotics: evaluation of the risk of bacterial resistance
to the antimicrobials in the catheters. Infect Control Hosp
Epidemiol 2001;22:640e646.
Journal of Hospital Infection (2009) 73, 331e337
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Noroviruses in healthcare settings: a challenging

problem
M. Koopmans a,b,*
a
Erasmus Medical Centre, Rotterdam, The Netherlands
b
Laboratory for Infectious Diseases, National Institute of Public Health and the Environment, Bilthoven,
The Netherlands

Available online 22 September 2009

KEYWORDS Summary Current knowledge about noroviruses in relation to healthcare-

Healthcare-associated associated infections (HCAIs) merely scratches the surface. Most data come
infections; from outbreak-based studies, which represent only a small piece of the puz-
Norovirus GII4;
zle. Nevertheless, the data available show that the clinical impact of noro-
Norovirus outbreak
viruses is particularly severe in outbreaks in healthcare settings, and that it
may be increasing. Coupled with the projected increases of the population
aged >65 years, especially those needing healthcare, it is timely to consider
noroviruses in discussions around HCAIs. In particular, broadening the scope
from a field mostly discussing bacteriological problems and a more holistic
approach to dealing with HCAIs may be needed to avoid introducing new
risks when trying to deal with the antimicrobial resistance problem.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

Introduction investigator from the National Institutes of Health

One of the diseases that has been making headlines
in the popular media in recent years is the condition
popularly known as ‘gastric flu’, caused by noro-
viruses. In contrast to what the recent attention
would suggest, a clinical syndrome known as ‘winter
vomiting disease’ caused by these viruses was
described as early as 1928. Forty years later, an
* Corresponding author. Address: Laboratory for Infectious
Diseases, National Institute of Public Health and the Environ-
ment, Antonie van Leeuwenhoeklaan 9, 3720BA Bilthoven,
The Netherlands. Tel.: þ31 30 2743945; fax: þ31 30 2744418.
E-mail address: marion.koopmans@rivm.nl
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.06.028
discovered small round particles in stool samples
from children involved in an outbreak in a school
in Norwalk, Ohio, USA, now known to be the aetio-
logical agent.1 To date, the Norwalk virus is the proto-
type of what has turned out to be a very diverse
family of viruses infecting humans and animals.
Properties of noroviruses and
epidemiology
Many of those active in the healthcare setting have
experienced one or more norovirus outbreaks and
the challenges they pose to infection control.
332
These can be better understood when looking at
some of the basic properties of noroviruses. The
typical person can become infected by ingesting
only a few virus particles, but then will start to
shed high loads of viruses through stool and vomit.2
Viral loads >109 particles/mL of stool have been
found in the first days following infection.3e5
There is some evidence that higher viral loads
are associated with more severe symptoms, but
other studies have not confirmed this and found
similar levels of shedding in persons with asymp-
tomatic norovirus infection.3,4,6 Vomiting can be
abrupt and violent (projectile vomiting) which
contributes to contamination of the environ-
ment.7,8 Virus shedding gradually decreases over
a period of several weeks, long after people have
recovered from illness, but both infection and
shedding may become chronic (>1 month) in per-
sons with comorbidities.5,9e11 Once in the environ-
ment, noroviruses can remain infectious for a long
time, depending on the specific conditions and
cleaning and disinfection procedures used.
Population-based studies have demonstrated
that viral pathogens e particularly rotavirus, noro-
virus and sapovirus e are by far the most com-
monly identified causes of gastroenteritis in the
community (Figure 1).12,13 This picture changes
hospital
Staff, visitor, patients
evolution Patients
staff
visitors
Day care center,
school
community
Virus evolution
Restaurant,
caterer, Multiple infections,
retail recombination
Introductions via travel, food
Figure 1 Epidemiology of noroviruses.
M. Koopmans
when looking at aetiology of illnesses for which
a physician is consulted, when the aetiological
fraction of bacterial pathogens is higher, reflecting
more severe symptoms on average.14 Recent stud-
ies have shown that clinical impact has been
underestimated, with reports of severe disease
and mortality associated with norovirus outbreaks
in risk groups.9,11,15,16 All of these factors merge
when outbreaks develop in healthcare settings.
It is important to realise that most of what is
currently known comes from studies of outbreaks
of gastroenteritis caused by noroviruses, whereas
data on clinical impact remain sparse.17 This is
particularly true for data on the role of noroviruses
as causes of childhood diarrhoea in developing
countries. Diarrhoeal disease remains a major
cause of childhood mortality in these regions,
and viruses (notably rotaviruses) play an important
role. Whether the same applies for noroviruses
remains to be seen, and this topic has been identi-
fied as a major data gap by an expert committee
commissioned by the World Health Organization
(WHO) to discuss viruses in relation to food safety.
In parts of the world with high standards of
living, the greatest impact is seen in the outbreaks
caused by noroviruses (Figure 1). As norovirus
infection is not a notifiable disease in most
Outbreaks, mild illness
Outbreaks, severe illness
Community, mild illness
Virus transmissions
Elderly Epidemic/
Home pandemic spread
Novel strains
Noroviruses in healthcare settings
countries, it is difficult to extrapolate trends from
available surveillance data which are mostly based
on passive reporting of outbreaks with clear differ-
ences in the set-up in different countries.18 Never-
theless, the combined analysis of epidemiological
and molecular data, as has been done by
a Europe-wide collaborative research network
(the Foodborne Viruses in Europe network;
www.fbve.nl), has yielded some interesting new
findings.19 In past years, outbreaks frequently
have been reported from healthcare institutions,
both hospitals and homes for elderly (Figure 2).
A sharp increase was witnessed in 2002, co-
inciding with emergence of a specific genogroup II
genotype 4 strain (GII4). This increase was real,
based on data from countries that already had
stable reporting for several years beforehand.20e23
Noroviruses can be transmitted by direct per-
son-to-person contact with persons shedding
viruses. Since the infection route is faecaleoral,
personal hand-washing hygiene is an important
determinant of risk of spreading e as demon-
strated, for example, from analysis of outbreaks
on cruise ships.24 Pre- and post-symptomatic shed-
ding may lead to outbreaks but it is unclear how
relevant this mode of transmission really is.25e28
Outbreak control protocols for workers in health-
care or food industry usually recommend sick leave
until at least 48 h after resolution of symptoms,
under the assumption that people without symp-
toms are less likely to be spreaders. Whether this
assumption is correct is unknown. An extension
of this mode of spread is contamination by food
handlers preparing or serving food. At least 20%
of outbreaks reported in the FBVE network are
thought to be food-related, many of them linked
to food handlers.29 More difficult to detect are
diffuse outbreaks that may occur when a food is
contaminated higher up in the production chain,
for instance with oysters and mussels or berries.
Here, contamination with sewage of water used
in production leads to presence of multiple
Elderly homes
Hospitals
Daycare
Hotel/catering
Other
Figure 2 Most commonly reported settings for noro-
virus outbreaks. Source: the Foodborne Viruses in
Europe network (1112 outbreaks).19
333
pathogens. With the globalisation of food produc-
tion and marketing, outbreaks linked to such prod-
ucts may be very difficult to detect and verify.
Nevertheless, studies have shown that this is not
uncommon, and recently the global organisation
that provides guidance for food industry, the Co-
dex Alimentarius (www.codexalimentarius.net)
has decided to commission the production of
guidelines for control of virus infection via food.
Molecular epidemiology
It is important to realise that noroviruses consti-
tute a widely diverse group. They belong to the
Caliciviridae family, which segregates into five
genera containing human and animal pathogens
(Figure 3).30 The genus norovirus consists of five
main genogroups, three of which have been found
to infect humans. Each genogroup can further be
segregated into genotypes. The clinical relevance
of this is that population immunity does not even
protect from infection with related strains, and
infections remain possible. This, coupled with
relatively poor immunogenicity of noroviruses,
probably explains why norovirus illness is observed
in people of all age groups. The success of the GII4
strains was explained by additional analyses show-
ing that they evolve rapidly by mutations in the
parts of the viral genome that encode the most
surface-exposed parts.31e33 By doing so, these vi-
ruses change their antigenic make-up and possibly
also the way they bind to the human host
(Figure 1). This finding resembles what is known
about influenza A viruses, where human vaccines
are updated frequently to match the immunity
needed to protect from illness due to circulating
strains. Like influenza, the successful norovirus
strains have been found across the world, spread-
ing rapidly once they emerge in one continent. It
is unclear what drives this evolutionary process
but immune-driven selection of strains is the
most plausible explanation, somewhat contrasting
the prior notion of poor immunogenicity of norovi-
ruses. Detailed knowledge of the molecular diver-
sity in the norovirus genus is required to allow the
use of molecular typing in support of investigations
(Figure 3).
Clinical impact
The disease caused by noroviruses is popularly
known as ‘gastric flu’. Typically symptoms develop
abruptly with acute onset of vomiting, and some-
times fever as soon as 8 h after exposure. Vomiting
334
Example
Family Caliciviridae
Genus(5) Norovirus
Species / genogroup (5) GII
Subtype/ genotype (>40) GGIL4
Lineage / variant (only in GII4?) GGIL4 2002
Outbreak/ patient strain Hu/NV/Amsterdaml/2002/NL
Figure 3 Taxonomic structure of noroviruses within the family Caliciviridae, and possible questions to address using
genetic typing at different taxonomic levels.
may be violent, and is seen as one of the effective
ways of spreading viruses into/onto the envir-
onment of a patient. Studies in which environmen-
tal surfaces were tested for viral RNA by using
reverse transcriptionepolymerase chain reaction
tests on cotton swabs have confirmed widespread
contamination.7,8 Even if it is not possible to test
for infectivity, presence of viral RNA does demon-
strate that the environment has been contami-
nated with faeces or vomit. The consequences of
this may be protracted outbreaks such as those
demonstrated on some cruise ships.34 Here, new
groups of passengers have become infected re-
peatedly following an initial outbreak.
Norovirus patients usually have vomiting for 1e2
days on average, but diarrhoea may last longer,
especially in children. In a community-based study
of acute gastroenteritis in The Netherlands, about
25% of all cases had symptoms lasting a week or
more.35 Longer duration of illness has been re-
ported for young children and elderly.10,36 Several
studies have shown that symptoms may be more
severe in patients with underlying illnesses, based
on outbreaks in hospitals.9,11,36,37 A recent finding
is that chronic infection with norovirus is much
more common than previously recognised. Chronic
diarrhoea and shedding has been observed in im-
munocompromised individuals, but it is unknown
how common this is, or if it is associated with
specific deficiencies in the immune system. Non-
specific symptoms such as vertigo, thirst, and
headaches after resolution of vomiting and diar-
rhoea may delay full recovery in elderly.38,39 No
other extraintestinal manifestations of norovirus
M. Koopmans
Relevance of typing Routine
Identification of new viruses No
Identification of new viruses No
Potential for zoonotic infection No
Monitoring of trends, surveillance, Some countries
international comparison
Surveillance, new epidemics, Some countries
virulence, international comparison
Hospital epidemiology, source Rare
tracing
infection have been reported, but clinical research
into these pathogens is limited. With the increased
media attention for outbreaks following the epi-
demics with GII4 strains, the number of hospital-
based laboratories offering services for norovirus
diagnostics is rapidly increasing, and new data on
clinical impact of noroviruses can be expected in
the coming years.17 Carefully designed studies
are needed because noroviruses are such a diverse
group of pathogens, and viruses belonging to dif-
ferent genotypes may behave differently. For in-
stance, GII4 strains are more often found as
causes of outbreaks in healthcare settings than
non-GII4 noroviruses, whereas the genogroup I
strains are more common in outbreaks related to
food or environmental contamination, pointing to
possible differences in virus behaviour that may
affect their epidemiology.19 In waste water, geno-
group I strains are found almost as abundantly as
genogroup II viruses, whereas the latter are much
more frequently found as causes of human disease.
This indicates possible differences in virulence of
viruses belonging to different clades.40 At present,
research addressing clinical impact in healthcare
settings should focus on GII noroviruses, particu-
larly the recombinant strains that have emerged
as the most common norovirus in childhood diar-
rhoea, recognised by a combination of genes la-
belled IIb pol/II3 capsid (IIbp/II3c).41 Remarkably,
IIbp/II3c strains were first recognised following
a large international foodborne outbreak in
2001.42 The capsid gene sequences matching the
IIb polymerase gene have never been identified
from human surveillance, raising questions about
Noroviruses in healthcare settings
its origin. Genetic recombination is common,
occurring when viruses belonging to the same
genogroup exchange parts of their genes when
a host cell is infected simultaneously with two
parent viruses (Figure 1). Noroviruses clustering
within genogroup II have been identified in pigs,
and pigs can be infected with human noroviruses,
raising speculation about their potential as reser-
voir for new viruses, or for viruses resulting from
recombination between human and pig norovirus
strains (Figure 3).43e45 At present, there is no evi-
dence that recombination occurs between human
and animal strains, but monitoring of trends of
human illness should include characterisation of
a systematic sample of strains to increase our
understanding of the complex epidemiology of
these viruses. An additional reason for doing so
comes from studying related animal viruses
belonging to the same family: there are examples
of highly successful species jumps within the vesi-
virus genus and the lagovirus genus, demonstrating
that caliciviruses are quite flexible.46 Therefore,
from a virological perspective it is prudent to
monitor what is going on.
Specific issues related to norovirus in
healthcare settings
The above shows that there are three important
reasons for wanting to control norovirus outbreaks
in healthcare settings. First, nosocomial norovirus
infection may prolong the duration of hospital-
isation and cause complications in vulnerable per-
sons, which should be avoided. Second, control of
outbreaks may be costly when not implemented
immediately, not only in terms of added cost for
cleaning and disinfection, but also in terms of
absenteeism of staff, longer hospitalisation of
patients, long-term treatment of patients that
develop chronic disease, closure of wards including
intensive care units, and postponement of surgery.47,48
The third reason is that in patients with chronic in-
fection the viruses can evolve by accumulation of
mutations.11 Besides the health problems for the
individual patient, these persons therefore may
eventually be the reservoir in which new variants
develop, thus perpetuating the cycle of infection
(Figure 1). Although this is speculative at this stage,
the size of the group that is at risk for chronic dis-
ease due to this ‘simple’ diarrhoea virus is large
and expanding. A five-year survey in which patients
diagnosed with norovirus infection were tested
with at least a one-month interval found chronic in-
fection in 22% of all patients in this group, which
was 6% of all patients diagnosed with norovirus
335
throughout that time period.41 These numbers
may be biased, however, and need to be evaluated
in other studies. Nevertheless, low estimates of the
proportion of the population with impaired immu-
nity are 1e2%, and knowing that one in four persons
is likely to contract a norovirus infection each year
suggests that there may be numerous patients with
chronic infection. An important question to answer
is what are the best ways of controlling the disease
in healthcare settings. Clearly, we are faced with
multiple possible sources of virus triggering out-
breaks. These could be any of the patients, staff
or visitors walking into the hospital given the high
incidence of norovirus in the community.49 Addi-
tional sources may be long-stay patients in the hos-
pital, or food brought in by visitors. Understanding
the details of norovirus epidemiology will help to
design more evidence-based strategies to reduce
their impact.
Issues around infection control
Because norovirus infection is so common, pre-
vention of introduction of the virus in a healthcare
setting is an unrealistic goal. One could speculate
about the possibility of screening patients at
admission, but that requires rapid tests with very
high sensitivity to be effective. A recent study
comparing infection control protocols of different
level of stringency was inconclusive due to poor
compliance to a particular protocol. The only
measure that seemed to have an effect was the
speed of starting any protocol, i.e. taking di-
arrhoea and vomiting in one or two patients as
a trigger for enhanced hygienic measures.50
Another issue of some debate among virologists
is whether alcohol-based hand hygiene products
have sufficient activity against these viruses.51 The
problem is that noroviruses cannot be grown in cell
culture, so that experimental inactivation studies
are not possible.52 Model viruses have been used
but their validity has been disputed because
specific characteristics can differ substantially
between highly related viruses. The WHO recom-
mends the use of alcohol-based hand sanitation
globally, because of its advantage over no hand
sanitation at all, e.g. in areas where access to
clean water is limited.53 Data on virus inactivation
for the diarrhoeal viruses where available shows
some effect on virus infectivity and thus the use
of alcohol-based hand hygiene products is likely
to have some advantage for these viruses as
well.54 Virologists, however, do recommend the
use of regular hand washing with water and soap
during outbreak situations.
336
In conclusion, the recent increase in incidence
of norovirus outbreaks has triggered studies show-
ing that the clinical impact of noroviruses is
greater than previously recognised. A clear picture
emerges: outbreaks of norovirus illness in health-
care settings may be disruptive and costly, and can
cause significant complications including chronic
illness and deaths. This should be sufficient moti-
vation for swift action in the face of new out-
breaks, using stringent infection control measures.
Acknowledgements
I am greatly indebted to my coworkers at the RIVM,
Erasmus MC, and the FBVE network for the contin-
uous efforts to advance our understanding of
norovirus.
References
1. Kapikian AZ. The discovery of the 27 nm Norwalk virus: an
historic perspective. J Infect Dis 2000;181:S295eS302.
2. Teunis PF, Moe CL, Liu P, et al. Norwalk virus: how infec-
tious is it? J Med Virol 2008;80:1468e1476.
3. Lee N, Chan MC, Wong B, et al. Fecal viral concentration
and diarrhea in norovirus gastroenteritis. Emerg Infect Dis
2007;13:1399e1401.
4. Atmar RL, Opekun AR, Gilger MA, et al. Norwalk virus shed-
ding after experimental human infection. Emerg Infect Dis
2008;14:1553e1557.
5. Tu ET, Bull RA, Kim MJ, et al. Norovirus excretion in an
aged-care setting. J Clin Microbiol 2008;46:2119e2121.
6. Ozawa K, Oka T, Takeda N, Hansman GS. Norovirus infection
in symptomatic and asymptomatic food handlers in Japan.
J Clin Microbiol 2007;45:3996e4005.
7. Wu HM, Fornek M, Schwab KJ, et al. A norovirus outbreak at
a long term care facility: the role of environmental surface
contamination. Infect Control Hosp Epidemiol 2005;26:
802e810.
8. Boxman IL, Dijkman R, te Loeke NA, et al. Environmental
swabs as a tool in norovirus outbreak investigation, including
outbreaks on cruise ships. J Food Prot 2009;72:111e119.
9. Ludwig A, Adams O, Laws HJ, Schroten H, Tenenbaum T.
Quantitative detection of norovirus excretion in pediat-
ric patients with cancer and prolonged gastroenteritis
and shedding of norovirus. J Med Virol 2008;80:
1461e1467.
10. Murata T, Katsushima N, Mizuta K, Muraki Y, Hongo S,
Matsuzaki Y. Prolonged norovirus shedding in infants 6
months of age with gastroenteritis. Pediatr Infect Dis
J 2007;26:46e49.
11. Siebenga JJ, Beersma MF, Vennema H, van Biezen P,
Hartwig NJ, Koopmans M. High prevalence of prolonged
norovirus shedding and illness among hospitalized
patients: a model for in vivo molecular evolution. J Infect
Dis 2008;198:994e1001. Erratum in: J Infect Dis 2008;
198:1575.
12. de Wit MA, Koopmans MP, Kortbeek LM, et al. Sensor, a pop-
ulation-based cohort study on gastroenteritis in the Nether-
lands: incidence and etiology. Am J Epidemiol 2001;154:
666e674.
M. Koopmans
13. Wheeler JG, Sethi D, Cowden JM, et al. Study of infectious
intestinal disease in England: rates in the community,
presenting to general practice, and reported to national
surveillance. The Infectious Intestinal Disease Study Exe-
cutive. Br Med J 1999;318:1046e1050.
14. de Wit MA, Kortbeek LM, Koopmans MP, et al. A comparison of
gastroenteritis in a general practice-based study and a com-
munity-based study. Epidemiol Infect 2001;127:389e397.
15. Jansen A, Stark K, Kunkel J, et al. Aetiology of community-
acquired, acute gastroenteritis in hospitalised adults:
a prospective cohort study. BMC Infect Dis 2008;8:143.
16. Harris JP, Edmunds WJ, Pebody R, Brown DW, Lopman BA.
Deaths from norovirus among the elderly, England and
Wales. Emerg Infect Dis 2008;14:1546e1552.
17. Koopmans M. Progress in understanding norovirus epidemi-
ology. Curr Opin Infect Dis 2008;21:544e552.
18. Kroneman A, Harris J, Vennema H, et al. Data quality of
5 years of central norovirus outbreak reporting in the Euro-
pean Network for food-borne viruses. J Public Health (Oxf)
2008;30:82e90.
19. Kroneman A, Verhoef L, Harris J, et al. Analysis of inte-
grated virological and epidemiological reports of norovirus
outbreaks collected within the foodborne viruses in Europe
Network from 1 July 2001 to 30 June 2006. J Clin Microbiol
2008;46:2959e2965.
20. Bruggink L, Marshall J. Molecular changes in the norovirus
polymerase gene and their association with incidence of
GII.4 norovirus-associated gastroenteritis outbreaks in Vic-
toria, Australia, 2001e2005. Arch Virol 2008;153:729e732.
21. Widdowson MA, Cramer EH, Hadley L, et al. Outbreaks of
acute gastroenteritis on cruise ships and on land: identifica-
tion of a predominant circulating strain of norovirus e
United States, 2002. J Infect Dis 2004;190:27e36.
22. Lopman B, Vennema H, Kohli E, et al. Increase in viral gas-
troenteritis outbreaks in Europe and epidemic spread of
new norovirus variant. Lancet 2004;363:682e688.
23. Siebenga JJ, Vennema H, Duizer E, Koopmans MP. Gastroen-
teritis caused by norovirus GGII.4, The Netherlands,
1994e2005. Emerg Infect Dis 2007;13:144e146.
24. Chimonas MA, Vaughan GH, Andre Z, et al. Passenger be-
havior associated with norovirus infection on board a cruise
ship e Alaska. J Travel Med 2008;15:177e183.
25. Lo SV, Connolly AM, Palmer SR, Wright D, Thomas PD,
Joynson D. The role of the pre-symptomatic food handler
in a common source outbreak of food-borne SRSV gastroen-
teritis in a group of hospitals. Epidemiol Infect 1994;113:
513e521.
26. Parashar UD, Dow L, Fankhauser RL, et al. An outbreak of
viral gastroenteritis associated with consumption of sand-
wiches: implications for the control of transmission by food
handlers. Epidemiol Infect 1998;121:615e621.
27. Götz H, de JB, Lindbäck J, et al. Epidemiological investiga-
tion of a food-borne gastroenteritis outbreak caused by
Norwalk-like virus in 30 day-care centres. Scand J Infect
Dis 2002;34:115e121.
28. Patterson T, Hutchings P, Palmer S. Outbreak of SRSV gas-
troenteritis at an international conference traced to food
handled by a post-symptomatic caterer. Epidemiol Infect
1993;111:157e162.
29. Verhoef LP, Kroneman A, van Duynhoven Y, Boshuizen H,
van Pelt W, Koopmans M. Foodborne viruses in Europe Net-
work. Selection tool for foodborne norovirus outbreaks.
Emerg Infect Dis 2009;15:31e38.
30. Green KY, Chanock RM, Kapikian AZ. Human caliciviruses.
In: Knipe DM, Howley PM, editors. Fields’ virology. 4th
edn. Philadelphia: Lippincott Williams & Wilkins; 2001. p.
841e874.
Noroviruses in healthcare settings
31. Lindesmith LC, Donaldson EF, Lobue AD, et al. Mechanisms
of GII.4 norovirus persistence in human populations. PLoS
Med 2008;5:e31.
32. Siebenga JJ, Vennema H, Renckens B, et al. Epochal evolu-
tion of GGII.4 norovirus capsid proteins from 1995 to 2006.
J Virol 2007;81:9932e9941.
33. Allen DJ, Gray JJ, Gallimore CI, Xerry J, Iturriza-Gómara M.
Analysis of amino acid variation in the P2 domain of the
GII-4 norovirus VP1 protein reveals putative variant-specific
epitopes. PLoS ONE 2008;3:e1485.
34. Verhoef L, Depoortere E, Boxman I, et al. Emergence of
new norovirus variants on spring cruise ships and prediction
of winter epidemics. Emerg Infect Dis 2008;14:238e243.
35. Rockx B, De Wit M, Vennema H, et al. Natural history of
human calicivirus infection: a prospective cohort study. Clin
Infect Dis 2002;35:246e253.
36. Lopman BA, Reacher MH, Vipond IB, Sarangi J, Brown DW.
Clinical manifestations of norovirus in health care settings.
Clin Infect Dis 2004;39:318e324.
37. Mattner F, Sohr D, Heim A, Gastmeier P, Vennema H,
Koopmans M. Risk groups for clinical complications of noro-
virus infections: an outbreak investigation. Clin Microbiol
Infect 2006;12:69e74.
38. Tsang OT, Wong AT, Chow CB, Yung RW, Lim WW, Liu SH.
Clinical characteristics of nosocomial norovirus in Hong
Kong. J Hosp Infect 2008;69:135e140.
39. Goller JL, Dimitriadis A, Tan A, Kelly H, Marshall JA. Long
term features of norovirus gastroenteritis in the elderly.
J Hosp Infect 2004;58:286e291.
40. da Silva AK, Le Saux JC, Parnaudeau S, Pommepuy M,
Elimelech M, Le Guyader FS. Evaluation of removal of noro-
viruses during waste water treatment, using RTePCR: dif-
ferent behaviors of GI and GII. Appl Environ Microbiol
2007;73:7891e7897.
41. Beersma MF, Schutten M, Vennema H, et al. Norovirus in
a Dutch tertiary care hospital (2002e2007): frequent noso-
comial transmission and dominance of GIIb strains in young
children. J Hosp Infect 2009;71:199e205.
42. Koopmans M, Vennema H, Heersma H, et al. Early identifi-
cation of common-source foodborne virus outbreaks in
Europe. Emerg Infect Dis 2003;9:1136e1142.
337
43. Sugieda M, Nakajima S. Viruses detected in the caecum con-
tents of healthy pigs representing a new genetic cluster in
genogroup II of the genus ‘‘Norwalk-like viruses’’. Virus
Res 2002;87:165e172.
44. Wang QH, Han MG, Cheetham S, Souza M, Funk JA, Saif LJ.
Porcine noroviruses related to human noroviruses. Emerg
Infect Dis 2005;11:1874e1881.
45. Cheetham S, Souza M, Meulia T, Grimes S, Han MG, Saif LJ.
Pathogenesis of a genogroup II human norovirus in gnoto-
biotic pigs. J Virol 2006;80:10372e10381.
46. Smith AW, Skilling DE, Cherry N, Mead JH, Matson DO. Cal-
icivirus emergence from ocean reservoirs: zoonotic and
interspecies movements. Emerg Infect Dis 1998;4:13e20.
47. Hansen S, Stamm-Balderjahn S, Zuschneid I, et al. Closure
of medical departments during nosocomial outbreaks: data
from a systematic analysis of the literature. J Hosp Infect
2007;65:348e353.
48. Lopman BA, Reacher MH, Vipond IB, et al. Epidemiology and
cost of nosocomial gastroenteritis, Avon, England,
2002e2003. Emerg Infect Dis 2004;10:1827e1834.
49. Mattner F, Mattner L, Borck HU, Gastmeier P. Evaluation of
the impact of the source (patient versus staff) on nosoco-
mial norovirus outbreak severity. Infect Control Hosp
Epidemiol 2005;26:268e272.
50. Friesema IHM, Vennema H, Heijne JCM, et al. Norovirus
outbreaks in nursing homes. The evaluation of infection control
measures. Epidemiol. Infect 2009 (e-pub ahead of print).
51. Lages SL, Ramakrishnan MA, Goyal SM. In vivo efficacy of
hand sanitisers against feline calicivirus: a surrogate for
norovirus. J Hosp Infect 2008;68:159e163.
52. Duizer E, Schwab KJ, Neill FH, Atmar RL, Koopmans MP,
Estes MK. Laboratory efforts to cultivate noroviruses.
J Gen Virol 2004;85:79e87.
53. World Alliance for Patient Safety. WHO guidelines on hand
hygiene in health care (advanced draft). Geneva: WHO;
2006.
54. Sandora TJ, Shih MC, Goldmann DA. Reducing absenteeism
from gastrointestinal and respiratory illness in elementary
school students: a randomized, controlled trial of an
infection-control intervention. Pediatrics 2008;121:
e1555ee1562.
Journal of Hospital Infection (2009) 73, 338e344
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Pseudomonas aeruginosa: a formidable and

ever-present adversary
K.G. Kerr a,*, A.M. Snelling b
a
Department of Microbiology, Harrogate District Hospital, Harrogate, North Yorkshire, UK
b
Division of Biomedical Sciences & Bradford Infection Group, University of Bradford, Bradford,
West Yorkshire, UK

Available online 21 August 2009

KEYWORDS Summary Pseudomonas aeruginosa is a versatile pathogen associated

Healthcare-associated with a broad spectrum of infections in humans. In healthcare settings
infection; the bacterium is an important cause of infection in vulnerable individuals
Hospital environment;
including those with burns or neutropenia or receiving intensive care. In
Infection control;
Pseudomonas
these groups morbidity and mortality attributable to P. aeruginosa infec-
aeruginosa tion can be high. Management of infections is difficult as P. aeruginosa is
inherently resistant to many antimicrobials. Furthermore, treatment is be-
ing rendered increasingly problematic due to the emergence and spread of
resistance to the few agents that remain as therapeutic options. A notable
recent development is the acquisition of carbapenemases by some strains
of P. aeruginosa. Given these challenges, it would seem reasonable to iden-
tify strategies that would prevent acquisition of the bacterium by hospital-
ised patients. Environmental reservoirs of P. aeruginosa are readily
identifiable, and there are numerous reports of outbreaks that have been
attributed to an environmental source; however, the role of such sources in
sporadic pseudomonal infection is less well understood. Nevertheless there
is emerging evidence from prospective studies to suggest that environmen-
tal sources, especially water, may have significance in the epidemiology of
sporadic P. aeruginosa infections in hospital settings, including intensive
care units. A better understanding of the role of environmental reservoirs
in pseudomonal infection will permit the development of new strategies
and refinement of existing approaches to interrupt transmission from these
sources to patients.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

* Corresponding author. Address: Department of Microbiology, Harrogate District Hospital, Lancaster Park Road, Harrogate, North
Yorkshire HG2 7SX, UK. Tel.: þ44 1423 553077.
E-mail address: kevin.kerr@hdft.nhs.uk

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.04.020
Prevention and control of P. aeruginosa
Introduction
Pseudomonas aeruginosa is associated with an
ever-widening spectrum of infections. Some mani-
festations of P. aeruginosa infection occur pre-
dominantly in the community although the
majority are seen in healthcare settings especially
in immunosuppressed individuals and other highly
vulnerable patients such as those in the intensive
care unit (ICU). Infections may be associated
with significant morbidity and mortality and thera-
peutic options are becoming increasingly limited
with the continued emergence and spread of anti-
microbial resistant strains. Given the potential
severity of specific P. aeruginosa infections and
problems in selecting optimal therapy for multi-
drug resistant (MDR) isolates, identification and
implementation of effective strategies to prevent
these infections are urgent priorities.
Clinical features of P. aeruginosa
infection
There can be no doubt about the versatility of
P. aeruginosa as a pathogen. The list of infections
associated with this bacterium is long and continues
to grow. Some infections are more likely to be en-
countered in community settings. These include
folliculitis and the green nail or hot foot syndromes
following recreational exposure to water sources
such as hot tubs, Jacuzzis and swimming pools,
septic arthritis (especially of smaller joints) in
injecting drug users and otitis externa. The last
may manifest as severe, potentially life-threaten-
ing malignant otitis externa in patients with
diabetes and immunocompromised individuals.
Underlying pulmonary disease, such as bronchi-
ectasis and cystic fibrosis, is a predisposing factor
for P. aeruginosa respiratory tract infections. Most
serious P. aeruginosa infections are seen in health-
care settings and include bacteraemia, pneumonia,
urosepsis and wound infection including secondary
infection of burns. Some patient groups are more
at risk of acquiring infection than others. These
include immunocompromised patients, especially
those with chemotherapy-induced neutropenia
who are at particular risk of pseudomonal blood-
stream infections; hence empiric antimicrobial
therapy for febrile neutropenia should include at
least one agent with anti-pseudomonal activity.
Physical breaches in host defences such as those
which follow surgical incisions, insertion of urinary
and vascular catheters and endotracheal tubes also
predispose patients to P. aeruginosa infection.
339
Patients in critical care units are particularly
at risk of these infections, partly due to the
presence of invasive devices and partly because of
underlying disease and comorbidities. The mortality
attributable to P. aeruginosa infection such as
ventilator-associated pneumonia or bacteraemia
is substantial, especially for patients who receive
inadequate empirical therapy.1e3 Morbidity also
may be considerable. For example, post-surgical
endophthalmitis is a feared complication which
may, even if treated promptly, result in significant
visual impairment or even total loss of sight.4
As well as their undoubted clinical importance,
P. aeruginosa infections are also quantitatively sig-
nificant as has been noted in numerous reports.
Recently, the US National Healthcare Safety Net-
work (which subsumes the National Nosocomial
Infection Surveillance system) published data on
>28 000 cases of nosocomial infections from 463
hospitals gathered over a 22 month period.5 Over-
all, P. aeruginosa was the sixth most frequently
occurring pathogen, the second commonest cause
of ventilator-associated pneumonia and the
seventh commonest cause of catheter-related
bloodstream infection.
Virulence factors of P. aeruginosa
An impressive arsenal of virulence factors can be
deployed by P. aeruginosa. This extensive armoury
partly explains why the range of infections caused
by the bacterium is so wide. A full discussion of
established and putative virulence factors is beyond
the scope of this paper, but they include enzymes
such as proteases and elastases and phenazine pig-
ments, such as pyocyanin and rhamolipids. The bac-
terium possesses several different export systems
that are involved in the secretion of virulence fac-
tors. Of particular importance is a type III secretion
system which enables injection of effector proteins
directly into the cytoplasm of host cells.6 Four ef-
fector proteins have been identified: ExoS and
ExoT, which are bifunctional enzymes with GTPase
activating protein and ADP ribosylase activities
(which ultimately cause disruption of the host cell
cytoskeleton) and ExoU (a cytotoxin with phospho-
lipase activity) and Exo Y (an adenylate cyclase).
Different combinations of these proteins have
profound and often devastating effects on epi-
thelial barrier function and wound healing. It is
also thought that they interfere with the function-
ing of macrophages and neutrophils.
The ability of P. aeruginosa to form biofilm is
important for the ability of the bacterium to per-
sist in environmental niches such as pipes and
340
taps.7 When growing as a complex mass of cells at-
tached to a surface, P. aeruginosa cells can be sig-
nificantly more resistant to biocides than when
they are in a planktonic (free-floating) state.8
This may have significant consequences for hospi-
tal cleaning regimens.
Biofilms can also occur inside the host and are
associated with the pathogenesis of chronic lung
infections in patients with cystic fibrosis and
bronchiectasis as well as infections associated
with indwelling devices and other prosthetic
material.
Antimicrobial therapy of P. aeruginosa
infection
Pseudomonas aeruginosa is intrinsically resistant to
several classes of antibiotics, thus limiting
therapeutic options. Unfortunately, antimicrobial
therapy is becoming even more problematic due
to acquired or mutational resistance. A wide range
of resistance mechanisms have been identified
including multi-drug efflux pumps (which can confer
resistance to, among others, cephalosporins,
ureidopenicillins, fluoroquinolones and aminogly-
cosides), aminoglycoside-modifying enzymes,
b-lactamases, and target site modifications, e.g.
of DNA gyrase and topoisomerase IV following
mutations in gyrA and parC, respectively.9 Resis-
tance to a particular class may be mediated by sev-
eral mechanisms. For example, isolates may lose
susceptibility to a carbapenem due to production
of a carbapenemase, loss of the OprD porin, or via
a multidrug efflux pump. A number of recently pub-
lished studies have highlighted the growing problem
of antimicrobial resistance in P. aeruginosa, and in
particular the emergence of MDR (resistance to 3
classes), extensively drug-resistant (resistance to
all but one or two classes) or even pandrug-resistant
strains. Souli et al. recently published data from
23 countries on the European Antimicrobial
Resistance Surveillance System (EARSS) website.10
Resistance rates for aminoglycosides, carbape-
nems, quinolones and ceftazidime were 0e51.9%;
9e50.5%; 7.2e51.9% and 4e48.5%, respectively.
Overall, 18% of isolates were recorded as being
MDR.
As noted earlier, patients in intensive care are
especially vulnerable to P. aeruginosa infection
and reports of increasing resistance in this setting
give particular cause for concern.11e14 Given the
frequent use of antibiotics in this setting, high
resistance rates in ICU-based studies are not unex-
pected but Handberger et al. report that half of
the units (10/20) which responded to the survey
K.G. Kerr, A.M. Snelling
did not have an antibiotic policy and that 43%
(6/14) did not routinely isolate patients colonised
or infected with MDR strains of P. aeruginosa.11
Perhaps the most concerning development in
recent years has been the emergence of carbapen-
emases in MDR strains of P. aeruginosa.15 Loss of
the carbapenems, which have been the mainstay
of therapy for infections caused by MDR strains,
severely limits therapeutic options, with colistin
becoming the ‘antibiotic of last resort’. The
emergence of colistin resistance in carbapenem-
resistant P. aeruginosa, creating effectively pan-
resistant strains, is thus a serious development.
This is highlighted by an outbreak of pan-resistant
P. aeruginosa in the ICU of a cancer centre in
Bratislava. Five of six patients with P. aeruginosa
(colistin minimum inhibitory concentration:
4 mg/L) bacteraemia died.16
Pseudomonas aeruginosa in the
hospital environment
Pseudomonas aeruginosa is capable of colonising
a range of niches such as the phyllosphere and rhi-
zosphere of plants, soil and a wide range of
aquatic habitats. Although the bacterium is gener-
ally not considered a component of the normal re-
gional flora of humans, faecal and e to a lesser
extent e oropharyngeal and skin carriage is not
uncommon.
The range of reservoirs in healthcare environ-
ments from which P. aeruginosa has been isolated
is wide ranging, with examples listed in Table I.
Sites most likely to harbour the bacterium are wa-
ter-related (e.g. taps and showers) or where mois-
ture or humidity is high (e.g. respiratory therapy
equipment). Several properties of P. aeruginosa
favour persistence in the hospital environment.
The bacterium is inherently resistant to several
disinfectants such as biguanides and quaternary
ammonium compounds through the action of
multidrug efflux pumps. Furthermore, the ability
to form biofilm on a range of inanimate surfaces
also contributes to disinfectant resistance as well
as impeding physical removal. Persistence in moist
environments is also favoured by the use of the
type III secretion system to kill free-living amoeba
which graze on environmental bacteria.17
The true importance of P. aeruginosa in the
hospital environment as it relates to healthcare-
associated infection (HCAI) remains a controversial
issue. Even when strains indistinguishable from
clinical isolates are isolated from the ward envir-
onment, it can be argued that patients have con-
taminated the environment directly or indirectly
Prevention and control of P. aeruginosa
Table I Environmental reservoirs of P. aeruginosa in
hospitals
Potable water
Taps/sinks/sink traps
Showers
Disinfectants/sanitisers/antiseptics/bar soap
Respiratory therapy equipment
Ice makers
Flower vases
Shaving/toothbrushes
Medication, e.g. eye drops; multi-dose vials;
mouthwash
Mop heads/buckets
Endoscopes/endoscope washers
Urometers
Water baths
Hydrotherapy pools
Infant feeding basins
Bathing basins
Bath toys
Cleaning equipment
rather than vice versa. Nevertheless, as discussed
below there is now accumulating evidence to sug-
gest that the environment can play an important
role in the epidemiology of HCAI caused by
P. aeruginosa.
P. aeruginosa infection in hospitals:
prevention is better than cure
Given the possible consequences of P. aeruginosa
infection, particularly for vulnerable patients,
and the continued emergence of MDR strains which
hamper effective antimicrobial therapy, it is clear
that strategies to prevent infection and reduce
the likelihood of antimicrobial resistance should
become a key priority.
Antimicrobial stewardship
One of the principal elements of an antimicrobial
stewardship programme is restriction of certain
classes of antimicrobials, to try to delay the
emergence of resistance to these agents. Unfortu-
nately, recent studies have reported only partial
success with this approach. Ntagiopolous et al. re-
stricted the use of ciprofloxacin and ceftazidime
over an 18 month period in an ICU.18 Although
resistance of P. aeruginosa to both agents fell
by the end of the intervention, this was at the
expense of resistance to meropenem which
increased significantly. Similarly, a 34 month
341
hospital-wide intervention to reduce ciprofloxacin
use was associated with a significant reduction in
the numbers of ciprofloxacin-resistant P. aerugi-
nosa isolates, but this was tempered by an
increase in carbapenem-resistant strains.19 In a
five-year study involving 35 ICUs Jonas et al. exam-
ined the genodiversity of ciprofloxacin- and imipe-
nem-resistant P. aeruginosa isolates and found
that this was significantly lower in ICUs with high
resistance rates and low antimicrobial usage com-
pared with ICUs with low resistance rates and high
antimicrobial use.20 This suggests that high resis-
tance rates in units with low antimicrobial use
occurs due to more frequent opportunities for
cross-transmission. Thus, efforts to reduce the
emergence and propagation of antibiotic resis-
tance in P. aeruginosa through prudent antimicro-
bial use may be undermined if infection control
practice is suboptimal.
Infection control
As illustrated by reports from outbreak investiga-
tions, the importance of standard infection control
techniques such as good hand hygiene for the
control and prevention of P. aeruginosa infection
cannot be overemphasised. Following an outbreak
in a neonatal ICU in which 16 babies died from
P. aeruginosa infection, multivariate analysis iden-
tified two nurses as being significantly associated
with colonisation/infection of neonates with the
outbreak strains.21 Furthermore, isolates indistin-
guishable on molecular typing from these strains
were cultured from the nurses’ nails, one of
whom had long nails and the other who had artifi-
cial nails. Outbreak control measures included in-
creased hand hygiene, banning artificial nails and
restricting the length of non-artificial nails. These
reduced but did not eliminate further cases of in-
fection.21 Similarly, a scrub nurse with onychmyco-
sis was identified as the likely source of an
outbreak of P. aeruginosa infection in a cardio-
thoracic unit. Improved hand hygiene is often
employed as a control measure during outbreaks;
however, as other interventions, such as isolation
of colonised or infected patients, are usually insti-
tuted contemporaneously, it may be unclear as to
which measure (or combination thereof) is most
successful in preventing the spread of infection.22
The hospital environment
Given the widespread distribution of the bacter-
ium, it is not surprising that efforts are made to
342
identify the presence of P. aeruginosa in the hospi-
tal environment, especially water supplies, sinks,
taps and showerheads. Earlier studies were ham-
pered by reliance on phenotypic typing techniques
such as serotyping or pyocin typing to compare the
relatedness of clinical and environmental iso-
lates.23 The introduction of molecular typing
methods has permitted improved characterisation
of environmental and clinical isolates, strengthen-
ing evidence that hospital water supplies and out-
lets may act as a source of infection.24e28
Nevertheless, the question as to whether water
and water outlets are infecting patients or whether
patients contaminate water outlets cannot be an-
swered definitively by investigations performed in
the outbreak setting. Furthermore, sources of pot-
able water other than mains water should be taken
into account when trying to trace reservoirs for in-
fection. For example, Eckmanns et al. recently re-
ported an outbreak of P. aeruginosa infection in six
ICUs in the same hospital.29 Extensive environmen-
tal sampling identified the outbreak strain in both
opened and unopened units of commercially avail-
able bottled water. The outbreak was terminated
when this water was withdrawn.
Several prospective studies have been under-
taken that acknowledge the limitations of outbreak
investigations.30e34 However, despite variations in
study design, these investigations, with some
exceptions, have provided evidence of patient
acquisition of P. aeruginosa from hospital water
(Table II).35 Rogues et al. investigated a 16-bedded
ICU over a 6 month period.36 Weekly tap water spec-
imens were obtained, as were throat, rectal, spu-
tum and urine cultures from patients on admission
and weekly thereafter. Isolates of P. aeruginosa
were typed using SpeI pulsed-field gel electrophore-
sis (PFGE). The authors concluded that P. aerugi-
nosa carriage by patients was occurring following
exposure to contaminated tap water, and that
colonised patients were the source of contamina-
tion for taps.36 Another study conducted over a 12
month period in five ICUs within the same hospital
compared patient and tap isolates using the same
methodology.31 These were indistinguishable in
56/132 (42%) cases, again suggesting that in non-
outbreak situations water outlets may be an
important reservoir for P. aeruginosa.
Prevention
Growing recognition that hospital water represents
an important source of P. aeruginosa in both
endemic or outbreak situations has prompted the
development of interventions to counter this
K.G. Kerr, A.M. Snelling
Table II Prospective investigations of P. aeruginosa
in hospital water: variables in study design
Frequency of patient sampling
Patient body sites sampled (single versus multiple)
Frequency of water sampling
Nature of environmental sampling (water only; taps
only; water and taps)
Sampling methods (volume of water/pre- vs
post-flush/culture media/temperature of
incubation, etc.)
Number of colonies selected for molecular typing
problem. Aumeran et al. reported that replace-
ment of shower heads and hoses and cleaning dis-
infection failed to control an outbreak of infection
associated with both P. aeruginosa and P. putida
on an oncohaematology unit.24 The outbreak was
terminated following the chlorination of the hospi-
tal water system and the fitting of point-of-use fil-
ters on taps and showers. Subsequently, a new
distribution system for the unit, in which water
was treated by daily chlorination and passed
through prefilters (5 mm and 1 mm) and terminal
filtration (0.22 mm), was installed.24
An outbreak of MDR P. aeruginosa serogroup 011
infection on a neurosurgery ICU was not controlled
by descaling of taps and replacement of faulty
pipework. The outbreak ended only when the
unit was temporarily closed and all the sinks re-
placed.27 Replacement of taps also finally halted
an outbreak of P. aeruginosa infection on a paedi-
atric surgical unit after other measures including
chlorination of the water supply and disinfection
of sinks were unsuccessful.26 Similarly, Hota
et al. were able to terminate an outbreak associ-
ated with contamination of sink drains only when
the sinks were replaced.25 An outbreak of MDR
P. aeruginosa infection on an ICU which proved re-
fractory to control measures, such as enhanced
barrier and hygiene precautions, cohorting of pa-
tients and changes to the unit’s antibiotic policy,
were not sufficient to end the outbreak.37 Thermal
disinfection of taps and the use of sterile water for
patient care purposes had to be instigated.37
Following a study where P. aeruginosa colonisa-
tion or infection in ICU patients was associated
with contaminated taps, Petignat et al. introduced
a range of measures: an increase in the water tem-
perature; copper and silver ionisation; replace-
ment of tap water by P. aeruginosa-free bottled
water for patient use; and reinforcement of stan-
dard precautions and hand hygiene.38 Three groups
of colonised/infected patients were identified.
Group 1 comprised patients whose isolate was
Prevention and control of P. aeruginosa
genotypically indistinguishable from an isolate
recovered from a tap. Group 2 patients had
isolates which were indistinguishable from that of
an isolate recovered from at least one other pa-
tient, but not grown from a tap. Group 3 consisted
of cases with a unique genotype. The control mea-
sures led to statistically significant reductions in
infection/colonisation in group 1 and 2 patients
but cases in group 3 patients, who were assumed
to represent cases of endogenous infection,
remained unchanged.38
Vianelli et al. reported successful control of an
outbreak of P. aeruginosa on a haematology unit
following the installation of point-of-use filtration.
This approach has also been investigated in non-
outbreak situations.39 Trautman et al. introduced
filters on an 11-bedded surgical ICU.40 P. aerugi-
nosa infections were initially deemed to be endog-
enous in origin and selective decontamination of
the digestive tract (SDD) was introduced. This
failed to make an impact on the numbers of infec-
tions, and other measures, such as use of sterile
water for oral care, face washing and cleaning of
taps, were introduced without success. Point-
of-use filters were installed with water outlets
sampled at two-weekly intervals beforehand. In
the pre-filter period 113/117 (97%) water speci-
mens grew P. aeruginosa with [109/113 (96%)]
yielding numbers in excesss of 102 cfu/mL. Molec-
ular typing revealed that all 113 water isolates
and 25/27 (92.6%) patient isolates belonged to
the same clonotype. After introduction of filters
0/52 samples yielded P. aeruginosa and the num-
ber of patients colonised or infected with the
bacterium was significantly reduced, as was the
amount of anti-pseudomonal antibiotics used.40
Discussion
Pseudomonas aeruginosa remains a major HCAI-
related pathogen and is associated with significant
morbidity and mortality, particularly for immuno-
compromised patients and vulnerable patients on
ICUs. Management of infections associated with
this bacterium is rendered increasingly difficult
due to the continued emergence of antibiotic resis-
tance and, more recently, the appearance of pan-
resistant strains. Against this backdrop, the need
for robust strategies for prevention and control of
these infections is self-evident. Results from studies
which have employed molecular typing techniques
undertaken as part of outbreak investigations or
conducted prospectively have highlighted the im-
portance of hospital water supplies as a potential
source of P. aeruginosa. A range of control
343
techniques have been employed with varying de-
grees of success, to reduce or eliminate the bacte-
rium from these sources and futher evaluations of
these are clearly indicated. However, it should be
acknowledged that we cannot rely solely on these
control measures and that they should be employed
as part of a comprehensive multifaceted approach
alongside other interventions such as good hand
hygiene and prudent use of antimicrobials.
Conflict of interest statement
K.K. and A.S. have received research funding
from PALL Europe; K.K. has received speaker
fees from PALL Europe.
Funding sources
None.
References
1. Lodise Jr TP, Patel N, Kwa A, et al. Predictors of 30-day
mortality among patients with Pseudomonas aeruginosa
bloodstream infections: impact of delayed appropriate
antibiotic selection. Antimicrob Agents Chemother 2007;
51:3510e3515.
2. Micek ST, Lloyd AE, Ritchie DJ, Reichley RM, Fraser VJ,
Kollef MH. Pseudomonas aeruginosa bloodstream infection:
importance of appropriate initial antimicrobial treatment.
Antimicrob Agents Chemother 2005;49:1306e1311.
3. Kollef KE, Schramm GE, Wills AR, Reichley RM, Micek ST,
Kollef MH. Predictors of 30-day mortality and hospital costs
in patients with ventilator-associated pneumonia attributed
to potentially antibiotic-resistant Gram-negative bacteria.
Chest 2008;134:281e287.
4. Eifrig CW, Scott IU, Flynn Jr HW, Miller D. Endophthalmitis
caused by Pseudomonas aeruginosa. Ophthalmology 2003;
110:1714e1717.
5. Hidron AI, Edwards JR, Patel J, et al. NHSN annual update:
antimicrobial-resistant pathogens associated with health-
care-associated infections: annual summary of data
reported to the National Healthcare Safety Network at
the centers for disease control and prevention, 2006e
2007. Infect Control Hosp Epidemiol 2008;29:996e1011.
6. Engel J, Balachandran P. Role of Pseudomonas aeruginosa type
III effectors in disease. Curr Opin Microbiol 2009;12:61e66.
7. Tart AH, Wozniak DJ. Shifting paradigms in Pseudomonas
aeruginosa biofilm research. Curr Top Microbiol Immunol
2008;322:193e206.
8. Smith K, Hunter IS. Efficacy of common hospital biocides
with biofilms of multi-drug resistant clinical isolates.
J Med Microbiol 2008;57:966e973.
9. Mesaros N, Nordmann P, Plésiat P, et al. Pseudomonas
aeruginosa: resistance and therapeutic options at the turn
of the new millennium. Clin Microbiol Infect 2007;13:
560e578.
10. Souli M, Galani I, Giamarellou H. Emergence of extensively
drug-resistant and pandrug-resistant Gram-negative bacilli
in Europe. Eurosurveillance 2008;13:1e11.
11. Hanberger H, Arman D, Gill H, et al. Surveillance of micro-
bial resistance in European intensive care units: a first
344
report from the CareeICU programme for improved infec-
tion control. Intensive Care Med 2009;35:91e100.
12. Jean SS, Hsueh PR, Lee WS, et al. Nationwide surveillance
of antimicrobial resistance among non-fermentative
Gram-negative bacteria in intensive care units in Taiwan:
SMART programme data 2005. Int J Antimicrob Agents
2009;33:266e271.
13. Walkty A, Decorby M, Nichol K, Mulvey MR, Hoban D,
Zhanel G. Antimicrobial susceptibility of Pseudomonas
aeruginosa isolates obtained from patients in Canadian
intensive care units as part of the Canadian National Inten-
sive Care Unit study. Diagn Microbiol Infect Dis 2008;61:
217e221.
14. Rosenthal VD, Maki DG, Mehta A, et al. International Noso-
comial Infection Control Consortium report, data summary
for 2002e2007, issued January 2008. Am J Infect Control
2008;36:627e637.
15. Walsh TR. Clinically significant carbapenemases: an update.
Curr Opin Infect Dis 2008;21:367e371.
16. Beno P, Krcmery V, Demitrovicova A. Bacteraemia in cancer
patients caused by colistin-resistant Gram-negative bacilli
after previous exposure to ciprofloxacin and/or colistin.
Clin Microbiol Infect 2006;12:497e498.
17. Matz C, Moreno AM, Alhede M, et al. Pseudomonas aerugi-
nosa uses type III secretion system to kill biofilm-associated
amoebae. ISME J 2008;2:843e852.
18. Ntagiopoulos PG, Paramythiotou E, Antoniadou A,
Giamarellou H, Karabinis A. Impact of an antibiotic restric-
tion policy on the antibiotic resistance patterns of Gram-
negative microorganisms in an intensive care unit in
Greece. Int J Antimicrob Agents 2007;30:360e365.
19. Cook PP, Das TD, Gooch M, Catrou PG. Effect of a program
to reduce hospital ciprofloxacin use on nosocomial Pseudo-
monas aeruginosa susceptibility to quinolones and other an-
timicrobial agents. Infect Control Hosp Epidemiol 2008;29:
716e722.
20. Jonas D, Meyer E, Schwab F, Grundmann H. Genodiversity of
resistant Pseudomonas aeruginosa isolates in relation to an-
timicrobial usage density and resistance rates in intensive
care units. Infect Control Hosp Epidemiol 2008;29:
350e357.
21. Moolenaar RL, Crutcher JM, San Joaquin VH, et al. A pro-
longed outbreak of Pseudomonas aeruginosa in a neonatal
intensive care unit: did staff fingernails play a role in
disease transmission? Infect Control Hosp Epidemiol 2000;
21:80e85.
22. McNeil SA, Nordstrom-Lerner L, Malani PN, Zervos M,
Kauffman CA. Outbreak of sternal surgical site infections
due to Pseudomonas aeruginosa traced to a scrub nurse
with onychomycosis. Clin Infect Dis 2001;33:317e323.
23. Trautmann M, Lepper PM, Haller M. Ecology of Pseudomonas
aeruginosa in the intensive care unit and the evolving role
of water outlets as a reservoir of the organism. Am J Infect
Control 2005;33:S41eS49.
24. Aumeran C, Paillard C, Robin F, et al. Pseudomonas aerugi-
nosa and Pseudomonas putida outbreak associated with
contaminated water outlets in an oncohaematology paedi-
atric unit. J Hosp Infect 2007;65:47e53.
25. Hota S, Hirji Z, Stockton K, et al. Outbreak of multidrug-
resistant Pseudomonas aeruginosa colonization and
infection secondary to imperfect intensive care unit room
design. Infect Control Hosp Epidemiol 2009;30:25e33.
26. Ferroni A, Nguyen L, Pron B, Quesne G, Brusset MC,
Berche P. Outbreak of nosocomial urinary tract infections
K.G. Kerr, A.M. Snelling
due to Pseudomonas aeruginosa in a paediatric surgical unit
associated with tap-water contamination. J Hosp Infect
1998;39:301e307.
27. Bert F, Maubec E, Bruneau B, Berry P, Lambert-Zechovsky N.
Multi-resistant Pseudomonas aeruginosa outbreak associ-
ated with contaminated tap water in a neurosurgery inten-
sive care unit. J Hosp Infect 1998;39:53e62.
28. Kolmos HJ, Thuesen B, Nielsen SV, Lohmann M,
Kristoffersen K, Rosdahl VT. Outbreak of infection in a burns
unit due to Pseudomonas aeruginosa originating from con-
taminated tubing used for irrigation of patients. J Hosp In-
fect 1993;24:11e21.
29. Eckmanns T, Oppert M, Martin M, et al. An outbreak of
hospital-acquired Pseudomonas aeruginosa infection
caused by contaminated bottled water in intensive care
units. Clin Microbiol Infect 2008;14:454e458.
30. Trautmann M, Bauer C, Schumann C, et al. Common RAPD
pattern of Pseudomonas aeruginosa from patients and tap
water in a medical intensive care unit. Int J Hyg Environ
Health 2006;209:325e331.
31. Blanc DS, Nahimana I, Petignat C, Wenger A, Bille J,
Francioli P. Faucets as a reservoir of endemic Pseudomonas
aeruginosa colonization/infections in intensive care units.
Intensive Care Med 2004;30:1964e1968.
32. Berthelot P, Grattard F, Mahul P, et al. Prospective study of
nosocomial colonization and infection due to Pseudomonas
aeruginosa in mechanically ventilated patients. Intensive
Care Med 2001;27:503e512.
33. Reuter S, Sigge A, Wiedeck H, Trautmann M. Analysis of
transmission pathways of Pseudomonas aeruginosa between
patients and tap water outlets. Crit Care Med 2002;30:
2222e2228.
34. Vallés J, Mariscal D, Cortés P, et al. Patterns of colonization
by Pseudomonas aeruginosa in intubated patients: a 3-year
prospective study of 1,607 isolates using pulsed-field gel
electrophoresis with implications for prevention of ventila-
tor-associated pneumonia. Intensive Care Med 2004;30:
1768e1775.
35. Cholley P, Thouverez M, Floret N, Bertrand X, Talon D. The
role of water fittings in intensive care rooms as reservoirs
for the colonization of patients with Pseudomonas aerugi-
nosa. Intensive Care Med 2008;34:1428e1433.
36. Rogues AM, Boulestreau H, Lashéras A, et al. Contribution
of tap water to patient colonisation with Pseudomonas aer-
uginosa in a medical intensive care unit. J Hosp Infect 2007;
67:72e78.
37. Bukholm G, Tannaes T, Kjelsberg AB, Smith-Erichsen N. An
outbreak of multidrug-resistant Pseudomonas aeruginosa
associated with increased risk of patient death in an inten-
sive care unit. Infect Control Hosp Epidemiol 2002;23:
441e446.
38. Petignat C, Francioli P, Nahimana I, et al. Exogenous
sources of Pseudomonas aeruginosa in intensive care unit
patients: implementation of infection control measures
and follow-up with molecular typing. Infect Control Hosp
Epidemiol 2006;27:953e957.
39. Vianelli N, Giannini MB, Quarti C, et al. Resolution of a Pseu-
domonas aeruginosa outbreak in a hematology unit with the
use of disposable sterile water filters. Haematologica 2006;
91:983e985.
40. Trautmann M, Halder S, Hoegel J, Royer H, Haller M. Point-
of-use water filtration reduces endemic Pseudomonas aeru-
ginosa infections on a surgical intensive care unit. Am J Infect
Control 2008;36:421e429.
Journal of Hospital Infection (2009) 73, 345e354
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Extended-spectrum b-lactamase-producing
organisms
M.E. Falagas a,b,c,*, D.E. Karageorgopoulos a
a
Alfa Institute of Biomedical Sciences, Athens, Greece
b
Department of Medicine, Henry Dunant Hospital, Athens, Greece
c
Department of Medicine, Tufts University School of Medicine, Boston, Massachusetts, USA

Available online 10 July 2009

KEYWORDS Summary Extended-spectrum b-lactamases (ESBLs), which hydrolyse

Bacterial drug extended-spectrum cephalosporins and are inhibited by clavulanic acid,
resistance; are spreading among Enterobacteriaceae. The CTX-M enzymes are repla-
b-Lactam resistance;
cing SHV and TEM enzymes as the prevalent type of ESBLs, principally in
Cefotaximases;
Genetic techniques;
community-acquired infections caused by Escherichia coli. Associated in-
Microbiological fectious syndromes include mainly urinary tract infections, and secondly
techniques; bloodstream and intra-abdominal infections, and may be serious enough
Plasmids to warrant hospitalisation. Affected patients commonly have various un-
derlying risk factors. This is also observed in hospital-acquired infections.
The rates of ESBL-expression among nosocomial Enterobacteriaceae
isolates, particularly Klebsiella pneumoniae, have risen substantially in
several countries. The hospital epidemiology of these infections is often
complex; multiple clonal strains causing focal outbreaks may co-exist with
sporadic ones. Relevant infection-control measures should focus on reducing
patient-to-patient transmission via the inanimate environment, hospital
personnel, and medical equipment. Wise use of antibiotics is also essen-
tial. The available therapeutic options for the treatment of ESBL-
associated infections are limited by drug resistance conferred by the
ESBLs, along with frequently observed co-resistance to various antibiotic
classes, including cephamycins, fluoroquinolones, aminoglycosides, tetra-
cyclines, and trimethoprim/sulfamethoxazole. Relevant clinical data
regarding the effectiveness of different regimens for ESBL-associated in-
fections are limited. Although certain cephalosporins may appear active
in vitro, associated clinical outcomes are often suboptimal. b-Lactam/
b-lactamase inhibitor combinations may be of value, but the supporting
evidence is weak. Carbapenems are regarded as the agents of choice,

* Corresponding author. Address: Alfa Institute of Biomedical Sciences, 9 Neapoleos Street, 151 23 Marousi, Athens, Greece.
Tel.: þ30 210 683 9604; fax: þ30 210 683 9605.
E-mail address: m.falagas@aibs.gr

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.02.021
346
and may be more effective than fluoroquinolones for serious infections. Ti-
gecycline and polymyxins have substantial antimicrobial activity against
ESBL-producing Enterobacteriaceae, and, along with fosfomycin, merit
further evaluation.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.
Introduction
The main mechanism of bacterial resistance to the
b-lactam class of antibiotics consists of the pro-
duction of b-lactamases, which are hydrolytic
enzymes with the ability to inactivate these
antibiotics before they reach the penicillin-binding
proteins located at the cytoplasmic membrane.
The extended-spectrum b-lactamases (ESBLs) are
classified in the molecular (Ambler) class A and
functional (BusheJacobyeMedeiros) group 2be;
they are characterised by the ability to hydrolyse
an oxyimino-b-lactam at a rate 10% of that for
benzylpenicillin along with inhibition by clavulanic
acid.1,2 The presence of ESBLs in various members
of the Enterobacteriaceae family, particularly
Klebsiella pneumoniae and Escherichia coli, is of
great microbiological and clinical importance.
ESBLs are also found in non-fermentative Gram-
negative bacteria, such as Pseudomonas aerugi-
nosa and Acinetobacter baumannii.3
The ESBL enzymes were initially recognised in
clinical isolates in the 1980s; they derived mainly
from the TEM or SHV types of b-lactamases, by point
mutations in the parent enzymes which did not
possess extended-spectrum b-lactam substrate
activity.3 The CTX-M type of ESBLs is becoming in-
creasingly more prevalent, particularly in E. coli
and K. pneumoniae.4,5 More than 50 enzymes of
the latter type have so far been identified, which
can be divided into five main groups on the basis
of amino acid changes (CTX-M1, CTX-M2, CTX-M8,
CTX-M9 and CTX-M25, respectively).6 The origin of
some of these enzymes has been traced to chromo-
somally encoded enzymes of the Kluyvera spp. of
environmental bacteria. The relevant genes are
thought to have been mobilised into conjugative
plasmids and thus transferred to pathogenic bac-
teria.6 Additional clinically relevant types of ESBLs
include mainly the VEB, PER, GES, TLA, IBC,
SFO-1, BES-1 and BEL-1 types.3
Global epidemiology
The rate of ESBL production among Enterobacter-
iaceae varies worldwide. In a recent study based
M.E. Falagas, D.E. Karageorgopoulos
on the Tigecycline Evaluation and Surveillance
Trial (TEST) global surveillance database, the
rate of ESBL production was highest among the
K. pneumoniae isolates collected in Latin America,
followed by Asia/Pacific Rim, Europe, and North
America (44.0%, 22.4%, 13.3% and 7.5%, respec-
tively).7 The same ranking order between the
different geographical regions was observed
regarding the prevalence of ESBLs among the E.
coli isolates, although the corresponding rates
were lower (13.5%, 12.0%, 7.6%, and 2.2%, respec-
tively).7 It should be mentioned that the above
data refer to isolates related to hospital-acquired
infections obtained from various clinical
specimens.
Detailed data derived from the TEST database
regarding the prevalence of ESBL production
among Enterobacteriaceae isolates in Europe
have recently been presented.8 According to these
data that refer to 22 European countries for the
period of 2004 to 2007, the rate of ESBL production
among 515 K. pneumoniae isolates and 794 E. coli
isolates was 15.5% and 9.8%, respectively. Marked
differences were observed in the country-specific
data; the highest rate of ESBL production was
noted in Greece, while the lowest was noted in
Denmark. Relevant data collected by the European
Antimicrobial Resistance Surveillance System
(EARSS Annual Report 2007. Bilthoven. The Nether-
lands. ISBN:978-90-6960-214-1. available at:
http://www.rivm.nl/earss/) regarding resistance
rates to third-generation cephalosporins of
K. pneumoniae and E. coli clinical isolates col-
lected in 31 European countries are consistent
with those of the TEST database.
Particular attention should be paid to the in-
creasing prevalence of the CTX-M type ESBLs
worldwide.6 This can be attributed to the spread
of CTX-M genes among bacterial species by plas-
mids or other mobile genetic elements, as well as
to the clonal expansion of epidemic strains carry-
ing these genes.5 The prevalence of specific types
or groups of CTX-M ESBLs has acquired endemic
proportions in many countries. Among European
countries, relevant examples include CTX-M-1 en-
zymes in Italy, CTX-M-9 and CTX-M-14 enzymes in
Spain, CTX-M-3 enzymes in Poland, and CTX-M-15
ESBL-producing organisms
enzymes in the UK.5,6,9 Notably, the CTX-M-15
ESBLs exhibit a nearly worldwide distribution.5
Nevertheless CTX-M enzymes have rarely been
found responsible for ESBL production among clin-
ical isolates collected in the USA. However, a re-
cent study highlights the increasing prevalence of
CTX-M type ESBLs in a large US institution.10
Laboratory detection
In the microbiological laboratory, detection of
ESBLs can be done with phenotypic or genotypic
tests. The phenotypic tests are routinely used in
clinical diagnostic laboratories, whereas the geno-
typic tests are mainly used in reference or re-
search laboratories.
The phenotypic tests for ESBL detection involve
screening and confirmatory steps. The screening
step consists of testing for resistance to cefpodox-
ime (which is hydrolysed by all TEM, SHV, and CTX-M
types of ESBLs), cefotaxime, ceftazidime, ceftriax-
one, or aztreonam.11 The confirmatory step is based
on the demonstration of synergy between the above
agents and clavulanic acid.11 Several methods in-
cluding the double disc synergy test, the combina-
tion disc method, or specific ESBL Etests can be
used in this regard.11,12 Poor sensitivity of these
tests may be observed when the evaluated ESBL-
producing isolate additionally produces a b-lacta-
mase not inhibited by clavulanic acid, such as an
AmpC b-lactamase or metallo-b-lactamase.12
Methods to overcome this limitation include the
use of cefepime, which is a weak substrate for
most AmpC b-lactamases, the use of chromogenic
agar, cloxacillin-containing agar, or the addition of
EDTA to inactivate metallo-b-lactamases.11
The above-mentioned principles for the detec-
tion of ESBLs by phenotypic methods have been
incorporated in most commercial semi-automated
bacterial identification and antimicrobial suscep-
tibility testing systems. However, their perfor-
mance in this regard is variable and their
accuracy appears to be lower compared with the
conventional phenotypic methods.12
The genotypic tests for the detection of ESBLs
primarily consist of polymerase chain reaction-
based amplification of the specific genes. Regard-
ing the TEM and SHV type ESBLs, additional
molecular techniques, such as sequencing or
restriction fragment length polymorphism, are re-
quired for the identification of specific point
mutations that differentiate these enzymes from
parent enzymes without ESBL activity.9 Although
technically challenging, the genotypic methods
have the advantage of identification of the specific
347
type of ESBL present in a micro-organism, which
may be particularly useful for epidemiological
purposes.9 Moreover, they can detect low-level
resistance, and can be performed without prior
culture of the microbiological specimen.
Clinical relevance and impact of ESBL-
associated infections
It is increasingly being recognised that the pro-
duction of ESBLs is not relevant to nosocomial
infections only, but is becoming an important public
health issue also with regard to infections acquired
in the community. Community-onset ESBL-associ-
ated infections are principally caused by E. coli pro-
ducing CTX-M type ESBLs.9 Urinary tract infections
constitute the main clinical syndrome observed in
this setting. Bloodstream infections may also be
observed, mainly of urinary or biliary tract origin.
Community-acquired ESBL-associated infections
typically affect patients with various complicating
factors. A relevant caseecontrol study identified
various risk factors for community-acquired
infection by ESBL-producing E. coli, including
increased age, female sex, diabetes mellitus, re-
current urinary tract infection, previous instru-
mentation of the urinary tract, follow-up in
outpatient clinic, and previous receipt of amino-
penicillins, cephalosporins, or fluoroquinolones.13
Such findings raise the question whether commu-
nity-onset ESBL-associated infections are mainly
healthcare-associated. However, reports of truly
community-acquired infections are increasing,
while clusters of cases in the community, particu-
larly among members of the same family, may be
observed.14 Additionally, faecal carriage of ESBLs
has been reported in a considerable percentage
of healthy individuals residing in the community.15e17
Potential transmission of ESBL-producing organ-
isms from animal sources to humans through the
food chain or patient-to-patient transmission of
these organisms might contribute to the dissemi-
nation of ESBLs in the community, but these issues
require further study.16
ESBL-associated infections observed in hospitalised
patients represent either serious community-
acquired infections requiring hospital admission or
infections acquired during hospitalisation. The
degree that community-acquired infections con-
tribute to the isolation of ESBL-producing organisms
in hospitalised patients appears to be increasing.18
Furthermore, infection or colonisation of residents
in long-term care facilities by ESBL-producing
organisms may provide a means for the dissemina-
tion of ESBLs between the community and hospitals.
348
In this respect, the incidence of colonisation by
ESBL-producing organisms of residents in long-
term care facilities appears to be substantially in-
creasing, although relevant data are limited.19
Regarding hospital-acquired infections caused by
ESBL-producing organisms, the majority of relevant
studies refer to K. pneumoniae.9,14 Clinical syn-
dromes observed in this setting include respiratory
tract and wound infections, in addition to urinary
tract, bloodstream, and intra-abdominal ones.9
Risk factors for infection or colonisation of hospital-
ised patients by ESBL-producing organisms are simi-
lar to those referring to other common nosocomial
micro-organisms.20 Specifically, increasing length of
hospital or intensive care unit (ICU) stay, greater se-
verity of clinical status, insertion of various types of
indwelling catheters, performance of certain types
of invasive procedures or surgical interventions, re-
ceipt of renal replacement therapy or mechanical
ventilatory support have all been associated with
the isolation of ESBL-producing organisms from hos-
pitalised patients.3,21 The use of antibiotics, particu-
larly of oxyimino-b-lactams or fluoroquinolones,
constitutes an important additional risk factor.3
Studies assessing the epidemiology of hospital-
acquired ESBL-associated infections have often
recognised a complex pattern.18 Specifically, epi-
demic strains may co-exist with sporadic ones,20
while multiple predominant clones may also be ob-
served.18 Furthermore, the same types of ESBLs
may be identified in clonally unrelated isolates,
or, conversely, isolates with the same clonal origin
may encode different types of ESBLs.14 These ob-
servations are attributed to horizontal spread of
ESBL genes through mobile genetic elements.15
The clinical impact of ESBL-associated infections
has mainly been studied in hospitalised patients,
especially those with bloodstream infections. In this
respect, it has been found that bloodstream
infections by ESBL-producing Enterobacteriaceae
isolates compared with non-producing isolates is
associated with a delay in the institution of appro-
priate antimicrobial therapy, as empirically insti-
tuted antibiotics may be inactive.22 The above
factor is thought to be mainly responsible for the
increased mortality related to ESBL production in
K. pneumoniae or E. coli bloodstream infections.22
Hospital infection control
The general aspects of hospital infection control for
ESBL-producing Enterobacteriaceae are similar to
those applied for other common nosocomial Gram-
negative organisms.14,23 Specifically, infection con-
trol measures should focus on preventing the main
M.E. Falagas, D.E. Karageorgopoulos
modes of patient-to-patient transmission of ESBL-
producing organisms in the hospital setting, which
include transmission via colonisation of the
inanimate environment, the hands of healthcare
personnel, and of medical equipment.14
The identification of patients colonised with ESBL-
producing organisms can be done with surveillance
cultures of gastrointestinal tract specimens, partic-
ularly with rectal swabs.14 It has been shown that
a substantial percentage of patients who develop
nosocomial ESBL-associated infections have preced-
ing colonisation of the gastrointestinal tract. How-
ever, the identification of ESBL producers among
commensal Enterobacteriaceae is technically de-
manding and requires the use of selective culture
media. Whether a strategy of selective decontamina-
tion of the gastrointestinal tract of patients found to
be colonised with ESBL-producing organisms is effec-
tive in terms of infection control remains a controver-
sial issue.14 Although effective decolonisation may
reduce the likelihood of subsequent infection by
these organisms, as well as horizontal spread to
neighbouring patients, increased resistance rates of
nosocomial ESBL-producing Enterobacteriaceae iso-
lates to agents commonly used in this regard, such
as neomycin and norfloxacin, limit the utility of this
approach. Moreover, the use of microbiologically ac-
tive agents, such as polymyxins, may carry the risk of
selection for Gram-negative organisms resistant to
this class of agents, which is often used as a last resort
option for infections by highly resistant isolates.24
It should be mentioned that the selection of proper
antibiotic therapy is a key factor relating to the
effectiveness of infection control. Specifically, limit-
ing the institutional use of third generation cephalo-
sporins has been shown to aid towards the reduction
of the prevalence of ESBL-producing organisms.25 Use
of fluoroquinolones may contribute to the selection of
ESBL producers, because determinants of fluoro-
quinolone resistance are often carried in the same
mobile genetic elements as ESBL genes.15 Some stud-
ies have suggested that substitution of cephalosporins
for piperacillin/tazobactam may be useful in limiting
the nosocomial isolation rates of ESBL-producing
organisms.25,26 However, consideration should be given
in preventing the emergence of Gram-negative organ-
isms with advanced antimicrobial drug resistance.27
Treatment
Associated antimicrobial drug resistance
patterns
ESBLs hydrolyse penicillins, cephalosporins (with
the exception of cephamycins), and aztreonam.14
ESBL-producing organisms
However, the degree of hydrolytic activity against
the above substrates may considerably differ for
different types of ESBLs. Typically, the TEM and
SHV type ESBLs have greater hydrolytic activity
for ceftazidime than cefotaxime, in contrast to
the CTX-M type ESBLs.6,14 Consequently, ESBL-pro-
ducing organisms may appear susceptible to some
of the above agents in vitro. The Clinical and Lab-
oratory Standards Institute (CLSI) recommends
that ESBL-producing E. coli, K. pneumoniae, Kleb-
siella oxytoca and P. mirabilis should be reported
as resistant to penicillins, true cephalosporins
and aztreonam, regardless of the in-vitro suscepti-
bility data. The level of in-vitro antimicrobial drug
resistance conferred by the presence of ESBLs to
b-lactam/b-lactamase inhibitor combinations is
variable.14 Relevant minimum inhibitory concen-
trations can be in the susceptible range. An inocu-
lum effect regarding the in-vitro susceptibility of
ESBL-producing isolates to agents that are hydro-
lysed by the ESBLs, including b-lactam/b-lacta-
mase inhibitor combinations, has also been
found. However, clinical relevance of this phenom-
enon has not been firmly established, as it may
represent simply a laboratory artefact.21
An important factor that limits the array of
active antibiotics against ESBL-producing Entero-
bacteriaceae is the frequent co-expression of
resistance by these organisms to classes of antimi-
crobial agents other than those hydrolysed by the
ESBLs. This has been shown for fluoroquinolones,
aminoglycosides, tetracyclines (excluding glycylcy-
clines), and trimethoprim/sulfamethoxazole.9
Cephalosporins
Relatively few studies have assessed clinically
cephalosporin treatment for infections caused by
ESBL-producing organisms, with an agent of this
class showing activity in vitro against the causative
organism. In a small relevant clinical trial, clinical
success rates were similar in seven patients with
CTX-M-producing E. coli bacteraemia treated with
ceftazidime compared with eight such patients
treated with imipenem/cilastatin (86% compared
with 88%, respectively).28 Likewise, a retrospective
study did not find significant differences in the
clinical outcome of 44 ICU hospitalised patients
with TEM-24-producing Enterobacter aerogenes
infections, treated with cefepime-based versus car-
bapenem-based therapy.29 However, the microbio-
logical outcome was inferior in the cefepime
group. Moreover, in the context of a randomised
trial evaluating patients with nosocomial pneumo-
nia, cefepime treatment tended to be associated
with worse outcome compared with imipenem/
349
cilastatin regarding the subgroup of patients
infected with ESBL-producing organisms.30
Additional studies, although small, have reported
suboptimal effectiveness of cephalosporins for
the treatment of ESBL-producing Enterobacteria-
ceae, even if in-vitro antimicrobial activity is
shown.31,32 Thus, most experts argue against
cephalosporin use as treatment of choice for ESBL-
associated infections.
b-Lactams/b-lactamase inhibitor
combinations
The degree of inhibitory activity of b-lactamase
inhibitors against the hydrolysis of b-lactams by
the ESBL enzymes may vary by the type of inhibitor
as well as by the type of ESBL. In this respect,
tazobactam has been found to be more potent
compared with clavulanic acid against certain
CTX-M type ESBLs,33 while both of the above
agents appear to be more potent than sulbactam
in inhibiting TEM and SHV type ESBLs.34
The available clinical evidence regarding the
utility of b-lactam/b-lactamase inhibitor combina-
tions for the treatment of ESBL-associated infec-
tions is rather limited. Specifically, favourable
patient outcomes have been related to piperacil-
lin/tazobactam treatment in some small studies,
although such findings have not been consistently
reproduced.26,35 One pharmacokinetic/pharmaco-
dynamic modelling study concluded that the prob-
ability of pharmacodynamic target attainment
against infections caused by ESBL-producing E. coli
and K. pneumoniae is lower for piperacillin/tazo-
bactam than cefepime if conventional dosing regi-
mens are used.36 Moreover, increasing resistance
rates of ESBL Enterobacteriaceae to piperacillin/ta-
zobactam may limit the potential therapeutic util-
ity of this agent.26 Last but not least, it should be
mentioned that amoxicillin/clavulanate may have
considerable antimicrobial activity against
Enterobacteriaceae organisms isolated in the com-
munity, and it may constitute an effective thera-
peutic option for community-acquired urinary tract
infections caused by ESBL-producing isolates.13,37
Cephamycins
The cephamycins, mainly including cefoxitin, ce-
fotetan, and cefmetazole, do not by definition
constitute substrates for hydrolysis by the ESBL
enzymes. However, co-resistance to these agents
in ESBL-producing Enterobacteriaceae may be
observed, mainly due to porin loss or concomitant
expression of AmpC b-lactamases. Clinical data
350
regarding the potential value of cephamycins for
the treatment of ESBL-associated infections are
scarce. Specifically, a small retrospective study
evaluated treatment with flomoxef, which is
grouped along with latamoxef in the related
oxacephem class of b-lactams, or a carbapenem
for a total of 27 patients with K. pneumoniae
bacteraemia. Difference in mortality between
the two treatment groups was not evident.38
However, additional reports have noted that emer-
gence of resistance to cephamycins may be
observed during therapy with agents, and even
co-resistance to carbapenems also.14,39
Carbapenems
Carbapenems are considered to be the treatment
of choice against serious ESBL-associated infec-
tions.9 This is mainly because they are not inacti-
vated by these enzymes in vitro, and have
demonstrated adequate effectiveness for the
treatment of serious Gram-negative infections at
various body sites. However, specific data for their
clinical use against ESBL-associated infections are
rather limited, although generally supportive of
their effectiveness.30,32,35,40 A multicentre pro-
spective cohort study that evaluated 85 cases of
K. pneumoniae bacteraemia demonstrated that
use of a carbapenem in the initial five-day period
of the infection was a factor independently associ-
ated with lower mortality.41 In addition, in a small
clinical trial, ertapenem use for the treatment of
20 patients with early-onset ventilator-associated
pneumoniae caused by ESBL-producing Enterobac-
teriaceae resulted in a rather favourable overall
clinical success rate of 80%.42
Fluoroquinolones
As mentioned above, ESBL-producing organisms
may carry resistance determinants that confer
low- or high-level resistance to fluoroquinolones.
Potential resistance to fluoroquinolones may relate
to suboptimal patient outcomes when these agents
are elected as empirical therapy for infections
caused by ESBL-producing organisms. There are
some concerns also regarding the effectiveness of
fluoroquinolones for the treatment of serious ESBL-
associated infections caused by fluoroquinolone-
susceptible isolates, compared with carbapenems.
Studies that have provided relevant data e albeit
only for a small number of patients e are summar-
ised in Table I.35,40,41,43,44 Regarding the findings of
the two largest relevant studies, which both address
K. pneumoniae bacteraemia, one study favoured
M.E. Falagas, D.E. Karageorgopoulos
the use of carbapenems over fluoroquinolones,
whereas the other found similar effectiveness
with both these antibiotic classes.41,44
Tigecycline
Tigecycline, a derivative of minocycline, is the first
member of the glycylcycline class of antibiotics
available for clinical use. It has the property to
evade common mechanisms of resistance to
tetracyclines expressed in Gram-negative and
Gram-positive bacteria.45 Detailed data on the
antimicrobial activity of tigecycline against ESBL-
producing Enterobacteriaceae, as reported in a re-
cent systematic review of the literature, are
summarised in Table II.46 Specifically, excellent
activity of tigecycline has been shown against
ESBL-producing E. coli isolates. Additionally, sub-
stantial antimicrobial activity of tigecycline has
been demonstrated against ESBL-producing K.
pneumoniae isolates, although this depends on
the interpretive breakpoints of susceptibility
elected. Data regarding the antimicrobial activity
of tigecycline against other ESBL-producing En-
terobacteriaceae organisms are rather limited.
Clinical data regarding the effectiveness of
tigecycline for the treatment of infections caused
by ESBL-producing organisms are yet limited.46
Further, consideration of the pharmacokinetic
and pharmacodynamic parameters of this agent
casts doubt on the potential effectiveness of tige-
cycline for the treatment of specific infectious
syndromes, such as urinary tract and bloodstream
infections.45 In this respect, only a fraction of
10e15% of the tigecycline dose appears to be ex-
creted as active, unchanged drug in the urine.
Achievable serum concentrations of this agent
may also be inadequate (due to extensive tissue
drug distribution) for substantial antimicrobial
activity to be exerted against pathogenic micro-
organisms circulating in the bloodstream with
minimum inhibitory concentrations close to the
susceptibility breakpoint.
Considerations for further research
The re-evaluation of earlier-used antimicrobial
agents, that have had low clinical use in recent
decades, for potential antimicrobial activity and
clinical effectiveness against today’s resistant
micro-organisms may provide a temporary solution
to the problem of spreading and advancing bac-
terial drug resistance. Agents that may be useful
for the treatment of ESBL-associated infections
include polymyxins, fosfomycin, nitrofurantoin,
Table I Clinical outcome in patients with infections caused by ESBL-producing Enterobacteriaceae treated with fluoroquinolones, carbapenems, or other agents
showing in-vitro activity
ESBL-producing organisms

Study Study design Patient characteristics Outcome Specific agents: n/N (%)
Fluoroquinolones Carbapenems Other antibiotics
a
Kim et al. Retrospective Adult patients with Mortality Ciprofloxacin : 1/3 Imipenem: 2/12 (17) Aminoglycosidesa: 2/4 (50)
(2002)43 cohort study ESBL K. pneumoniae (33)
bacteraemia
Burgess et al. Retrospective Cases with infections by Clinical Fluoroquinolonesa: Carbapenemsa: 0/3 (0) Piperacillin/tazobactama: 1/3
(2003)35 cohort study ESBL-producing failure 0/3 (0) (33)
organisms (E. coli, K.
pneumoniae, K. oxytoca)
Endimiani et al. Retrospective Patients with TEM-52 Clinical Ciprofloxacina,b: Carbapenemsa,b: 3/11 Aminoglycosidesa,b: 0/2 (0)
(2004)40 cohort study ESBL-producing K. failure 2/7 (29) (27) [imipenem: 2/10
pneumoniae bacteraemia (20); meropenem: 1/1
(100)]
Kang et al. Retrospective Patients with ESBL- 30-day Ciprofloxacinc: Carbapenemsc: 8/62 NR
(2004)44 cohort study producing E. coli or K. mortality 3/29 (10) (12.9)
pneumoniae bacteraemia
Paterson et al. Prospective Patients aged >6 years Mortality Ciprofloxacina: Carbapenemsa: 1/27 (4) Cephalosporinsa: 2/5 (50); b-
(2004)41 multicentre with ESBL-producing K. 4/11 (36) [imipenem: 1/24 (4); lactam/b-lactamase inhibitora:
cohort study pneumoniae meropenem: 0/3 (0)] 2/4 (50)
ESBL, extended-spectrum b-lactamase; NR, not specifically reported.
a
Representing therapy as the only microbiologically active agents.
b
Patients who received adequate treatment and dose for 7 days.
c
Definitive therapy.
351
352
Table II Cumulative susceptibility to tigecycline of ESBL-producing Enterobacteriaceae isolates identified in
different studiesa
Micro-organisms No. of studies
E. coli 16
Klebsiella spp. 17
Enterobacter spp. 4 91.3% (69)
a
Adapted from Kelesidis et al.46
b
Food and Drug Administration (FDA) and European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoints of
susceptibility: minimum inhibitory concentrations 2 and 1 mg/L, respectively.
and temocillin. Furthermore, combining available
cephalosporins with b-lactamase inhibitors could
enhance the effectiveness of the former agents
against ESBL-associated infections.
With regard to polymyxins (of which colistin and
polymyxin B are currently available for clinical
use), these have retained excellent antimicrobial
activity against ESBL-producing organisms.47 How-
ever, reported clinical use of these agents for the
treatment of such infections is scarce, since they
are typically reserved for the treatment of infec-
tions caused by Gram-negative bacteria with
more advanced resistance patterns.23
Fosfomycin can also have good antimicrobial
activity against ESBL-producing Enterobacteri-
aceae.48 Despite concerns that increasing use of
this agent may lead to resistance development, re-
sistance rates among urinary tract isolates have re-
mained low.49 Moreover, a recent study revealed
high effectiveness of fosfomycin in the treatment
of community-acquired lower urinary tract infec-
tion caused by ESBL-producing E. coli.13 The po-
tential value of this drug for the treatment of
systemic infections e of origin other than the
urinary tract e is of marked interest.49 With regard
to nitrofurantoin, this agent may also be effective
in the treatment of ESBL-associated uncompli-
cated urinary tract infections, but this could be
limited by co-resistance to this agent of ESBL-
producing organisms.50
b-Lactamases viewed from a clinical
perspective
Clinicians involved in the treatment of infectious
diseases may be puzzled by the continuous expan-
sion of knowledge concerning b-lactamases, since
new enzymes with potentially varying properties
are increasingly being recognised. From a clinical
standpoint, translating and grouping the complex
microbiological information into therapeutically
meaningful categories might facilitate the choice
of appropriate therapy. In this regard, the need for
M.E. Falagas, D.E. Karageorgopoulos
Susceptibility, % (no. of isolates)
FDA criteriab EUCAST criteriab
99.8% (1636) 99.7% (737)
92.3% (2030) 72.3% (1284)
77.6% (49)
a pragmatic new definition for ESBLs has been
emphasised.51 In fact, re-definition of the ESBLs to
include enzymes of different classes of b-lacta-
mases that confer resistance to extended-spectrum
cephalosporins has recently been proposed.52
Conclusion
The increasing prevalence and shifting epidemiol-
ogy of ESBL-producing organisms, particularly of K.
pneumoniae and E. coli, render the infections
caused by these pathogenic micro-organisms an
important public health problem. The resistance
to extended-spectrum b-lactams, which by defini-
tion these enzymes confer, along with frequently
observed co-resistance to other antibiotics renders
ineffective many of the regimens traditionally used
for the empirical therapy of various types of associ-
ated infections. This may be of particular impor-
tance for community-acquired infections, since
options for oral antibiotic therapy against ESBL-pro-
ducing organisms appear to be limited. Regarding
nosocomial infections caused by these organisms,
carbapenems appear as the most reliable therapeu-
tic agents. Further research is required on appropri-
ate strategies to limit the emergence and spread of
resistant organisms, both in the community and the
hospital settings, as well as to evaluate the avail-
able therapeutic agents and identify new ones.
Conflict of interest statement
None declared.
Funding sources
None.
References
1. Bush K, Jacoby GA, Medeiros AA. A functional classification
scheme for beta-lactamases and its correlation with
molecular structure. Antimicrob Agents Chemother 1995;
39:1211e1233.
ESBL-producing organisms
2. Ambler RP, Meadway RJ. Chemical structure of bacterial
penicillinases. Nature 1969;222:24e26.
3. Jacoby GA, Munoz-Price LS. The new beta-lactamases. N
Engl J Med 2005;352:380e391.
4. Jones CH, Tuckman M, Keeney D, Ruzin A, Bradford PA.
Characterization and sequence analysis of extended spec-
trum b-lactamase encoding genes from Escherichia coli,
Klebsiella pneumoniae and Proteus mirabilis isolates col-
lected during tigecycline phase 3 clinical trials. Antimicrob
Agents Chemother 2009;53:465e475.
5. Livermore DM, Canton R, Gniadkowski M, et al. CTX-M:
changing the face of ESBLs in Europe. J Antimicrob Chemo-
ther 2007;59:165e174.
6. Bonnet R. Growing group of extended-spectrum beta-lacta-
mases: the CTX-M enzymes. Antimicrob Agents Chemother
2004;48:1e14.
7. Reinert RR, Low DE, Rossi F, Zhang X, Wattal C, Dowzicky MJ.
Antimicrobial susceptibility among organisms from the
Asia/Pacific Rim, Europe and Latin and North America
collected as part of TEST and the in vitro activity of tigecy-
cline. J Antimicrob Chemother 2007;60:1018e1029.
8. Hackel M, Badal R, Bouchillon S, et al. Extended-spectrum
beta-lactamase production in Europe. In: Abstracts of the
18th European Congress of Clinical Microbiology and Infec-
tious Diseases (ECCMID), Barcelona, Spain, 2008.
9. Pitout JD, Laupland KB. Extended-spectrum beta-lacta-
mase-producing Enterobacteriaceae: an emerging public-
health concern. Lancet Infect Dis 2008;8:159e166.
10. Lewis 2nd JS, Herrera M, Wickes B, Patterson JE,
Jorgensen JH. First report of the emergence of CTX-M-type
extended-spectrum beta-lactamases (ESBLs) as the
predominant ESBL isolated in a U.S. health care system.
Antimicrob Agents Chemother 2007;51:4015e4021.
11. Drieux L, Brossier F, Sougakoff W, Jarlier V. Phenotypic
detection of extended-spectrum beta-lactamase produc-
tion in Enterobacteriaceae: review and bench guide. Clin
Microbiol Infect 2008;14(Suppl. 1):90e103.
12. Wiegand I, Geiss HK, Mack D, Sturenburg E, Seifert H.
Detection of extended-spectrum beta-lactamases among
Enterobacteriaceae by use of semiautomated microbiology
systems and manual detection procedures. J Clin Microbiol
2007;45:1167e1174.
13. Rodriguez-Bano J, Alcala JC, Cisneros JM, et al. Community
infections caused by extended-spectrum beta-lactamase-
producing Escherichia coli. Arch Intern Med 2008;168:
1897e1902.
14. Paterson DL, Bonomo RA. Extended-spectrum beta-lacta-
mases: a clinical update. Clin Microbiol Rev 2005;18:
657e686.
15. Canton R, Novais A, Valverde A, et al. Prevalence and
spread of extended-spectrum beta-lactamase-producing
Enterobacteriaceae in Europe. Clin Microbiol Infect 2008;
14(Suppl. 1):144e153.
16. Rodriguez-Bano J, Lopez-Cerero L, Navarro MD, Diaz de
Alba P, Pascual A. Faecal carriage of extended-spectrum
beta-lactamase-producing Escherichia coli: prevalence, risk
factors and molecular epidemiology. J Antimicrob Chemo-
ther 2008;62:1142e1149.
17. Falagas ME, Karageorgopoulos DE. Clinical Microbiology and
Infectious Diseases (ECCMID) e 18th European Congress.
Drug resistance among Gram-negative and Gram-positive
bacteria. IDrugs 2008;11:409e411.
18. Valverde A, Coque TM, Garcia-San Miguel L, Baquero F,
Canton R. Complex molecular epidemiology of extended-
spectrum beta-lactamases in Klebsiella pneumoniae:
a long-term perspective from a single institution in Madrid.
J Antimicrob Chemother 2008;61:64e72.
353
19. Nicolas-Chanoine MH, Jarlier V. Extended-spectrum beta-
lactamases in long-term-care facilities. Clin Microbiol
Infect 2008;14(Suppl. 1):111e116.
20. Rodriguez-Bano J, Navarro MD, Romero L, et al. Clinical and
molecular epidemiology of extended-spectrum beta-lacta-
mase-producing Escherichia coli as a cause of nosocomial
infection or colonization: implications for control. Clin
Infect Dis 2006;42:37e45.
21. Pfaller MA, Segreti J. Overview of the epidemiological pro-
file and laboratory detection of extended-spectrum beta-
lactamases. Clin Infect Dis 2006;42(Suppl. 4):S153eS163.
22. Schwaber MJ, Carmeli Y. Mortality and delay in effective
therapy associated with extended-spectrum beta-
lactamase production in Enterobacteriaceae bacteraemia:
a systematic review and meta-analysis. J Antimicrob Che-
mother 2007;60:913e920.
23. Karageorgopoulos DE, Falagas ME. Current control and
treatment of multidrug-resistant Acinetobacter baumannii
infections. Lancet Infect Dis 2008;8:751e762.
24. Matthaiou DK, Michalopoulos A, Rafailidis PI, et al. Risk fac-
tors associated with the isolation of colistin-resistant gram-
negative bacteria: a matched caseecontrol study. Crit Care
Med 2008;36:807e811.
25. Kim JY, Sohn JW, Park DW, Yoon YK, Kim YM, Kim MJ. Con-
trol of extended-spectrum beta-lactamase-producing Kleb-
siella pneumoniae using a computer-assisted management
program to restrict third-generation cephalosporin use. J
Antimicrob Chemother 2008;62:416e421.
26. Peterson LR. Antibiotic policy and prescribing strategies for
therapy of extended-spectrum beta-lactamase-producing
Enterobacteriaceae: the role of piperacillin-tazobactam.
Clin Microbiol Infect 2008;14(Suppl. 1):181e184.
27. Falagas ME, Rafailidis PI, Kofteridis D, et al. Risk factors of
carbapenem-resistant Klebsiella pneumoniae infections:
a matched case control study. J Antimicrob Chemother
2007;60:1124e1130.
28. Bin C, Hui W, Renyuan Z, et al. Outcome of cephalosporin
treatment of bacteremia due to CTX-M-type extended-
spectrum beta-lactamase-producing Escherichia coli. Diagn
Microbiol Infect Dis 2006;56:351e357.
29. Goethaert K, Van Looveren M, Lammens C, et al. High-dose
cefepime as an alternative treatment for infections caused
by TEM-24 ESBL-producing Enterobacter aerogenes in
severely-ill patients. Clin Microbiol Infect 2006;12:56e62.
30. Zanetti G, Bally F, Greub G, et al. Cefepime versus imi-
penem-cilastatin for treatment of nosocomial pneumonia
in intensive care unit patients: a multicenter, evaluator-
blind, prospective, randomized study. Antimicrob Agents
Chemother 2003;47:3442e3447.
31. Paterson DL, Ko WC, Von Gottberg A, et al. Outcome of ceph-
alosporin treatment for serious infections due to apparently
susceptible organisms producing extended-spectrum
beta-lactamases: implications for the clinical microbiology
laboratory. J Clin Microbiol 2001;39:2206e2212.
32. Wong-Beringer A, Hindler J, Loeloff M, et al. Molecular
correlation for the treatment outcomes in bloodstream
infections caused by Escherichia coli and Klebsiella pneu-
moniae with reduced susceptibility to ceftazidime. Clin
Infect Dis 2002;34:135e146.
33. Bush K, Macalintal C, Rasmussen BA, Lee VJ, Yang Y. Kinetic
interactions of tazobactam with beta-lactamases from all
major structural classes. Antimicrob Agents Chemother
1993;37:851e858.
34. Payne DJ, Cramp R, Winstanley DJ, Knowles DJ. Compara-
tive activities of clavulanic acid, sulbactam, and
tazobactam against clinically important beta-lactamases.
Antimicrob Agents Chemother 1994;38:767e772.
354
35. Burgess DS, Hall 2nd RG, Lewis 2nd JS, Jorgensen JH,
Patterson JE. Clinical and microbiologic analysis of a hospi-
tal’s extended-spectrum beta-lactamase-producing isolates
over a 2-year period. Pharmacotherapy 2003;23:
1232e1237.
36. Ambrose PG, Bhavnani SM, Jones RN. Pharmacokinetics-
pharmacodynamics of cefepime and piperacillin-tazobactam
against Escherichia coli and Klebsiella pneumoniae strains
producing extended-spectrum beta-lactamases: report from
the ARREST program. Antimicrob Agents Chemother 2003;
47:1643e1646.
37. Falagas ME, Polemis M, Alexiou VG, Marini-
Mastrogiannaki A, Kremastinou J, Vatopoulos AC. Antimicro-
bial resistance of Esherichia coli urinary isolates from
primary care patients in Greece. Med Sci Monit 2008;14:
CR75eCR79.
38. Lee CH, Su LH, Tang YF, Liu JW. Treatment of ESBL-producing
Klebsiella pneumoniae bacteraemia with carbapenems or
flomoxef: a retrospective study and laboratory analysis of
the isolates. J Antimicrob Chemother 2006;58:1074e1077.
39. Lee CH, Chu C, Liu JW, Chen YS, Chiu CJ, Su LH. Collateral
damage of flomoxef therapy: in vivo development of porin
deficiency and acquisition of blaDHA-1 leading to ertape-
nem resistance in a clinical isolate of Klebsiella pneumo-
niae producing CTX-M-3 and SHV-5 beta-lactamases. J
Antimicrob Chemother 2007;60:410e413.
40. Endimiani A, Luzzaro F, Perilli M, et al. Bacteremia due to
Klebsiella pneumoniae isolates producing the TEM-52 ex-
tended-spectrum beta-lactamase: treatment outcome of
patients receiving imipenem or ciprofloxacin. Clin Infect
Dis 2004;38:243e251.
41. Paterson DL, Ko WC, Von Gottberg A, et al. Antibiotic ther-
apy for Klebsiella pneumoniae bacteremia: implications of
production of extended-spectrum beta-lactamases. Clin
Infect Dis 2004;39:31e37.
42. Bassetti M, Righi E, Fasce R, et al. Efficacy of ertapenem in
the treatment of early ventilator-associated pneumonia
caused by extended-spectrum beta-lactamase-producing
organisms in an intensive care unit. J Antimicrob Chemo-
ther 2007;60:433e435.
M.E. Falagas, D.E. Karageorgopoulos
43. Kim BN, Woo JH, Kim MN, Ryu J, Kim YS. Clinical implica-
tions of extended-spectrum beta-lactamase-producing
Klebsiella pneumoniae bacteraemia. J Hosp Infect 2002;
52:99e106.
44. Kang CI, Kim SH, Park WB, et al. Bloodstream infections due to
extended-spectrum beta-lactamase-producing Escherichia
coli and Klebsiella pneumoniae: risk factors for mortality
and treatment outcome, with special emphasis on antimicro-
bial therapy. Antimicrob Agents Chemother 2004;48:
4574e4581.
45. Falagas ME, Karageorgopoulos DE, Dimopoulos G. Clinical
significance of the pharmacokinetic and pharmacodynamic
characteristics of tigecycline. Curr Drug Metab 2009;10:
13e21.
46. Kelesidis T, Karageorgopoulos DE, Kelesidis I, Falagas ME.
Tigecycline for the treatment of multidrug-resistant
Enterobacteriaceae: a systematic review of the evidence
from microbiological and clinical studies. J Antimicrob
Chemother 2008;62:895e904.
47. Gales AC, Jones RN, Sader HS. Global assessment of the an-
timicrobial activity of polymyxin B against 54 731 clinical
isolates of Gram-negative bacilli: report from the SENTRY
antimicrobial surveillance programme (2001e2004). Clin
Microbiol Infect 2006;12:315e321.
48. Falagas ME, Kanellopoulou MD, Karageorgopoulos DE, et al.
Antimicrobial susceptibility of multidrug-resistant Gram
negative bacteria to fosfomycin. Eur J Clin Microbiol Infect
Dis 2008;27:439e443.
49. Falagas ME, Giannopoulou KP, Kokolakis GN, Rafailidis PI.
Fosfomycin: use beyond urinary tract and gastrointestinal
infections. Clin Infect Dis 2008;46:1069e1077.
50. Garau J. Other antimicrobials of interest in the era of ex-
tended-spectrum beta-lactamases: fosfomycin, nitrofuran-
toin and tigecycline. Clin Microbiol Infect 2008;14
(Suppl. 1):198e202.
51. Livermore DM. Defining an extended-spectrum beta-
lactamase. Clin Microbiol Infect 2008;14(Suppl. 1):3e10.
52. Giske CG, Sundsfjord AS, Kahlmeter G, et al. Redefining
extended-spectrum beta-lactamases: balancing science
and clinical need. J Antimicrob Chemother 2009;63:1e4.
Journal of Hospital Infection (2009) 73, 355e363
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Acinetobacter: an old friend, but a new enemy

K.J. Towner*

Department of Clinical Microbiology, Nottingham University Hospitals NHS Trust, Queen’s Medical
Centre, Nottingham NG7 2UH, UK

Available online 22 August 2009

KEYWORDS Summary Acinetobacter emerged as a significant nosocomial pathogen

Acinetobacter; during the late 1970s, probably as a consequence, at least in part, of in-
Epidemiology; creasing use of broad-spectrum antibiotics in hospitals. Most clinically sig-
Infection control
nificant isolates belong to the species Acinetobacter baumannii or its close
measures;
Nosocomial infection;
relatives, with many infections concentrated in intensive care, burns or
Therapeutic options high dependency units treating severely ill or debilitated patients. Large
outbreaks can occur in such units, involving the infection or colonisation
of numerous patients by specific epidemic strains of A. baumannii.
Recently, a particular problem has concerned cross-infection of injured
military patients repatriated from combat regions of the world (e.g. Iraq
and Afghanistan). Carbapenems have previously been the treatment of
choice for infected patients, but increasing reports worldwide now
describe A. baumannii isolates resistant to all conventional antimicrobial
regimens. Data to support therapeutic use of the limited number of new
antimicrobial agents (e.g. tigecycline) with in-vitro activity against these
pathogens are still very limited. Detailed advice concerning prevention
and control of outbreaks caused by multidrug-resistant strains of acineto-
bacter is available from the UK Health Protection Agency. In addition to an-
tibiotic prescribing policies and audit, these measures focus on reinforcing
standard infection control procedures and precautions, with particular
attention to thorough cleaning of patient areas to take account of the
long-term survival of acinetobacter after drying and inadequate disinfec-
tion. Despite these measures, the problem continues to escalate, with
many hospitals worldwide now reporting outbreaks caused by multidrug-
resistant strains of acinetobacter.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

* Tel.: þ44 115 9709163; fax: þ44 115 9422190.

E-mail address: Kevin.Towner@nuh.nhs.uk

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.03.032
356 K.J. Towner

Introduction antimicrobial agents, such as ampicillin (60e70% of

isolates susceptible), gentamicin (92.5%), chloram-
The genus Acinetobacter has a long and convo- phenicol (57%) and nalidixic acid (97.8%), so that in-
luted taxonomic history. It was in 1911 that Beijer- fections caused by these organisms could be treated
inck, a Dutch microbiologist working in Delft, relatively easily.2 However, multidrug-resistant
isolated and described the first example of an (MDR) clinical isolates of Acinetobacter spp. have
organism that would now be recognised as Acine- been reported increasingly during the last two de-
tobacter.1 Since then, members of the genus cades, almost certainly as a consequence of exten-
have been classified under a variety of different sive use of potent broad-spectrum antimicrobial
names (e.g. Bacterium anitratum, Herellea vagini- agents in hospitals throughout the world.
cola, Mima polymorpha, Moraxella lwoffi), and In the clinical environment, Acinetobacter bau-
this confusion resulted in difficulties in establish- mannii and its close relatives (genomic species 3
ing the epidemiology and true clinical importance and 13TU), together forming the ‘A. baumannii
of these organisms. Today, the use of molecular complex’, are the genomic species of greatest clin-
methods has established the identity of at least ical importance, together accounting for the vast
33 different species belonging to the genus Acine- majority of infections and hospital outbreaks in-
tobacter, of which 18 have now been assigned for- volving Acinetobacter spp.3e5 The members of
mal species names (Table I). A further 28 unnamed the complex are very difficult for routine diagnos-
groups have been identified that contain multiple tic laboratories to distinguish accurately; there-
strains, and there are also at least 21 ungrouped fore, reports of A. baumannii in the scientific
single strains. However, although the genus Acine- and medical literature should be assumed to in-
tobacter comprises an ever-growing number of de- clude the other members of the complex unless
fined bacterial species, only a minority of these this possibility has been specifically excluded.
species are of clinical significance (see below). Many of the infections caused by these organisms
Members of the genus Acinetobacter first began occur in intensive care or high dependency units
to be recognised as significant nosocomial pathogens in which severely ill or debilitated patients are
during the early 1970s. In early in-vitro studies, most treated extensively with broad-spectrum antimi-
clinical isolates were susceptible to commonly used crobial agents. Members of the A. baumannii com-
plex seem to have an extraordinary ability to
develop resistance to even the most potent antimi-
Table I Species with validly described names that crobial compounds, and are able to rapidly fill the
belong to the genus Acinetobacter a,b ecological niche left vacant by the elimination of
A. calcoaceticus competing bacteria with broad-spectrum com-
A. baumannii pounds. Genetic interchange with other bacterial
A. haemolyticus species that are capable of growing at lower envir-
A. junii onmental temperatures is also possible, with
A. johnsonii
significant implications for the wider spread of
A. lwoffii
A. radioresistens
antibiotic resistance genes in the environment.
A. ursingii
A. schindleri
A. parvus Common misconceptions
A. baylyi
A. bouvetii Many misconceptions concerning acinetobacter still
A. towneri appear repeatedly in the scientific and medical liter-
A. tandoii ature. Chief among these are the statements that
A. tjernbergiae A. baumannii is (i) ubiquitous or highly prevalent in
A. gerneri nature, (ii) that it can be recovered easily from
A. beijerinckii
soil, water and animals, and (iii) that it is a frequent
A. gyllenbergii
a
skin and oropharyngeal commensal of humans. While
A. grimontii has recently been demonstrated to be a het- these statements certainly apply to members of the
erotypic synonym of A. junii, while A. venetianus currently
remains as a provisional designation awaiting further
genus Acinetobacter when considered as a whole,
investigation. A. baumannii (and its close relatives of clinical
b
In addition to the 18 named species listed, a further 26 importance) are not ubiquitous organisms. While it
unnamed groups have been identified that contain multiple is certainly true that A. baumannii can be isolated
strains, and there are also at least 21 ungrouped single from patients and hospital environmental sources
strains that exist in culture collections.
during outbreaks, this species has no known natural
Acinetobacter
habitat outside the hospital.5 This species can be iso-
lated only very rarely from soil, water and other en-
vironmental samples; indeed, during non-outbreak
periods it is often isolated only rarely inside
hospitals.
In summary, three major overlapping popula-
tions of the genus Acinetobacter been identified to
date: (i) MDR isolates found primarily in hospitals
during outbreak situations e these isolates are
capable of colonising and infecting hospitalised
patients, and comprise mainly A. baumannii and
its close relatives; (ii) ‘sensitive’ isolates that
form part of the normal commensal skin flora of
humans (25e70% of individuals) and animals, and
which can also be found as part of the spoilage
flora of many different foodstuffs e these isolates
comprise mainly A. johnsonii, A. lwoffii and A.
radioresistens; and (iii) ‘sensitive’ isolates that
can be found in the environment, soil or wastewa-
ters e these isolates comprise mainly A. calcoace-
ticus and A. johnsonii. The natural habitats of the
many other species belonging to the genus Acine-
tobacter (Table I) are still poorly defined. Most
species have been isolated on an occasional basis
from humans, including clinical specimens, but in
terms of infection control and clinical importance,
concern should be reserved for the MDR isolates
belonging to the A. baumannii complex. Such or-
ganisms are not ubiquitous and are generally iso-
lated only from colonised or infected patients, or
from the hospital environment during outbreaks.
Acinetobacter as a nosocomial pathogen
The main problems caused by acinetobacter in the
hospital setting mostly concern critically ill pa-
tients in intensive care units (ICUs), particularly
those requiring mechanical ventilation, and pa-
tients with wound or burn injuries (trauma pa-
tients). Infections associated with acinetobacter
include ventilator-associated pneumonia, skin and
soft-tissue infections, wound infections, urinary
tract infections, secondary meningitis and blood-
stream infections.2,5 Such infections are caused
predominantly by members of the A. baumannii
complex; nosocomial infections caused by other
species belonging to the genus Acinetobacter are
relatively unusual and are restricted mainly to
catheter-related bloodstream infections and rare
outbreaks related to point-source contamination.
There have also been a few reports of commu-
nity-acquired infections, usually in patients with
comorbidities in tropical or subtropical areas.5
Problems caused by members of the A. bauman-
nii complex in the hospital setting are exacerbated
357
by the high degree of resistance of these organisms
to drying and disinfectants, leading to long-term
persistence in the hospital environment and the
occurrence of outbreaks of infection involving
many patients, and the ever-increasing proportion
of isolates that are MDR. Numerous studies have in-
dicated an upward trend in MDR A. baumannii, but
resistance rates can vary widely according to the
individual hospital, city or country involved.
Thus, in the UK, the latest available surveillance
data show that resistance to carbapenems has in-
creased from <0.5% in 1990 to 24% in 2007 (UK
Health Protection Agency; unpublished data).
Even allowing for the distortions caused by re-
peated sampling of predominant MDR epidemic
strains, it is clear that A. baumannii has a remark-
able ability to acquire resistance genes and to de-
velop resistance against virtually all available
antibiotic classes (see below).
The ‘reservoir’ for nosocomial infection with
A. baumannii still remains uncertain, and may vary
among healthcare facilities. Large numbers of
A. baumannii infections have recently been re-
ported among military casualties repatriated
from war zones in Iraq and Afghanistan.6 Based
on the misconception that A. baumannii is ubiqui-
tous in the environment, it was originally thought
that such infections might have originated from
contact with contaminated soil in the war zones.
However, it has now become clear that the main
cause was contamination of the environment of
field hospitals and other healthcare facilities
involved in the repatriation of military casualties.7
Potential sources of infection in the hospital
environment from which these organisms have
been isolated during outbreaks are numerous
(Table II). Airborne transmission and patient-
to-patient transmission have also been demon-
strated. However, although the hands of hospital
personnel, coupled with contamination of environ-
mental surfaces and medical equipment, may play
a role in the spread of A. baumannii during an out-
break, it seems likely that the infected patient
forms the primary reservoir of infection, with
such patients shedding extremely large numbers
of A. baumannii cells into their surrounding envir-
onment. Factors facilitating the spread of A. bau-
mannii are listed in Table III.
Infection control measures
Once endemic in a healthcare unit, A. baumannii is
extremely difficult to eradicate. Nevertheless, it is
still possible to eradicate these organisms from
a unit when an uncompromising approach is taken
358 K.J. Towner

the environment, and the use of nebulised colistin

Table II Potential sources from which Acineto-
bacter baumannii has been isolated in the hospital for patients with evidence of mild-to-moderate
environment ventilator-associated pneumonia.8 Nevertheless,
there are also numerous examples in which it has
Hands of staff
been necessary to implement patient isolation
Ventilators and tubing
Oxygen analysers and/or ward closures for periods of up to four
Bronchoscopes weeks in order to combat A. baumannii out-
Bed frames breaks.9e16 Detailed guidance concerning contact
Sinks isolation precautions, risk factors for colonisation
Jugs or infection, antibiotic prescribing policies, patient
Soap transfer procedures (internal and external), use of
Plastic screens dedicated equipment, screening strategies, and
Bed linen, pillows and mattresses cleaning and decontamination procedures has been
Resuscitation bags made available by a Working Party of the UK Health
Blood pressure cuffs
Protection Agency (http://www.hpa.org.uk/web/
Parenteral nutrition solution
HPAwebFile/HPAweb_C/1194947325341).17
Gloves
Humidifiers To reiterate, the most important source of A.
Patients baumannii in a potential outbreak situation is the
Respirometers already colonised or infected patient. If an in-
Lotion dispensers crease in the number of cases is detected, the iso-
Rubbish bins lates should first be identified and typed, the
Air supply patients involved should be traced and isolated
Bowls where possible, hygiene and infection control pro-
Hand cream cedures should be re-emphasised and enhanced,
Bedside charts antibiotic policies should be reviewed, and the
Service ducts/dust
unit should be cleaned thoroughly.
Computer keyboards
Cell phones
Epidemiology

Numerous studies have investigated the epidemiol-

to infection control. Normal infection control mea-
sures are often insufficient to halt the transmission
of MDR A. baumannii, but a range of enhanced
measures that focused on controlling environmen-
tal contamination of an outbreak strain proved to
be successful in eradicating A. baumannii from
the ICU of a major London teaching hospital.8
Key measures included use of a closed tracheal
suction system for all patients receiving mechani-
cal ventilation, improved hand decontamination
using alcohol gels, clearer designation of responsi-
bilities and strategies for cleaning equipment and
Table III Factors facilitating the spread of Acineto-
bacter baumannii
Increased length of hospital stay
Prior antibiotics
Mechanical ventilation
Exposure to patients colonised or infected with
A. baumannii
Environmental contamination
Understaffing
Poor adherence of staff to hand hygiene
ogy of A. baumannii. Various patterns of spread of
individual A. baumannii strains have been reported,
including outbreaks caused by the same strain in
multiple hospitals within a city, outbreaks in mul-
tiple cities within a country, and outbreaks in hospi-
tals in several different countries.18e25 In the UK,
a survey in 1999e2001 identified 34 different A.
baumannii genotypes in 46 UK hospitals.26 These
genotypes belonged to 10 different clusters, with
particular strains generally being characteristic of
particular hospitals.26 Between 2003 and 2006,
two carbapenem-resistant A. baumannii lineages
(SE clone and OXA-23 clone) became prevalent in
over 40 UK hospitals each, with these lineages being
susceptible only to colistin and tigecycline.22 More
recently, a further lineage (the Northwest strain)
has become prevalent in several hospitals in the
northern/midland regions of the UK.
The question of whether specific carbapenem-
resistant clones are spreading in European hospi-
tals has also been examined. As part of the EU
ARPAC project, 169 hospitals in 32 countries pro-
vided data concerning MDR isolates of Acineto-
bacter spp.27 In total, 130 hospitals reported
encountering carbapenem-resistant isolates of
Acinetobacter
acinetobacter, ranging from rare sporadic isolates
to an endemic/epidemic situation. Diverse clusters
were identified in European hospitals by random
amplified polymorphic DNA, pulsed-field gel elec-
trophoresis and polymerase chain reaction-based
sequence typing, but three major European lin-
eages were identified. As in the UK, multiple iso-
lates from a single hospital generally belonged to
the same clone.27
Antimicrobial therapy
Table IV lists the classes of antimicrobial agents that
are currently considered to have potential activity
against A. baumannii. However, successive surveys
have shown increasing resistance among clinical
isolates, and high proportions of isolates are
now resistant to most commonly used antibacterial
agents, including aminopenicillins, ureido-
penicillins, broad-spectrum cephalosporins, most
aminoglycosides, quinolones, chloramphenicol and
tetracyclines.2e4 As a consequence, the emergence
of MDR strains of acinetobacter, probably following
the increasing use of broad-spectrum antibiotics in
hospitals, has resulted in carbapenems (especially
imipenem and meropenem) becoming the mainstay
of treatment for acinetobacter infections. How-
ever, reports of carbapenem resistance in clinical
isolates of Acinetobacter spp. are now accumulat-
ing worldwide, with some isolates being resistant
to all conventional antimicrobial agents.18,28e30 It
is therefore vital that therapeutic strategies
optimise the use of existing antimicrobial agents
and minimise the possibilities for the evolution of
drug resistance.
Potential therapeutic approaches for treatment
of A. baumannii infections have been considered
in detail in several recent review articles.5,31e34
The antibiotic classes listed in Table IV have poten-
tial activity against individual strains of acineto-
bacter, but therapy should ideally be based on
Table IV Currently available classes of antimicro-
bial agents with potential activity against Acineto-
bacter baumannii
Anti-pseudomonal broad-spectrum penicillins
Anti-pseudomonal broad-spectrum cephalosporins
Monobactams
Aminoglycosides
Fluoroquinolones
Carbapenems
Polymyxins
Sulbactam
Glycylcyclines
359
the results of properly performed antimicrobial
susceptibility tests. Antibiotic selection for empir-
ical therapy is challenging and should rely,
wherever possible, on institutional-level data con-
cerning the phenotypes and genotypes of A. bau-
mannii strains endemic in a particular hospital
environment. No large prospective controlled clin-
ical trials for the treatment of A. baumannii infec-
tions have been reported to date. Knowledge of
the best therapeutic approaches is therefore
based on in-vitro susceptibility data, small case
series and retrospective analysis of observational
studies. Of the available options (Table IV), wide-
spread resistance means that broad-spectrum
penicillins, broad-spectrum cephalosporins and
monobactams cannot be recommended as suitable
candidates for the empirical treatment of severe
A. baumannii infections, although cefepime and
cefpirome retain some potentially useful residual
activity.
Aminoglycosides
Resistance rates to all the clinically useful amino-
glycoside antibiotics are higher in Acinetobacter
spp. than in most other groups of pathogens.3,35
Most aminoglycoside resistance in Acinetobacter
spp. involves production of aminoglycoside-modi-
fying enzymes, and all three classes of aminoglyco-
side-modifying enzymes have been found in
Acinetobacter spp. Nevertheless, depending on
the results of susceptibility tests, aminoglycosides,
particularly amikacin, continue to show useful ac-
tivity against 40e50% of strains in some centres.
Fluoroquinolones
Until 1990, quinolones had reasonably good activity
against A. baumannii, but resistance to these anti-
biotics subsequently emerged rapidly in clinical iso-
lates. Newer fluoroquinolones, such as
moxifloxacin, sometimes have increased activity
in vitro against A. baumannii in comparison with
older agents such as ciprofloxacin. Spence and
Towner demonstrated that isolates with a single
mutation in the gyrA gene at codon Ser-83 had
a ciprofloxacin minimum inhibitory concentration
(MIC) of 2 to >32 mg/L, but a moxifloxacin MIC of
0.25e1 mg/L.36 All moxifloxacin-resistant (MIC
>2 mg/L) clinical isolates possessed a second
mutation in codon Ser-80 of the parC gene. Addi-
tional single-step mutation studies with ciprofloxa-
cin-resistant/moxifloxacin-susceptible isolates
revealed that such stable moxifloxacin-resistant
mutants were difficult to generate, but that growth
could occur in the presence of moxifloxacin,
360
perhaps indicating the presence of an efflux pump
that could be induced in the presence of
subinhibitory concentrations of moxifloxacin.
Carbapenems
As mentioned above, carbapenems (especially imi-
penem and meropenem) have been a mainstay of
treatment for acinetobacter infections for the past
decade. However, although these agents are still
active against many strains, increasing numbers of
clinical isolates of A. baumannii resistant to carba-
penems are now being reported worldwide, reach-
ing levels of 90% in some centres.34 In northern
European countries (e.g. the UK), carbapenem
resistance is now reported in w15% of A. baumannii
isolates, but levels are much higher in southern
European countries (e.g. Spain, Italy, Greece).
Polymyxins
Polymyxins (e.g. colistin) show bactericidal activ-
ity against A. baumannii, and resistance rates have
remained relatively low. Favourable clinical re-
sponses have been reported for intravenous colis-
tin therapy in case series of ICU patients with
various types of infections, including ventilator-
associated pneumonia and nosocomial meningi-
tis.34 Perhaps surprisingly, reports of colistin toxic-
ity are relatively rare. Polymyxins can be
administered in a nebulised form via the respira-
tory tract in patients with nosocomial pneumonia,
and several small studies have reported clinical ef-
fectiveness.34 Some studies have indicated that
colistin may be useful for treating infections
caused by carbapenem-resistant isolates, but in-
creasing use of polymyxins in critically ill patients
may lead rapidly to the emergence of resistan-
ce.37e39 Indeed, high rates of resistance to colistin
have recently been reported in A. baumannii
isolates from two South Korean hospitals.40
Sulbactam
A few reports have described the successful ther-
apeutic use of sulbactam, which has unusual in-
trinsic activity against Acinetobacter spp., either
alone or in combination with ampicillin.41 Urban
et al. investigated the interaction of sulbactam,
clavulanic acid and tazobactam with the penicil-
lin-binding proteins (PBPs) of imipenem-resistant
and -susceptible isolates of A. baumannii, and
showed in competition-binding experiments that
all three of these b-lactamase inhibitors bound to
the PBPs of the imipenem-susceptible isolates.42
This observation may explain the in-vitro intrinsic
K.J. Towner
antimicrobial properties of these agents against
A. baumannii. However, as currently formulated
for clinical use, sulbactam is likely to be the only
one of these inhibitors to be active in vivo against
A. baumannii. Recent evidence has suggested that
the activity of sulbactam against A. baumannii iso-
lates has declined significantly, perhaps in response
to the increased clinical use of this compound.34
Tigecycline
Tigecycline is a member of a new class of tetracy-
cline-related antimicrobial agents, the glycylcy-
clines, and was found to have good in-vitro activity
against some carbapenem-resistant acinetobacter
isolates in preliminary studies.43,44 However, clini-
cal experience regarding the use of tigecycline for
the treatment of patients with MDR A. baumannii
infections has been somewhat mixed. In the avail-
able clinical reports, tigecycline has mainly been
used, often in combination regimens, to treat
a small number of critically ill patients with infec-
tions such as ventilator-associated pneumonia and
primary or secondary bacteraemia.45,46 Most pa-
tients had a good outcome, but failure of tigecy-
cline to clear A. baumannii bacteraemia was
noted in a few cases, perhaps because of subopti-
mal concentrations of tigecycline in blood. In-vitro
studies have revealed that resistance to tigecy-
cline can be mediated by overexpression of
a multidrug efflux pump, and the emergence of re-
sistance to tigecycline during therapy has already
been reported.45,47
Combination therapy
Various combinations of a carbapenem with sul-
bactam, tobramycin, amikacin, colistin, rifampicin
and aztreonam have been assessed, both in vitro
and in vivo, with somewhat mixed results.34 Other
reports have described the successful use of sul-
bactam, which has unusual intrinsic activity
against Acinetobacter spp. (see above), in combi-
nation with ampicillin, or have proposed unusual
combinations of antibiotics, such as polymyxin B,
imipenem and rifampicin.48e50 Clinical data are
too few to recommend the use of specific combina-
tion regimens for the treatment of infections
caused by MDR strains of A. baumannii, but combi-
nation regimens might be considered by clinicians
in order to achieve synergistic activity and to max-
imise antimicrobial effectiveness (as well as to
minimise the possibility of emergence of further
resistance) in severely ill patients for whom thera-
peutic options are otherwise limited.
Acinetobacter
Discussion
A. baumannii and its close relatives have become
a major cause of hospital-acquired infections pri-
marily because of the remarkable ability of these
bacteria to survive and spread in the hospital en-
vironment and to rapidly acquire resistance deter-
minants to a wide range of antibacterial agents.
Epidemic spread of MDR strains among patients in
hospitals, particularly in ICUs, has been observed
frequently, with infected patients disseminating
large numbers of these organisms into their en-
vironment. The problem is then compounded by
the long-term survival of these organisms on
numerous surfaces and inanimate objects, and by
their high degree of resistance to drying,
disinfectants and antibiotics. Consequently, once
endemic, these organisms are extremely difficult
to eradicate from a particular healthcare unit.
Other special features of the genus Acineto-
bacter include the fact that, perhaps by accident,
it has evolved a range of its own special resistance
genes (particularly carbapenemases), as well as
the capacity to overexpress them in response to
antibiotic challenge. It has also evolved molecular
mechanisms to capture resistance genes from
other organisms, and has developed a range of ex-
pression mechanisms (e.g. provision of promoters
on insertion sequences) that enables ‘foreign’ re-
sistance genes to be expressed. Consequently, an-
timicrobial resistance now greatly limits the
available therapeutic options, especially if isolates
are resistant to the carbapenem class of antimicro-
bial agents, which would otherwise be the agents
of choice for the treatment of severe infections
caused by these organisms. Knowledge of the sus-
ceptibility patterns of strains endemic in particular
geographical areas is essential, but therapy of in-
fections caused by carbapenem-resistant strains
frequently requires the use of unusual drugs (e.g.
colistin or sulbactam), often in novel combina-
tions, or of drugs for which there is presently
very little clinical experience (e.g. tigecycline).
Anecdotes, small case series and case reports sug-
gest that such regimens can be efficacious in indi-
vidual patients, although reports of resistance to
new drugs and unusual combinations are already
beginning to appear.
Well-controlled clinical trials of existing and
new therapies, combined with greater emphasis
on the prevention of healthcare-associated A. bau-
mannii infections, are essential in order to combat
the spread of these MDR organisms. Worldwide
spread of MDR lineages of A. baumannii is now be-
ing observed, although the question of whether
361
these lineages are selected primarily on the basis
of the resistance genes that they carry or because
there is something special about certain lineages
that confers epidemic potential currently remains
unanswered.
Conflict of interest statement
None.
Funding sources
Work on acinetobacter in K.J.T.’s laboratory is
supported, in part, by a grant from the Medical
Research Council (grant no. RA0119).
References
1. Dijkshoorn L, Nemec A. The diversity of the genus Acineto-
bacter. In: Gerischer U, editor. Acinetobacter molecular
biology. Norfolk: Caister Academic Press; 2008. p. 1e34.
2. Bergogne-Bérézin E. Resistance of Acinetobacter spp. to an-
timicrobials e overview of clinical resistance patterns and
therapeutic problems. In: Bergogne-Bérézin E, Joly-
Guillou ML, Towner KJ, editors. Acinetobacter, microbiol-
ogy, epidemiology, infections, management. Boca Raton:
CRC Press; 1996. p. 133e183.
3. Bergogne-Bérézin E, Towner KJ. Acinetobacter spp. as nos-
ocomial pathogens: microbiological, clinical, and epidemio-
logical features. Clin Microbiol Rev 1996;9:148e165.
4. Van Looveren M, Goosens H, ARPAC Steering Group. Antimi-
crobial resistance of Acinetobacter spp. in Europe. Clin
Microbiol Infect 2004;10:684e704.
5. Peleg AY, Seifert H, Paterson DL. Acinetobacter baumannii:
emergence of a successful pathogen. Clin Microbiol Rev
2008;21:538e582.
6. Centers for Disease Control and Prevention. Acinetobacter
baumannii infections among patients at military medical
facilities treating injured US service members, 2002e
2004. Morb Mortal Wkly Rep 2004;53:1063e1066.
7. Scott P, Deye G, Srinivasan A, et al. An outbreak of multi-
drug-resistant Acinetobacter baumannii e calcoaceticus
complex infection in the US military health care system
associated with military operations in Iraq. Clin Infect Dis
2007;44:1577e1584.
8. Wilks M, Wilson A, Warwick S, et al. Control of an outbreak
of multidrug-resistant Acinetobacter baumannii coloniza-
tion and infection in an intensive care unit (ICU) without
closing the ICU or placing patients in isolation. Infect Con-
trol Hosp Epidemiol 2006;27:654e658.
9. Idzenga D, Schouten MA, van Zanten AR. Outbreak of Acine-
tobacter genomic species 3 in a Dutch intensive care unit.
J Hosp Infect 2006;63:485e487.
10. Kraniotaki E, Manganelli R, Platsouka E, Grossato A,
Paniara O, Palù G. Molecular investigation of an outbreak
of multidrug-resistant Acinetobacter baumannii, with char-
acterisation of class 1 integrons. Int J Antimicrob Agents
2006;28:193e199.
11. Longo B, Pantosti A, Luzzi I, et al. An outbreak of Acineto-
bacter baumannii in an intensive care unit: epidemiological
and molecular findings. J Hosp Infect 2006;64:303e305.
362
12. Carbonne A, Naas T, Blanckaert K, et al. Investigation of
a nosocomial outbreak of extended-spectrum beta-lacta-
mase VEB-1-producing isolates of Acinetobacter baumannii
in a hospital setting. J Hosp Infect 2005;60:14e18.
13. Pimentel JD, Low J, Styles K, Harris OC, Hughes A, Athan E.
Control of an outbreak of a multi-drug-resistant Acineto-
bacter baumannii in an intensive care unit and a surgical
ward. J Hosp Infect 2005;59:249e253.
14. De Jong G, Duse A, Richards G, Marais E. Back to basics e
optimizing the use of available resources during an out-
break of multi-drug resistant Acinetobacter spp. J Hosp
Infect 2004;57:186e187.
15. Denton M, Wilcox MH, Parnell P, et al. Role of environmen-
tal cleaning in controlling an outbreak of Acinetobacter
baumannii on a neurosurgical intensive care unit. J Hosp In-
fect 2004;56:106e110.
16. Maragakis LL, Tucker MG, Miller RG, Carroll KC, Perl TM. In-
cidence and prevalence of multidrug-resistant acineto-
bacter using targeted active surveillance cultures. J Am
Med Assoc 2008;299:2513e2514.
17. Health Protection Agency. Working party guidance on the
control of multi-resistant Acinetobacter outbreaks.
18. Landman D, Quale JM, Mayorga D, et al. Citywide clonal
outbreak of multiresistant Acinetobacter baumannii and
Pseudomonas aeruginosa in Brooklyn, NY: the preanti-
biotic era has returned. Arch Intern Med 2002;162:
1515e1520.
19. Turton JF, Kaufmann ME, Warner M, et al. A prevalent mul-
tiresistant clone of Acinetobacter baumannii in Southeast
England. J Hosp Infect 2004;58:170e179.
20. Marais E, de Jong G, Ferraz V, Maloba B, Dusé AG. Inter-
hospital transfer of pan-resistant Acinetobacter strains in
Johannesburg, South Africa. Am J Infect Control 2004;32:
278e281.
21. Nemec A, Dijkshoorn L, van der Reijden TJ. Long-term pre-
dominance of two pan-European clones among multi-resis-
tant Acinetobacter baumannii strains in the Czech
Republic. J Med Microbiol 2004;53:147e153.
22. Coelho JM, Turton JF, Kaufmann ME, et al. Occurrence of
carbapenem-resistant Acinetobacter baumannii clones at
multiple hospitals in London and Southeast England. J Clin
Microbiol 2006;44:3623e3627.
23. Naas T, Coignard B, Carbonne A, et al. VEB-1 extended-
spectrum beta-lactamase-producing Acinetobacter
baumannii, France. Emerg Infect Dis 2006;12:1214e1222.
24. van Dessel H, Dijkshoorn L, van der Reijden T, et al. Identi-
fication of a new geographically widespread multiresistant
Acinetobacter baumannii clone from European hospitals.
Res Microbiol 2004;155:105e112.
25. Seifert H, Dolzani L, Bressan R, et al. Standardization and
interlaboratory reproducibility assessment of pulsed-field
gel electrophoresis-generated fingerprints of Acinetobacter
baumannii. J Clin Microbiol 2005;43:4328e4335.
26. Spence RP, Towner KJ, Henwood CJ, James D, Woodford N,
Livermore DM. Population structure and antibiotic resis-
tance of Acinetobacter DNA group 2 and 13TU isolates from
hospitals in the UK. J Med Microbiol 2002;51:1107e1112.
27. Towner KJ, Levi K, Vlassiadi M, ARPAC Steering Group.
Genetic diversity of carbapenem-resistant isolates of Acine-
tobacter baumannii in Europe. Clin Microbiol Infect 2008;
14:161e167.
28. Coelho J, Woodford N, Turton J, Livermore DM. Multi-resis-
tant acinetobacter in the UK: how big a threat? J Hosp
Infect 2004;58:167e169.
29. Poirel L, Nordmann P. Carbapenem resistance in Acineto-
bacter baumannii: mechanisms and epidemiology. Clin
Microbiol Infect 2006;12:826e836.
K.J. Towner
30. Falagas ME, Bliziotis IA. Pandrug-resistant gram-negative
bacteria: the dawn of the post-antibiotic era? Int J Antimi-
crob Agents 2007;29:630e636.
31. Gilad J, Carmeli Y. Treatment options for multidrug-resis-
tant Acinetobacter species. Drugs 2008;68:165e189.
32. Maragakis LL, Perl TM. Acinetobacter baumannii: epidemi-
ology, antimicrobial resistance, and treatment options. Clin
Infect Dis 2008;46:1254e1263.
33. Dijkshoorn L, Nemec A, Seifert H. An increasing threat in
hospitals: multidrug-resistant Acinetobacter baumannii.
Nat Rev Microbiol 2007;5:939e951.
34. Karageorgopoulos DE, Falagas ME. Current control and
treatment of multidrug-resistant Acinetobacter baumannii
infections. Lancet Infect Dis 2008;8:751e762.
35. Miller JH, Sabatelli FJ, Naples L, Hare RS, Shaw KJ. The
most frequently occurring aminoglycoside resistance mech-
anisms e combined results of surveys in eight regions of the
world. J Chemother 1995;7(Suppl.):17e30.
36. Spence RP, Towner KJ. Frequencies and mechanisms of resis-
tance to moxifloxacin in nosocomial isolates of Acinetobacter
baumannii. J Antimicrob Chemother 2003;52:687e690.
37. Garnacho-Montero J, Ortiz-Leyba C, Jiménez-Jiménez FJ,
et al. Treatment of multidrug-resistant Acinetobacter bau-
mannii ventilator-associated pneumonia (VAP) with intra-
venous colistin: a comparison with imipenem-susceptible
VAP. Clin Infect Dis 2003;36:1111e1118.
38. Michalopoulos A, Kasiakou SK, Rosmarakis ES, Falagas ME.
Cure of multidrug-resistant Acinetobacter baumannii bac-
teraemia with continuous intravenous infusion of colistin.
Scand J Infect Dis 2005;37:142e145.
39. Matthaiou DK, Michalopoulos A, Rafailidis PI, et al. Risk
factors associated with the isolation of colistin-resistant
gram-negative bacteria: a matched caseecontrol study.
Crit Care Med 2008;36:807e811.
40. Ko KS, Suh JY, Kwon KT, et al. High rates of resistance to
colistin and polymyxin B in subgroups of Acinetobacter
baumannii isolates from Korea. J Antimicrob Chemother
2007;60:1163e1167.
41. Levin AS, Levy CE, Manrique AE, Medeiros EA, Costa SF.
Severe nosocomial infections with imipenem-resistant
Acinetobacter baumannii treated with ampicillin/sulbac-
tam. Int J Antimicrob Agents 2003;21:58e62.
42. Urban C, Go E, Mariano N, Rahal JJ. Interaction of sulbac-
tam, clavulanic acid and tazobactam with penicillin-binding
proteins of imipenem-resistant and -susceptible Acineto-
bacter baumannii. FEMS Microbiol Lett 1995;125:193e198.
43. Henwood CJ, Gatward T, Warner M, et al. Antibiotic resis-
tance among clinical isolates of Acinetobacter in the UK,
and in vitro evaluation of tigecycline (GAR-936). J Antimi-
crob Chemother 2002;49:479e487.
44. Pachón-Ibáñez ME, Jiménez-Mejı́as ME, Pichardo C,
Llanos AC, Pachón J. Activity of tigecycline (GAR-936)
against Acinetobacter baumannii strains, including those
resistant to imipenem. Antimicrob Agents Chemother
2004;48:4479e4481.
45. Karageorgopoulos DE, Kelesidis T, Kelesidis I, Falagas ME.
Tigecycline for the treatment of multidrug-resistant
(including carbapenem-resistant) Acinetobacter infections:
a review of the scientific evidence. J Antimicrob Chemo-
ther 2008;62:45e55.
46. Gordon NC, Wareham DW. A review of clinical and microbi-
ological outcomes following treatment of infections involv-
ing multidrug-resistant Acinetobacter baumannii with
tigecycline. J Antimicrob Chemother, 2009;63:775e780.
47. Peleg AY, Adams J, Paterson DL. Tigecycline efflux as a mech-
anism for nonsusceptibility in Acinetobacter baumannii.
Antimicrob Agents Chemother 2007;51:2065e2069.
Acinetobacter
48. Levin AS. Multiresistant Acinetobacter infections: a role for
sulbactam combinations in overcoming an emerging world-
wide problem. Clin Microbiol Infect 2002;8:144e153.
49. Yoon J, Urban C, Terzian C, Mariano N, Rahal JJ. In vitro
double and triple synergistic activities of polymyxin B,
imipenem, and rifampin against multidrug-resistant
363
Acinetobacter baumannii. Antimicrob Agents Chemother
2004;48:753e757.
50. Wareham DW, Bean DC, Urban CM, Rahal JJ. In vitro activ-
ities of polymyxin B, imipenem, and rifampin against multi-
drug-resistant Acinetobacter baumannii. Antimicrob Agents
Chemother 2006;50:825e826.
Journal of Hospital Infection (2009) 73, 364e370
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Community-associated meticillin-resistant
Staphylococcus aureus as a cause of
hospital-acquired infections
R.L. Skov*, K.S. Jensen

National Centre for Antimicrobials and Infection Control, Statens Serum Institut, Copenhagen, Denmark

Available online 27 September 2009

KEYWORDS Summary The worldwide emergence of community-associated meticillin-

Community-associated resistant Staphylococcus aureus (CA-MRSA) during the last decade repre-
MRSA; sents a significant change in the biology of MRSA strains and is changing
Hospital-acquired
the epidemiology of MRSA infections. CA-MRSA infections are caused by
infections;
Mathematical
strains belonging to lineages distinct from HA-MRSA. In the community,
modelling; CA-MRSA strains typically cause skin and soft tissue infections in children
Staphylococcus aureus and younger adults. However, CA-MRSA strains increasingly cause health-
care-acquired infections including surgical site infections, ventilator-asso-
ciated pneumonia and bacteraemia. A mathematical model showing the
influence of MRSA transmission in the community on the prevalence of
MRSA in hospitals is presented. The increasing prevalence of MRSA in the
community also results in an increase in community-onset MRSA (CO-MRSA)
among S. aureus bacteraemia and other invasive infections. These patients
do not have typical risk factors for MRSA. Such changes may have profound
implications for the choice of empirical therapy for serious infections
where S. aureus is a possible cause. The new and potentially very large
reservoir of MRSA in production animals with subsequent transmission to
humans represents an additional serious threat to the control of MRSA both
in general and as a cause of healthcare-acquired infections. CA-MRSA is
thus a matter of serious concern and should be suppressed.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

* Corresponding author. Address: National Centre for Antimicrobials and Infection Control, Statens Serum Institut, 5 Artillerivej,
2300 Copenhagen, Denmark. Tel.: þ45 3268 8348; fax: þ45 3268 3231.
E-mail address: rsk@ssi.dk

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.07.004
CA-MRSA as a cause of HAI
Introduction
Meticillin-resistant Staphylococcus aureus (MRSA)
is one of the most significant causes of health-
care-associated (HA) infections worldwide; it is
associated with significant morbidity and mortality
and imposes a huge burden on healthcare
resources.1,2
Meticillin resistance is encoded by the mecA
gene which has no counterpart in meticillin-
susceptible S. aureus. The origin of the mecA
gene remains to be established. The regulatory
operons and mecA itself are carried on mobile
genetic elements known as the staphylococcal
chromosome cassettes mec (SCCmec).3,4
Recently, the epidemiology of MRSA infections
has changed significantly. Since the first description
of MRSA in 1961 and until the late 1990s, MRSA was
almost exclusively a HA infection.5 Multilocus se-
quence typing (MLST) has shown that >90% of these
MRSA infections were caused by MRSA strains that
belonged to only five genetic lineages/clonal com-
plexes (CC), i.e. CC5, CC8, CC22, CC 30 and CC45
carrying SCCmec cassettes IeVI.6
Today, MRSA transmission and infections are no
longer confined to healthcare institutions but are
increasingly seen in the community, and MRSA
strains are found in yet more genetic lineages.
Similarly new SCCmec types and subtypes are
emerging rapidly and currently at least nine
SCCmec types plus several subtypes have been
described.
Early reports of CA-MRSA
Before this, MRSA in persons without connection to
healthcare institutions [community-associated
(CA)-MRSA] had been described only sporadically
and was confined to particular groups or regions.
The first description of CA-MRSA was by Savoralatz
et al. in 1982 when MRSA spread among intra-
venous drug abusers.7,8 This outbreak was followed
by secondary cases of hospital-acquired infections
(HAIs). In 1993, Udo et al. reported the endemic
presence of CA-MRSA in aboriginal communities
in Western Australia.9 This was followed by a report
from New Zealand including isolates from 1988 to
1996.10 Common to all these reports was that the
causative strains, i.e. ST74-IVa in USA, WA MRSA
5 (ST8-IV) in Australia and ST30-IV in New Zealand,
were all different from the types causing health-
care-associated infections.
In 1998, Herold et al. reported a marked in-
crease in the number of community-acquired
MRSA infections in children admitted to the
365
University of Chicago Children’s Hospital with no
established HA risk factors.11 Again, the MRSA iso-
lates differed from MRSA isolates causing nosoco-
mial infections in the same institution. However,
as stated in the accompanying editorial, this was,
at that time, not perceived as indicative of a gen-
eral change in the epidemiology of MRSA.12 Later,
CA-MRSA from the same time period was reported
as retrospective cases or case series from a number
of European countries including The Netherlands
(1988), Denmark (1993), Finland (1997), Greece
(1998) and France (1998).13e17
However, the general change in the perception
that CA-MRSA actually constitutes a separate en-
tity began with a report of the death of four
children caused by CA-MRSA.18 Since then,
a steadily increasing number of papers have con-
firmed this significant change in the epidemiology
of MRSA transmission and infection.
Characteristics of CA-MRSA
Cases of CA-MRSA differ from the traditional HA-
MRSA in several aspects, both microbiologically
and epidemiologically. CA-MRSA strains belong to
distinct lineages and are not just feral descendants
of HA-MRSA strains but seem to have evolved
separately. Presently, five CA-MRSA lineages are
found worldwide: ST1-IV (USA400); ST8-IV
(USA300); ST30-IV (Pacific/Oceania); ST59-IV and
V (USA1000, Taiwan) and ST80-IV (European).19 Re-
cently, MRSA of the CC398 lineage has emerged
worldwide. This lineage appears to have its habitat
in livestock e especially in pigs e from where
it spreads to people in close contact with MRSA-
positive animals.20
CA-MRSA is characterised by possession of the
smaller SCC mec type IV and, to a lesser degree,
type V or new variants. Depending on the lineage
they are often less resistant to other classes of
antibiotics than HA-MRSA strains.21 Furthermore,
CA-MRSA very often produce potent toxins/viru-
lence factors such as PantoneValentine leucocidin
(PVL), arginine catabolic mobile element (ACME)
and phenol-soluble modulins.19
Epidemiologically, the type of infections, the
age distribution of patients with CA-MRSA infec-
tions and the risk factors differ significantly from
HA-MRSA. Risk factors for CA-MRSA include: history
of colonisation/infection with CA-MRSA; close con-
tact (family) with a person colonised/infected
with CA-MRSA; membership of an indigenous
community; and being a member of ‘special
communities’, e.g. participation in contact sports,
injection drug use, living in correctional facilities
366
or shelters, military personnel, and men who have
sex with men.21,22 Common for these ‘special com-
munities’ is that they share the presence of one
or more of the following ‘5 C’ risk factors: poor
hygiene/Cleanliness; Compromised skin; frequent
skin Contact; Contaminated surfaces and shared
items; and Crowding in addition to frequent anti-
biotic use and overuse.23 Furthermore, MRSA in
these groups tends to build up and eventually
MRSA is likely to disseminate from these groups
into the population in general.
Despite characteristic microbiological and epi-
demiological features, none of these are specific
and there is no clear definition of CA-MRSA. This
has led to the use of several different definitions
including strain characteristics and/or epidemio-
logical criteria including: susceptibility to
clindamycin/ciprofloxacin, being non-multidrug
resistant, carrying SCCmec IV or V, being PVL pos-
itive; community onset combined with a lack of
HA risk factors and without previous history of be-
ing MRSA positive.24 As mentioned previously, none
of the above criteria are exclusive for CA-MRSA and
among the microbiological features it is worth not-
ing that EMRSA-15 (ST 22), which is one of the five
large traditional HA-MRSA lineages, carries SCCmec
IV. Similarly, the characterisation of MRSA infec-
tions in persons previously found MRSA positive as
healthcare-related was very valuable in the begin-
ning of the CA-MRSA epidemic; however, it has
become increasingly problematic and, in several
cases, directly misleading with the increasing num-
ber of CA-MRSA cases.24,25
A definition of CA-MRSA is complicated further
by the nature of S. aureus/MRSA carriage. The bac-
terium can be carried for a prolonged period of
time without/before causing infections, this
means that MRSA carriage can be acquired in the
community, silently carried into the hospital and
then cause infection, which may then be classified
as nosocomial and vice versa.
CA-MRSA as a cause of HAI
CA-MRSA has mainly been associated with rela-
tively mild skin and soft tissue infections in other-
wise healthy people. An important question for the
handling of the change in epidemiology is the
extent to which CA-MRSA strains can cause severe
HAI. In UK hospitals, the situation is still dominated
almost exclusively by EMRSA-15 (ST22-IV) and, to
a lesser degree, EMRSA-16 (ST36-II). However,
there have been reports of HAI caused by isolates
perceived to be CA-MRSA.26,27 Furthermore, full
epidemiological investigation including typing is
R.L. Skov, K.S. Jensen
performed for only a fraction of MRSA cases, which
means that the occurrence of HAI caused by CA-
MRSA strains may be underestimated.
In other countries, HAI caused by CA-MRSA is
already a significant problem. In Greece, Chini
et al. found that 40% of HAI MRSA infections were
caused by strains circulating in the community
(CC30-IVb and CC80-IV).28 In Denmark, indirect ev-
idence suggests that HAI is caused to a still higher
degree by CA-MRSA strain types since 72% of HAI
cases were caused by strains carrying SCCmec IV
or cassette types not belonging to SCCmec types
IeVI (our own unpublished data). The most striking
change has, however, been seen in the USA espe-
cially after the emergence of USA 300 (ST8-IV).
This strain was first recognised around 2000 as
a significant cause of CA-MRSA, but within three
to four years it spread rapidly and became a fre-
quent cause of HAI.29e32 In fact, CA-MRSA might
displace the traditional HA-MRSA in the future, as
indicated by Popovich et al. who found that the in-
cidence density rate of infections caused by strains
classified as community genotypes increased 20-
fold from 0.05 to 0.1 in 2003e2006 compared
with 2000e2003, whereas strains with hospital
genotypes decreased from 0.16 to 0.11.33
Mathematical modelling
The consequences of changes in the epidemiology
and/or the effect of different interventions can
be investigated using mathematical models.34e37
Until very recently, most of the modelling studies
(e.g. Cooper et al.34) looked at ‘the old scenario’
in which transmission was confined to the hospital
environment.
We use a slightly modified version of the de-
terministic model by Cooper et al.34 to simulate
the effect of community transmission of MRSA
(Figure 1, Table I and equations in Appendix).
The simulation model was programmed in Delphi
(Borland Delphi Professional Ver 7.0) using
a fourth-order RungeeKutta method for numerical
integration of the differential equations.
In the model, persons can either be hospitalised
(N1 þ z); recently discharged from hospital, i.e.
having a greater risk of transmission (N2) or belong
to the general population (N3) either as colonised/
infected (y) or uncolonised (x). When MRSA car-
riers are identified in the hospital, they are moved
to ‘isolation’ (z) where no transmission is assumed.
The various transmission terms, bxy, are normal-
ised with respect to the corresponding value of N
in order to facilitate comparison of b-values from
the three population layers.
CA-MRSA as a cause of HAI 367

Uncolonized Colonized Isolated Here b1 is unchanged (b1 ¼ 0.1115 per day), b3 is

chosen as 0.001 b1 (arbitrarily) and b2 is set to
double the community prevalence to 2% in 2 or 4
Hospital

years, respectively, for an isolation capacity of

20e50 beds.38
Finally, we examined the effect of active
screening by increasing 4 to 0.1 per day, i.e. an
average time of 10 days for MRSA-positive patients
to be identified and isolated (or a 45% reduction in
Community

non-isolated MRSA patient-days) with an isolation

N1 = x1+y1
N2 = x2+y2
N3 = x3+y3
Figure 1 Flow diagram for the model with transmis-
sion in hospital and community (adapted and modified
from Cooper et al.34).
The simulations include two time periods, see
Table II and Figure 2: time course ‘0’ equals the
period where transmission was almost exclusively
confined to hospitals (as in the earlier model study,
i.e. b2 ¼ b3 ¼ 0) with no active surveillance for
MRSA, i.e. MRSA is only detected by clinical cul-
tures resulting in 4 ¼ 0.02 per day.34 Hospital
transmission is set to b1 ¼ 0.1115 per day, resulting
in a total prevalence in the community ((y2 þ y3)/
(N2 þ N3)) of 1% at equilibrium (reached after
about 30 years).38
Time segments ‘1 (a, b, c)’ denote the following
period with transmission also in the community.
capacity of 20, 40 and 50 beds. Effects of other in-
terventions have not been included. Throughout
the simulated time periods of intervention (cases
0r, 1ar, 1aR) the b-values were kept from the pre-
ceding period.
Figure 2 shows the steeply rising prevalence of
MRSA after introduction of transmission in the
community at t ¼ 30 years. A new equilibrium level
is reached after 7 to 14 years, leading to a preva-
lence of MRSA carriers in the hospital as high as
20% (case 1a; for a b2 ¼ 0.049 b1) if no interven-
tions are implemented. It further shows that by
using active screening as the intervention
(4 ¼ 0.1 per day), the demands to obtain control
over MRSA in the hospital varied from 5 years
with 20 isolation beds (case 0r) to 11 years with
50 beds (case 1aR).
The model simulations above indicate that even
a very small b-value for community transmission
(b2 z 0.05 b1 in our case 1a) will have a major
impact on the overall dynamics in the hospitale
community system.

Table I Parameter values used in the simulations shown for the model in Figure 1
Symbol Value Description
b1 0.1115 Transmission in hospital, per day
b2 0 or adjusteda Transmission in community layer 2, per day
b3 0 or 0.001 b1 Transmission in community layer 3, per day
Nimax 20 or 50 Number of beds in isolation unit
4 0.02 or 0.1 Isolation rate, per day
p 0.25 Cleared fraction of patients leaving isolation
1/r 20 Mean length of stay in isolation, days
1/l 370 Mean duration of carriage, days
1/m 8 Mean length of hospital stay, days
n 0.0057 High readmission rate, per day
x 0.00063 Low readmission rate, per day
g 0.03 Rate of change from high to low readmission, per day
The above values give rise to the following population levels (for Ntotal ¼ 170 000):
N1 þ z z1000 Number of hospitalised patients
N2 z3500 Number of recently discharged persons
N3 z165 500 General population
Except for b1, b2, b3 and Nimax, the parameter values are copied from the model by Cooper et al.34
a
b2 was adjusted to give a doubling in community prevalence from 1% to 2% within 2 years (cases 1a and 1b in Table II) or within
4 years (case 1c).
368 R.L. Skov, K.S. Jensen

Table II Parameter values (Nimax and 4) and simulation results for a number of cases (time segments) of which
some are shown in Figure 2
Case Nimax 4 (/day) R0 Hospital Community Ttrans (years)
prevalence (%) prevalence (%)
Baseline 0 >18 0.02 1.049 4.6 1.0 z 30
Response 0 0r 20 0.10 0.662 0.0 0.0 5.0
Two-year doubling 1a 20 0.02 1.120 20.1 4.6 7.0
1b 50 0.02 1.162 15.9 4.1 14.1
Four-year doubling 1c 20 0.02 1.079 16.7 3.6 9.7
Response 1 1ar 40 0.10 0.731 14.1 3.4 11.3
1aR 50 0.10 0.731 0.0 0.0 10.5
The prevalences listed are the equilibrium values obtained when the simulated time period is sufficiently long.
Ttrans is the time for the community prevalence to reach 95% of the total change from its case initial value till the new equilibrium
value.
R0 was calculated under the assumption that isolation capacity (free beds) will always be available, which does not hold for long
periods of time in most of the simulated cases (case 0 being the exception).

The N2 population in the model consists of pa-
tients recently discharged, including persons who
are residents in nursing homes; a sector where
transmission of MRSA is an increasing problem.39
In the simulations it is furthermore assumed that
the new MRSA strains have unchanged transmission
rates in hospitals compared with the old HA-MRSA
types; an increased intrahospital transmission
rate of these strains would lead to even worse
scenarios.
Community-onset MRSA
The increasing prevalence of MRSA in the commu-
nity also results in an increase in the proportion
of community onset MRSA (CO-MRSA) among
S. aureus bacteraemia (SAB) and other invasive
infections. In France during 2006e2007, 40% of
invasive MRSA infections had community onset
and a recent large multicentre investigation in
Australia and New Zealand reported that 18% of
community-onset SAB were MRSA.40,41
Furthermore, to a still higher degree these
patients do not have typical risk factors for MRSA
and that makes it difficult to predict whether
a given patient is at risk for having an MRSA
infection. The increase in CO-MRSA combined
with the lack of a typical patient profile may
have profound implications for the choice of
empirical therapy for serious infections where
S. aureus is a possible cause, especially with the
increasing evidence that inappropriate initial anti-
biotic treatment leads to higher mortality.42 The
increasing need for anti-MRSA coverage of empiri-
cal treatment will increase the overall consump-
tion of antibiotics and thereby drive resistance
further, not only in S. aureus, e.g. vancomycin-
resistant MRSA (VRSA), but also in other genera.
New Danish MRSA guidelines
Denmark has had a very low incidence of MRSA, i.e.
<100 new cases per year (both infections and
colonisations) since the 1970s. However, the num-
ber of MRSA cases has increased steeply from 2002
to a maximum of 851 new cases in 2005.21 Some of
the increase was explained by a large outbreak of
HA-MRSA in the six hospitals of Vejle County caused
by a ST22-IV strain, but general increase in MRSA, in
particular in CA-MRSA, was also seen in the rest of
the country.21 Therefore it became apparent that
the policy which had successfully kept MRSA to
a minimum in Denmark for decades simply was not
sufficient to cope with the changing epidemiology.
In consequence, new guidelines for MRSA were im-
plemented in 2006 (www.sst.dk/publ/publ2008/
CFF/MRSA_veijl_en_19mar08.pdf). They reinforce
the use of barrier precautions and single rooms for
all MRSA-positive hospital patients, but also include
interventions against MRSA in nursing homes and for
persons receiving professional healthcare at home.
Furthermore, all MRSA-positive persons are offered
eradication treatment as soon as they are in a condi-
tion that makes eradication of MRSA possible, i.e.
no active infection and preferably no intravenous
lines or catheters. Since the implementation of
the new guidelines, the number of MRSA HAIs, espe-
cially, has declined rapidly to a total of only eight
cases of MRSA bacteraemia in 2007.
In conclusion, the distinction between CA-MRSA
and HA-MRSA is becoming increasingly blurred and
the typical patient at risk for carrying MRSA is
CA-MRSA as a cause of HAI 369

50 Appendix
1ar
Number in isolation

40 The applied deterministic model is described by

the following differential equations:
30 1aR dx1 b x1 y1
1a ¼ 1 þ ly1  mx1 þ vx2 þ xx3
20 dt x1 þy1
0
dy1 b1 x1 y1
10 ¼  ly1  my1 þvy2 þ xy3  Fðy1 ; zÞ
0r dt x1 þ y1
0 dz
¼ Fðy1 ; zÞ  rz
dt
20
Hospital prevalence ( )

dx2 b x2 y2
¼ 2 þ ly2 þ mx1  vx2  gx2 þ prz
1ar dt x2 þ y2
15

1a dy2 b2 x2 y2
10 ¼  ly2 þ my1  vy2  gx2 þ ð1  pÞrz
dt x2 þ y2
5 0 dx3 b x3 y3
¼ 3 þ ly3 þ gx2  xx3
0r 1aR dt x3 þ y3
0
dy3 b3 x3 y3
5 ¼  ly3 þ gy2  xy3
dt x3 þ y3
Community prevalence ( )

4 
1ar 4y1 if z < Nimax
Fðy1 ; zÞ ¼
3
minð4y1 ; rNimax Þ if z ¼ Nimax
1a
2 1aR

0 References
1
0 Emergence and resurgence of meticillin-resistant Staphylo-
0 10 20
Time (year)
Figure 2 Simulation results for number of patients in
isolation, hospital prevalence, and community preva-
lence, respectively, during three scenarios: 0 þ 0r,
0 þ 1a þ 1ar, and 0 þ 1a þ 1aR. (See text and Tables I
and II for corresponding parameter values.)
difficult to define. In the UK, HA-MRSA still domi-
nates. However, in other countries such as Greece
and USA and low-level countries like Denmark, CA-
MRSA causes a significant portion of HAIs. Reports
from several countries document that CA-MRSA is
spreading from closed risk communities into the
general population. Mathematical modelling sug-
gests that transmission of MRSA in the community
influences the occurrence of MRSA and thereby of
HAI in hospitals significantly. CA-MRSA is thus a mat-
ter of serious concern and should be suppressed.
Conflict of interest statement
None declared.
Funding sources
None.
0r
1. Grundmann H, Aires-de-Sousa M, Boyce J, Tiemersma E.
30 40 50
coccus aureus as a public-health threat. Lancet 2006;368:
874e885.
2. Boyce JM, Cookson B, Christiansen K, et al. Meticillin-resistant
Staphylococcus aureus. Lancet Infect Dis 2005;5:653e663.
3. Katayama Y, Ito T, Hiramatsu K. A new class of genetic ele-
ment, staphylococcus cassette chromosome mec, encodes
methicillin resistance in Staphylococcus aureus. Antimicrob
Agents Chemother 2000;44:1549e1555.
4. Ito T, Katayama Y, Hiramatsu K. Cloning and nucleotide se-
quence determination of the entire mec DNA of pre-methi-
cillin-resistant Staphylococcus aureus N315. Antimicrob
Agents Chemother 1999;43:1449e1458.
5. Jevons MP. ‘‘Celbenin’’-resistant staphylococci. Br Med J
1961;1:124e125.
6. Enright MC, Robinson DA, Randle G, Feil EJ, Grundmann H,
Spratt BG. The evolutionary history of methicillin-resistant
Staphylococcus aureus (MRSA). Proc Natl Acad Sci U S A
2002;99:7687e7692.
7. Saravolatz LD, Pohlod DJ, Arking LM. Community-acquired
methicillin-resistant Staphylococcus aureus infections:
a new source for nosocomial outbreaks. Ann Intern Med
1982;97:325e329.
8. Saravolatz LD, Markowitz N, Arking L, Pohlod D, Fisher E.
Methicillin-resistant Staphylococcus aureus. Epidemiologic
observations during a community-acquired outbreak. Ann
Intern Med 1982;96:11e16.
9. Udo EE, Pearman JW, Grubb WB. Genetic analysis of com-
munity isolates of methicillin-resistant Staphylococcus au-
reus in Western Australia. J Hosp Infect 1993;25:97e108.
370
10. Mitchell JM, MacCulloch D, Morris AJ. MRSA in the commu-
nity. N Z Med J 1996;109:411.
11. Herold BC, Immergluck LC, Maranan MC, et al. Community-
acquired methicillin-resistant Staphylococcus aureus in
children with no identified predisposing risk. J Am Med
Assoc 1998;279:593e598.
12. Boyce JM. Are the epidemiology and microbiology of meth-
icillin-resistant Staphylococcus aureus changing? J Am Med
Assoc 1998;279:623e624.
13. Larsen AR, Bocher S, Stegger M, Goering R, Pallesen LV,
Skov R. Epidemiology of European community-associated
methicillin-resistant Staphylococcus aureus clonal complex
80 type IV strains isolated in Denmark from 1993 to 2004.
J Clin Microbiol 2008;46:62e68.
14. Chini V, Petinaki E, Foka A, Paratiras S, Dimitracopoulos G,
Spiliopoulou I. Spread of Staphylococcus aureus clinical iso-
lates carrying PantoneValentine leukocidin genes during
a 3-year period in Greece. Clin Microbiol Infect 2006;12:
29e34.
15. Salmenlinna S, Lyytikainen O, Vuopio-Varkila J. Community-
acquired methicillin-resistant Staphylococcus aureus, Fin-
land. Emerg Infect Dis 2002;8:602e607.
16. Wannet WJB, Spalburg E, Heck MEOC, et al. Emergence of
virulent methicillin-resistant Staphylococcus aureus strains
carrying PantoneValentine leucocidin genes in The Nether-
lands. J Clin Microbiol 2005;43:3341e3345.
17. Dufour P, Gillet Y, Bes M, et al. Community-acquired meth-
icillin-resistant Staphylococcus aureus infections in France:
emergence of a single clone that produces PantoneValen-
tine leukocidin. Clin Infect Dis 2002;35:819e824.
18. Anonymous. Four pediatric deaths from community-acquired
methicillin-resistant Staphylococcus aureus e Minnesota and
North Dakota, 1997e1999. Morb Mortal Wkly Rep 1999;48:
707e710.
19. Diep BA, Otto M. The role of virulence determinants in com-
munity-associated MRSA pathogenesis. Trends Microbiol
2008;16:361e369.
20. Wulf M, Voss A. MRSA in livestock animalsdan epidemic
waiting to happen? Clin Microbiol Infect 2008;14:519e521.
21. Larsen AR, Stegger M, Bocher S, Sorum M, Monnet DL, Skov RL.
Emergence and characterization of community-associated
methicillin-resistant Staphylococcus aureus infections in
Denmark, 1999 to 2006. J Clin Microbiol 2009;47:73e78.
22. Boucher HW, Corey GR. Epidemiology of methicillin-resistant
Staphylococcus aureus. Clin Infect Dis 2008;46(Suppl. 5):
S344eS349.
23. Tong SY, McDonald MI, Holt DC, Currie BJ. Global implica-
tions of the emergence of community-associated methicil-
lin-resistant Staphylococcus aureus in indigenous
populations. Clin Infect Dis 2008;46:1871e1878.
24. David MZ, Glikman D, Crawford SE, et al. What is commu-
nity-associated methicillin-resistant Staphylococcus au-
reus? J Infect Dis 2008;197:1235e1243.
25. Skov R, Gudlaugsson O, Hardardottir H, et al. Proposal for
common Nordic epidemiological terms and definitions for
methicillin-resistant Staphylococcus aureus (MRSA). Scand
J Infect Dis 2008;40:495e502.
26. Otter JA, French GL. The emergence of community-associ-
ated methicillin-resistant Staphylococcus aureus at a Lon-
don teaching hospital, 2000e2006. Clin Microbiol Infect
2008;14:670e676.
27. David MD, Kearns AM, Gossain S, Ganner M, Holmes A. Com-
munity-associated meticillin-resistant Staphylococcus
R.L. Skov, K.S. Jensen
aureus: nosocomial transmission in a neonatal unit. J Hosp
Infect 2006;64:244e250.
28. Chini V, Petinaki E, Meugnier H, et al. Emergence of a new
clone carrying PantoneValentine leukocidin genes and
staphylococcal cassette chromosome mec type V among
methicillin-resistant Staphylococcus aureus in Greece.
Scand J Infect Dis 2008;40:368e372.
29. Anonymous. Methicillin-resistant Staphylococcus aureus in-
fections among competitive sports participants e Colorado,
Indiana, Pennsylvania, and Los Angeles County, 2000e2003.
Morb Mortal Wkly Rep 2003;52:793e795.
30. Anonymous. Methicillin-resistant Staphylococcus aureus in-
fections in correctional facilities e Georgia, California,
and Texas, 2001e2003. Morb Mortal Wkly Rep 2003;52:
992e996.
31. Klevens RM, Morrison MA, Fridkin SK, et al. Community-asso-
ciated methicillin-resistant Staphylococcus aureus and
healthcare risk factors. Emerg Infect Dis 2006;12:1991e1993.
32. Seybold U, Kourbatova EV, Johnson JG, et al. Emergence of
community-associated methicillin-resistant Staphylococcus
aureus USA300 genotype as a major cause of health care-
associated blood stream infections. Clin Infect Dis 2006;
42:647e656.
33. Popovich KJ, Weinstein RA, Hota B. Are community-associ-
ated methicillin-resistant Staphylococcus aureus (MRSA)
strains replacing traditional nosocomial MRSA strains? Clin
Infect Dis 2008;46:787e794.
34. Cooper BS, Medley GF, Stone SP, et al. Methicillin-resistant
Staphylococcus aureus in hospitals and the community:
stealth dynamics and control catastrophes. Proc Natl Acad
Sci USA 2004;101:10223e10228.
35. Bootsma MC, Diekmann O, Bonten MJ. Controlling methicil-
lin-resistant Staphylococcus aureus: quantifying the effects
of interventions and rapid diagnostic testing. Proc Natl Acad
Sci USA 2006;103:5620e5625.
36. D’Agata EM, Webb GF, Horn MA, Moellering Jr RC, Ruan S.
Modeling the invasion of community-acquired methicillin-
resistant Staphylococcus aureus into hospitals. Clin Infect
Dis 2009;48:274e284.
37. Smith DL, Dushoff J, Perencevich EN, Harris AD, Levin SA.
Persistent colonization and the spread of antibiotic
resistance in nosocomial pathogens: resistance is a regional
problem. Proc Natl Acad Sci U S A 2004;101:3709e3714.
38. Gorwitz RJ, Kruszon-Moran D, McAllister SK, et al. Changes
in the prevalence of nasal colonization with Staphylococcus
aureus in the United States, 2001e2004. J Infect Dis 2008;
197:1226e1234.
39. Mody L, Kauffman CA, Donabedian S, Zervos M, Bradley SF.
Epidemiology of Staphylococcus aureus colonization in nurs-
ing home residents. Clin Infect Dis 2008;46:1368e1373.
40. Turnidge JD, Nimmo GR, Pearson J, Gottlieb T, Collignon PJ.
Epidemiology and outcomes for Staphylococcus aureus bac-
teraemia in Australian hospitals, 2005e06: report from the
Australian Group on Antimicrobial Resistance. Commun Dis
Intell 2007;31:398e403.
41. Dauwalder O, Lina G, Durand G, et al. Epidemiology of inva-
sive methicillin-resistant Staphylococcus aureus clones
collected in France in 2006 and 2007. J Clin Microbiol
2008;46:3454e3458.
42. Schramm GE, Johnson JA, Doherty JA, Micek ST, Kollef MH.
Methicillin-resistant Staphylococcus aureus sterile-site in-
fection: the importance of appropriate initial antimicrobial
treatment. Crit Care Med 2006;34:2069e2074.
Journal of Hospital Infection (2009) 73, 371e377
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Screening and isolation for infection control

E. Tacconelli*

Department of Infectious Diseases, Università Cattolica Sacro Cuore, Rome, Italy

Available online 21 August 2009

KEYWORDS Summary Control measures aimed to reduce the prevalence of health-

Healthcare-associated care-associated infections include active surveillance cultures (ASCs), con-
infection; tact isolation of patients colonised with epidemiologically significant
Isolation;
pathogens, and pre-emptive isolation of high risk patients. However, the
Meticillin-resistant
Staphylococcus
benefits of these measures are questionable. A systematic review of isola-
aureus; tion policies demonstrated that intensive concerted interventions includ-
Screening; ing isolation can substantially reduce nosocomial meticillin-resistant
Vancomycin-resistant Staphylococcus aureus (MRSA) infection. Monitoring of interventions is fun-
enterococcus damental. Surveillance data should be presented and fed back appropria-
tely. International guidelines suggest that only intensive care units
should apply extensive ASCs. However, legislation for mandatory screening
at hospital admission has been advocated in many countries. Targeted
screening could be used to limit the potential for dissemination of anti-
biotic-resistant pathogens from otherwise unsuspected carriers from the
start of patients’ hospitalisation, as opposed to other strategies, in which
screening programmes target patients already hospitalised. Although the
influx of antibiotic-resistant pathogens into the hospital would not change,
early detection would reduce the time colonised patients might have to
disseminate pathogens. Recently, rapid methods for molecular detection
of MRSA have been developed. Data on the impact of these tests on
the MRSA acquisition rate are extremely heterogeneous. Published studies
differ according to the settings in which they have been evaluated, the
choice of patient population to be screened, other infection control measures
employed and, most importantly, study design and baseline prevalence
of MRSA. Based on these studies, definitive recommendations cannot be
made.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

* Address: Largo F. Vito 8, 00168, Rome, Italy. Tel.: þ39 06 30155527; fax: þ39 06 3054519.
E-mail address: etacconelli@rm.unicatt.it

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.05.002
372
Introduction
Efforts to control healthcare-associated infections
(HCAIs) have, with few exceptions, failed, result-
ing in growing political and public alarm. Increased
morbidity and mortality due to antibiotic-resistant
bacteria acquired during hospital stay have also
been reported worldwide.1,2 The European Centre
for Disease Prevention and Control has estimated
that HCAIs affect about one in 20 hospital pa-
tients.3 In January 2009 the European Commission
published a Council Recommendation on patient
safety, including the prevention and control of
HCAIs.4 The Commission advised that a national
strategy, complementary to strategies targeted
towards the prudent use of antimicrobial agents,
should be developed incorporating prevention
and control of HCAIs into national public health
objectives and aiming to reduce the risk of such
infections within hospitals. Hence, the question
arises: how are public health professionals to
decide what is the most cost-effective approach
to preventing HCAIs? Unfortunately, a satisfactory
answer from evidence-based data cannot be
provided. Up to now, most studies have focused on
meticillin-resistant Staphylococcus aureus (MRSA)
and vancomycin-resistant enterococci (VRE) and
to a much lesser extent on multidrug-resistant
(MDR) Gram-negative bacteria. The control mea-
sures aimed to reduce the spread of HCAIs include:
education of healthcare workers (HCWs) with im-
plementation and observation of hand-washing
practices; active surveillance cultures (ASCs); con-
tact isolation of patients colonised with epidemio-
logically important bacteria (such as MRSA and
VRE); pre-emptive isolation of high risk patients;
and restriction of antibiotic use. Numerous studies
have demonstrated that the best way to face the
problem is a multidisciplinary approach. Marshall
et al. summarised in four steps the actions needed
to reduce the prevalence of HCAIs: (1) antibiotic
stewardship; (2) reduction in the number of
colonised patients; (3) prevention of infection in
colonised patients; and (4) prevention of cross-
transmission between hospitalised patients.5
Among the strategies aimed to reduce patient-to-
patient transmission, the role of screening and
isolation is still under debate.
The first assumption underlying a screening
strategy is that colonisation increases the risk of
infection. Huang et al. observed that MRSA infec-
tions developed in 29% of 209 colonised patients
in an 18 month follow-up period.6 Twenty-eight
percent of infections involved bacteraemia and
56% involved pneumonia, soft tissue infection,
E. Tacconelli
osteomyelitis, or septic arthritis. A systematic
review including ten studies and 1170 patients
estimated the risk of infection following MRSA
colonisation.7 The analysis showed a four-fold in-
crease (95% confidence interval: 2e7) in the risk
of infection after MRSA colonisation compared
with meticillin-susceptible S. aureus (MSSA) colo-
nisation. Significant heterogeneity was detected
between studies. The majority of reports differed
in the choice of patient population, severity of
illness, and frequency of sampling employed to
detect colonisation, further suggesting the need for
new research with appropriate statistical analysis.
Olivier et al. studied the risk of VRE bacteraemia
among patients previously colonised with the bac-
teria. Four percent of colonised patients devel-
oped bacteraemia caused by the same strain, as
confirmed by the genotypic analysis.8
Important clues in deciding whether screening
and contact isolation are needed are the pro-
portions of antibiotic resistance attributable to
hospital antibiotic use and cross-transmission in
hospitalised patients. In a multicentre prospective
Italian study, we determined the incidence of new
colonisation due to MRSA, VRE and ciprofloxacin-
resistant Pseudomonas aeruginosa per 1000 days of
antibiotic therapy.9 The highest risk was associ-
ated with carbapenem use, with eight new cases
of MRSA colonisation per 1000 days of therapy.
An accurate estimate of the attributable fraction
of antibiotic resistance due to cross-transmission
is not available. Many factors have contributed to
this lack of information. For example, Buke et al.
demonstrated significant epidemiological differ-
ences in the variables associated with colonisation
with MRSA and extended spectrum b-lactamase-
producing (ESBL) Gram-negative bacteria.10
Another parameter useful for a better under-
standing of the attributable fraction due to
cross-transmission is colonisation pressure, i.e.
the prevalence of resistance in the surrounding
environment. Williams et al. demonstrated that
ongoing monitoring of colonisation pressure pro-
vides the opportunity for early implementation
of enhanced infection prevention and control
practices, and can potentially decrease nosocomial
transmission of MRSA and prevent outbreaks.11
Should mandatory screening for MRSA
and VRE be performed?
‘Who’ should be screened for epidemiologically
important pathogens and ‘where’ and ‘when’ are
still open questions. In studies of infection control
Screening and isolation for HCAI prevention
measures, screening is never the primary interven-
tion and it is usually included in a multifaceted
approach. The majority of studies relate to hospital-
acquired MRSA (HA-MRSA). A recent systematic
review of the literature performed by McGinigle
et al. summarised the evidence of the efficacy of
ASCs in the intensive care unit (ICU) setting on sev-
eral clinical and economic outcomes.12 Although
most of observation-based studies reported a signifi-
cant reduction in MRSA infection after the applica-
tion of ASCs, the evidence was of poor quality and
could not allow definitive recommendations. The
role of ASCs might be different according to the base-
line prevalence of MRSA and the hospital setting. In
three intensive care units (ICUs) in France the acqui-
sition of MRSA was reduced from 7% to 3% through
multiple interventions, including ASC, contact pre-
cautions and use of alcohol hand rub.13 All patients
were screened for nasal carriage on hospital admis-
sion and weekly thereafter and placed under contact
precautions if screen positive. Similar results were
obtained in a 18 month study in Colorado, USA.14
All patients admitted to ICUs and medical wards
were screened at admission and weekly thereafter.
The overall rate of HA-MRSA infections compared
with historical controls decreased significantly,
although an increased number of community-ac-
quired MRSA skin infections was admitted to the
wards. The majority occurred in surgery ICU (SICU)
because the medical ICU (MICU) had a higher
incidence of colonisation and lower percentage of
compliance with hand washing. Physicians in SICU
were also less likely to enter isolation rooms.
Some countries have maintained low endemic
levels of MRSA by implementing nationwide control
measures targeting MRSA such as the ‘search
and destroy’ (S&D) strategy.15 Measures of S&D
include: contact isolation for MRSA-positive
patients; pre-emptive isolation and screening for
high risk patients; screening of patients and
personnel following an unexpected case of MRSA;
screening of HCWs with furlough of those found
to be carriers until decontamination is achieved;
closure of wards to new admissions if there
is more than one carrier among hospitalised
patients. Interestingly, Bootsma et al. generated
a stochastic three-hospital model and an analytical
one-hospital model to quantify the effectiveness
of different infection control measures and to
predict the effect of rapid diagnostic testing on
isolation needs.16 The analysis showed that the
application of the S&D strategy in a hospital with
high endemicity would reduce the prevalence to
<1% within 6 years.
The costs of a hospital-wide selective screen-
ing programme were analysed for a 19 month
373
period in a 700-bed hospital with 23 000 admis-
sions annually in Germany.17 The screening pro-
gramme was able to prevent 48% of predicted
nosocomial MRSA, saving a predicted V200,782
with a net saving of V110,000 annually. The
screening programme became cost-effective at
a low MRSA incidence rate. The major limit of
the study was that the authors did not use real
costs calculation but instead hospital reimburse-
ment as a surrogate parameter.
For VRE control, it has been estimated that ASCs
would be cost-beneficial for hospital units where
the number of patients with VRE bacteraemia is at
least 6e9 patients per year or if the savings from
fewer VRE bacteraemic patients in combination
with decreased antimicrobial use ranged from
US$100,000 to 150,000 per year.18
Targeted surveillance based on patients’ risk
factors in medical or surgical wards might be
preferred as the most effective use of hospital
resources. Risk factors for HCAIs due to MRSA at
hospital admissions were analysed in 127 patients
with bacteraemia in an endemic setting in Boston,
USA.19 Two logistic regression models were gener-
ated. In the first model, previous MRSA infection or
colonisation, cellulitis, presence of a central ve-
nous catheter (CVC), and skin ulcers were indepen-
dently associated with MRSA bacteraemia. Since
prior MRSA colonisation or infection status may
not be known at the time of hospitalisation, a sec-
ond model was generated which excluded prior
history of MRSA colonisation or infection. This
model identified the presence of a CVC, hospital-
isation(s) within the previous six months, use of
quinolones within the previous 30 days, and dia-
betes mellitus as independent risk factors for
MRSA bacteraemia. Both models had high sensitiv-
ity and specificity. Harbarth et al. defined risk fac-
tors for unknown carriage of MRSA at hospital
admission in a university hospital in Geneva.20
Male sex, age >75 years, previous exposure to qui-
nolones, cephalosporins and carbapenems, previ-
ous hospitalisation, intrahospital transfer and
urinary catheterisation were associated with
MRSA colonisation at hospital admission.
A prediction score for VRE identification was
derived and validated in a high-endemic setting in
the USA.21 A risk index was derived by using six in-
dependent risk factors associated with VRE recov-
ery within 48 h of hospital admission: previous
isolation of MRSA (4 points), whether the patient
was receiving long-term haemodialysis (3), trans-
fer from a long-term care facility (3), antibiotic ex-
posure (3), prior hospitalisation (3), and age >60
years (2). On the basis of a point score 10, the
sensitivity, specificity, and positive and negative
374
predictive values of this prediction rule were 44%,
98%, 81%, and 90%, respectively.
For several years conventional culture methods,
which can take 48e72 h to obtain a result, have
been considered the gold standard to detect
MRSA colonisation. In the absence of pre-emptive
room isolation this time might be enough to spread
the bacterium among patients. Recently, rapid
methods for molecular detection of MRSA-colon-
ised patients with available results in 2 h have
been developed with high sensitivity and specifi-
city.22 Major limits for the wide application of mo-
lecular diagnosis for MRSA are related to the high
costs and varying benefits in different studies.
Nine studies (eight cohort studies and one clus-
ter-randomised, cross-over trial) analysed the effi-
cacy of screening for MRSA using molecular tests
compared with no screening or screening with con-
ventional cultures.23e30 Studies were performed in
different settings: ICUs (two studies), medical
and/or surgery wards (four) and throughout the
hospital (two). Four papers reported a significant
decrease in the rate of HA-MRSA, two observed
no effects, and one showed uncertain results.
Variability in the efficacy of polymerase chain
reaction (PCR) might be partially explained by
different comparisons applied in the studies. The
main limitations other than the study design
were systematic screening not performed at
discharge, PCR result not confirmed by conven-
tional culture, lack of analysis of possible variation
in MRSA epidemiology during the study period,
absence of monitoring of compliance with decolo-
nisation treatment or contact precautions. Based
on these studies, definitive recommendations
cannot be made.
A rapid real-time PCR test that detects the
presence of vanA and/or vanB genes has been
proposed for rapid screening to identify patients
harbouring VRE at hospital admission.31 For 502
rectal swabs and stool specimens, sensitivity and
specificity was 98% and 87%, and 95% and 87%, re-
spectively. When the test is used in settings with
high prevalence of vanA isolates, results would
not require conventional culture confirmation. On
the contrary, the high number of false positives
for vanB would require culture confirmation.
In the USA, several large organisations such as
the Veterans Affairs and Evanston Northwestern
University and a few states such as Illinois, New
Jersey and Pennsylvania introduced mandatory
screening for both VRE and MRSA. In the UK it
was announced in October 2007 that all patients
admitted to National Health Service hospitals
would be screened for carriage of MRSA. Screening
was to be in place for all elective admissions by
E. Tacconelli
March 2009 and for all admissions by March 2011.
This extended the guidance issued as part of the
‘Saving Lives’ package in September 2006, which
recommended a risk-based approach to screening
to include high-risk surgical patients, renal dialysis
patients, those admitted to ICUs, high risk medical
admissions, and transfers from other hospitals,
long-term care facilities, and nursing homes. In
response to new legislation, the Society for
Healthcare Epidemiology of America (SHEA) and
the Association for Professionals in Infection
Control and Epidemiology Inc. (APIC) developed
a position statement asserting that, although
conspicuous evidence to support the use of ASCs
for high risk patients and during outbreaks is
available, there is not sufficient evidence to justify
the mandatory use of this control measure.32 The
lack of support for legislation is due to the uncer-
tainties and the potential unintended con-
sequences of legislation mandating application of
ASCs, namely: exclusion of local infection control
professionals from their role of leading risk assess-
ment and resource allocation; unresolved contro-
versies regarding the epidemiological, biological
and clinical implications of ASCs; and the poten-
tially negative effects on patients of contact
isolation.
Should contact isolation precautions be
implemented for patients colonised or
infected with epidemiologically
important bacteria?
Sixty-five percent of HCWs contaminate their
uniforms or gowns during routine care of patients
with MRSA. On >25% of occasions HCW’s clean
hands became recontaminated after contact with
their contaminated clothing. Cooper et al. demon-
strated that intensive concerted interventions,
including isolation, can substantially reduce MRSA.33
A before-and-after study evaluated the efficacy of
contact and droplet precautions in reducing the
incidence of HA-MRSA infections.34 After a baseline
period, contact and droplet precautions were
implemented in all ICUs. Reductions in the inci-
dence of HA-MRSA infection in ICUs also led to the
implementation of contact precautions in non-ICU
patient care areas. Combined rates of HA-MRSA
infection in the MICU and SICU decreased with no
significant change in other ICUs. After the discon-
tinuation of droplet precautions, the combined
rate in the MICU and SICU decreased further
although not significantly. After the implementation
of contact precautions the rate decreased signifi-
cantly in non-ICU areas.
Screening and isolation for HCAI prevention
Although >100 studies have reported control of
MRSA with screening linked to isolation and cohort-
ing, two studies suggested that reporting culture
results and isolating colonised patients had no
impact on the prevalence of HA-MRSA.35 The first
study assessed the effect of daily microbiological
surveillance alone on the spread of S. aureus.36
During a 10 week period, surveillance cultures
were performed in 158 patients. Surveillance cul-
tures and genotyping of MRSA and MSSA isolates
demonstrated the absence of cross-transmission
among patients. The second study was a prospec-
tive one-year study in the ICUs of two teaching
hospitals. Admission and weekly screens were
used to ascertain the incidence of MRSA colonisa-
tion.37 In the middle six months, MRSA-positive
patients were neither moved to a single room nor
cohort-nursed unless they were carrying other
multiresistant bacteria. Patients’ characteristics
and MRSA acquisition rates were similar in the
periods when patients were moved and not moved.
The crude (unadjusted) Cox proportional hazards
model showed no evidence of increased transmis-
sion during the non-move phase. However, the
studies have many limitations. First, the time
elapsed from admission to the results of cultures
was up to four days, and patients with a length
of hospitalisation <48 h were excluded. This time
would have been sufficient, especially in highly en-
demic situations, to allow person-to-person trans-
mission of MRSA. Second, the low number of
carers and patients might have had an important
role on the final outcome.
For VRE the role of contact precautions is still
under evaluation. Our group is currently reviewing
the available evidence within the Cochrane Wound
Group.38 Many reports showed that intensive infec-
tion control strategies including isolation enabled
the timely termination of VRE outbreaks, even
those involving VRE strains with high epidemic po-
tential on high risk wards. Premature discontinua-
tion of infection control measures caused
recurrence of VRE transmission.39
A combination of control measures including
ASCs, contact precautions for all colonised and
infected patients and antimicrobial stewardship
has also been reported to significantly reduce the
incidence of ESBL-producing bacteria.40
Contact precautions have been associated with
several adverse effects. Their application should
be preceded by accurate training of HCWs and it
should not be accepted as a substitute for hand
washing. In the MICU at Duke University Medical
Center it was observed that HCWs were approxi-
mately half as likely to enter the rooms of patients
in contact isolation.41 The results supported
375
suggestions made by other studies that patients
in isolation receive less frequent care and may
even suffer psychologically.
Decolonisation
Following positive screening, a patient should be
placed in contact precaution and decolonised. The
majority of protocols for MRSA decolonisation in-
clude mupirocin alone or in combination with body
washes, and systemic therapy with one or two
antibiotics such as rifampicin, doxycycline and tri-
methoprim-sulfamethoxazole. Less popular proto-
cols include the tea tree oil, a popular natural
antiseptic, topical gentian violet and arbekacin
inhalation in patients with percutaneous endo-
scopic gastrostomy. The duration of the decolonisa-
tion treatment is usually 7e10 days. Decolonisation
has been proven to be efficacious in reducing
infections in high risk groups.42 In dialysis patients
the use of mupirocin reduced the rate of S. aureus
infections by 68%.43 Risk reduction was higher in
peritoneal dialysis patients. Significant reduction
of S. aureus infections was also observed in patients
after cardiothoracic and orthopaedic surgery.42
Different protocols were used for VRE decoloni-
sation with poor response. After one month >50%
of the patients are recolonised.44 The most prom-
ising agent for gastrointestinal decolonisation was
ramoplanin, the first of a new class of antibiotics,
the glycolipodepsipeptides. The drug is bacteri-
cidal against Gram-positive bacteria and is not ab-
sorbed when taken orally and thus achieves high
faecal concentrations. In animal studies ramopla-
nin inhibits VRE colonisation due to re-expansion
of the colonic microflora. A randomised placebo-
controlled double-blind trial showed that VRE
decolonisation was unsuccessful in 79% of the
patients.44
Screening and decolonisation of HCWs has been
attempted for MRSA but its use remains contro-
versial.45 The prevalence of MRSA among profes-
sionals is related to geographical regions (Eastern
Europe 1.6%, Africa 15.5%), location (ICU 4.7%,
general wards 6.3%), type of HCW (nurse 8%, med-
ical 7.4%), type of room (private cohorting 2.4%, no
private/no cohorting 7.7%), and contact pre-
cautions in force (3.3%) or not applied (5.6%).46
However, little evidence suggests that the
exclusion of MRSA-positive HCWs improves the
control of HA-MRSA with the exception of hospital
outbreaks. Screening might have psychological
implications for professionals who test positive.
Cost-effectiveness analysis for this approach in
this population is also unavailable. Only staff
376
members with colonised or infected hand lesions
should be off work while receiving courses of
clearance therapy.47
Discussion
Patients and the public increasingly view rates of
HCAIs as indicators of quality of patient care.
There has been much debate about the evidence
and cost-effectiveness of various infection control
policies. A multidisciplinary approach should be
employed in all settings including: hand hygiene
programmes, ASCs, training and feedback to
HCWs, bundles for ventilator-associated pneumo-
nia and CVC-related infections and environmental
programmes. Target pathogens should be locally
defined. Programmes focused solely on MRSA might
not be the appropriate answer to impact on HCAIs.
Intensive promotion of alcohol-based hand rubs
and behavioural change interventions might rep-
resent a more cost-effective approach compared
with universal screening policies. On the grounds
of lack of evidence-based clinical and cost-effec-
tiveness data, routine screening of all admissions
to hospital is not advocated at this time.48e51
Rapid molecular tests for MRSA screening are real-
istic only when combined with adequate infection
control measures in high risk patients.
Conflict of interest statement
None.
Funding sources
None.
References
1. Boucher HW, Corey GR. Epidemiology of methicillin-
resistant Staphylococcus aureus. Clin Infect Dis 2008;46:
S344eS349.
2. Harris AD, McGregor JC, Furuno JP. What infection control
interventions should be undertaken to control multidrug-
resistant Gram-negative bacteria? Clin Infect Dis 2006;
43(Suppl. 2):S57eS61.
3. Anonymous. Healthcare-associated infections. European
Centre for Disease Prevention and Control. Available from:
http://ecdc.gov; 2006.
4. Proposal for a Council Recommendation on patient safety,
including the prevention and control of healthcare associ-
ated infections {COM(2008) 836 final} {SEC(2008) 3004}
{SEC(2008) 3005}. EC Brussels; 20 January 2009.
5. Marshall C, Wesselingh S, McDonald M, Spelman D. Control
of endemic MRSA e what is the evidence? A personal view.
J Hosp Infect 2004;56:253e268.
6. Huang SS, Platt R. Risk of methicillin-resistant Staphylococ-
cus aureus infection after previous infection or coloniza-
tion. Clin Infect Dis 2003;36:281e285.
E. Tacconelli
7. Safdar N, Bradley EA. The risk of infection after nasal colo-
nisation with Staphylococcus aureus. Am J Med 2008;12:
310e315.
8. Olivier CN, Blake RK, Steed LL, Salgado CD. Risk of vanco-
mycin-resistant enterococcus (VRE) bloodstream infection
among patients colonized with VRE. Infect Control Hosp
Epidemiol 2008;29:404e409.
9. Tacconelli E, De Angelis G, Cataldo MA, et al. Usage of car-
bapenems significantly increases the rate of new colonisa-
tion due to antibiotic-resistant bacteria in hospitalised
patients. Barcelona: ECCMID; 2008. Oral communication.
10. Buke C, Armand-Lefevre L, Lolom I, et al. Epidemiology of
multidrug-resistant bacteria in patients with long hospital
stays. Infect Control Hosp Epidemiol 2007;28:1255e1260.
11. Williams VR, Callery S, Vearncombe M, Simor AE. The role of
colonization pressure in nosocomial transmission of methi-
cillin-resistant Staphylococcus aureus. Am J Infect Control
2009;37:106e110.
12. McGinigle KL, Gourlay ML, Buchanan IB. The use of active
surveillance cultures in adult intensive care units to reduce
methicillin-resistant Staphylococcus aureus-related mor-
bidity, mortality, and costs: a systematic review. Clin Infect
Dis 2008;46:1717e1725.
13. Lucet JC, Paoletti X, Lolom I, et al. Successful long-term
program for controlling methicillin-resistant Staphylococcus
aureus in intensive care units. Intensive Care Med 2005;31:
1051e1057.
14. Clancy M, Graepler A, Wilson M, Douglas I, Johnson J,
Price CS. Active screening in high-risk units is an effective
and cost-avoidant method to reduce the rate of methicil-
lin-resistant Staphylococcus aureus infection in the hospi-
tal. Infect Control Hosp Epidemiol 2006;27:1009e1017.
15. Vos MC, Ott A, Verbrugh HA. Successful search-and-destroy
policy for methicillin-resistant Staphylococcus aureus in
The Netherlands. J Clin Microbiol 2005;43:2034.
16. Bootsma MC, Diekmann O, Bonten MJ. Controlling methicil-
lin-resistant Staphylococcus aureus: quantifying the effects
of interventions and rapid diagnostic testing. Proc Natl Acad
Sci USA 2006;103:5620e5625.
17. Wernitz MH, Keck S, Swidsinski S, Schulz S, Veit SK. Cost
analysis of a hospital-wide selective screening programme
for methicillin-resistant Staphylococcus aureus (MRSA)
carriers in the context of diagnosis related groups (DRG)
payment. Clin Microbiol Infect 2005;11:466e471.
18. Montecalvo MA, Jarvis WR, Uman J, et al. Costs and savings
associated with infection control measures that reduced
transmission of vancomycin-resistant enterococci in an
endemic setting. Infect Control Hosp Epidemiol 2001;22:
437e442.
19. Tacconelli E, Venkataraman L, De Girolami PC, D’Agata EM.
Methicillin-resistant Staphylococcus aureus bacteraemia
diagnosed at hospital admission: distinguishing between
community-acquired versus healthcare-associated strains.
J Antimicrob Chemother 2004;53:474e479.
20. Harbarth S, Sax H, Fankhauser-Rodriguez C, Schrenzel J,
Agostinho A, Pittet D. Evaluating the probability of previ-
ously unknown carriage of MRSA at hospital admission. Am
J Med 2006;119:275. e15e23.
21. Tacconelli E, Karchmer AW, Yokoe D, D’Agata EM. Prevent-
ing the influx of vancomycin-resistant enterococci into
health care institutions, by use of a simple validated predic-
tion rule. Clin Infect Dis 2004;39:964e970.
22. Carroll MC. Rapid diagnostics for methicillin-resistant
Staphylococcus aureus: current status. Mol Diagn Ther
2008;12:15e24.
23. Conterno LO, Shymanski J, Ramotar K, et al. Real-time
polymerase chain reaction detection of methicillin-
Screening and isolation for HCAI prevention
resistant Staphylococcus aureus: impact on nosocomial
transmission and costs. Infect Control Hosp Epidemiol
2007;28:1134e1141.
24. Cunningham R, Jenks P, Northwood J, Wallis M, Ferguson S,
Hunt S. Effect on MRSA transmission of rapid PCR testing of
patients admitted to critical care. J Hosp Infect 2007;65:
24e28.
25. Harbarth S, Masuet-Aumatell C, Schrenzel J, et al. Evalua-
tion of rapid screening and pre-emptive contact isolation
for detecting and controlling methicillin-resistant Staphylo-
coccus aureus in critical care: an interventional cohort
study. Crit Care 2006;10:R25.
26. Jeyaratnam D, Whitty CJ, Phillips K, et al. Impact of rapid
screening tests on acquisition of meticillin resistant Staph-
ylococcus aureus: cluster randomised crossover trial. Br
Med J 2008;336:927e930.
27. Jog S, Cunningham R, Cooper S, et al. Impact of reoperative
screening for meticillin-resistant Staphylococcus aureus by
real-time polymerase chain reaction in patients undergoing
cardiac surgery. J Hosp Infect 2008;69:124e130.
28. Keshtgar MR, Khalili A, Coen PG, et al. Impact of rapid mo-
lecular screening for meticillin-resistant Staphylococcus au-
reus in surgical wards. Br J Surg 2008;95:381e386.
29. Robicsek A, Beaumont JL, Paule SM, et al. Universal surveil-
lance for methicillin-resistant Staphylococcus aureus in 3
affiliated hospitals. Ann Intern Med 2008;148:409e418.
30. Harbarth S, Fankhauser C, Schrenzel J, et al. Universal
screening for methicillin-resistant Staphylococcus aureus
at hospital admission and nosocomial infection in surgical
patients. J Am Med Assoc 2008;299:11.
31. Stamper PD, Cai M, Lema C, Eskey K, Carroll KC. Comparison
of the BD GeneOhm VanR assay to culture for identification
of vancomycin-resistant enterococci in rectal and stool
specimens. J Clin Microbiol 2007;45:3360e3365.
32. Weber SG, Huang SS, Oriola S, et al. Legislative mandates for
use of active surveillance cultures to screen for meti-
cillin-resistant Staphylococcus aureus and vancomycin-
resistant enterococci: position statement from the Joint SHEA
and APIC Task Force. Am J Infect Control 2007;35:73e85.
33. Cooper BS, Stone SP, Kibbler CC, et al. Isolation measures in
the hospital management of methicillin resistant Staphylo-
coccus aureus (MRSA): systematic review of the literature.
Br Med J 2004;329:533e538.
34. Mangini E, Segal-Maurer S, Burns J, et al. Impact of contact
and droplet precautions on the incidence of hospital-
acquired methicillin-resistant Staphylococcus aureus infec-
tion. Infect Control Hosp Epidemiol 2007;28:1261e1266.
35. Cepeda JA, Whitehouse T, Cooper B, et al. Isolation of pa-
tients in single rooms or cohorts to reduce spread of MRSA
in intensive-care units: prospective two-centre study. Lan-
cet 2005;365:295e304.
36. Nijssen S, Bonten MJ, Weinstein RA. Are active microbiolog-
ical surveillance and subsequent isolation needed to pre-
vent the spread of methicillin-resistant Staphylococcus
aureus? Clin Infect Dis 2005;40:405e409.
377
37. Farr BM. What to think if the results of the National Insti-
tutes of Health randomized trial of methicillin-resistant
Staphylococcus aureus and vancomycin-resistant entero-
coccus control measures are negative (and other advice to
young epidemiologists): a review and an au revoir. Infect
Control Hosp Epidemiol 2006;27:1096e1106.
38. Tacconelli E, Cauda R, Cataldo MAA, Carmeli Y, De
Angelis G. Control interventions for preventing spread
of vancomycin-resistant enterococci (VRE) in hospitals.
Cochrane Database Syst Rev 2008;(4).
39. Schmidt-Hieber M, Blau IW, Schwartz S, et al. Intensified
strategies to control vancomycin-resistant enterococci in
immunocompromised patients. Int J Hematol 2007;86:
158e162.
40. Conterno LO, Shymanski J, Ramotar K, Toye B, Zvonar R,
Roth V. Impact and cost of infection control measures to re-
duce nosocomial transmission of extended-spectrum beta-
lactamase-producing organisms in a non-outbreak setting.
J Hosp Infect 2007;65:354e360.
41. Kirkland KB, Weinstein JM. Adverse effects of contact isola-
tion. Lancet 1999;354:1177e1178.
42. Kallen AJ, Wilson CT, Larson RJ. Perioperative intranasal
mupirocin for the prevention of surgical-site infections:
systematic review of the literature and meta-analysis.
Infect Control Hosp Epidemiol 2005;26:916e922.
43. Tacconelli E, Carmeli Y, Aizer A, Ferreira G, Foreman MG,
D’Agata EM. Mupirocin prophylaxis to prevent Staphylococ-
cus aureus infection in patients undergoing dialysis: a meta-
analysis. Clin Infect Dis 2003;37:1629e1638.
44. Wong MT, Kauffman CA, Standiford HC, et al. Ramoplanin
VRE2 Clinical Study Group. Effective suppression of vanco-
mycin-resistant Enterococcus species in asymptomatic gas-
trointestinal carriers by a novel glycolipodepsipeptide,
ramoplanin. Clin Infect Dis 2001;33:1476e1482.
45. Simpson AH, Dave J, Cookson B. The value of routine
screening of staff for MRSA. J Bone Joint Surg Br 2007;89:
565e566.
46. Albrich WC, Harbarth S. Health-care workers: source, vec-
tor, or victim of MRSA? Lancet Infect Dis 2008;8:289e301.
47. Nathwani D, Morgan M, Masterton RG, et al. British Society
for Antimicrobial Chemotherapy Working Party on commu-
nity-onset MRSA Infections. Guidelines for UK practice for
the diagnosis and management of meticillin-resistant
Staphylococcus aureus (MRSA) infections presenting in the
community. J Antimicrob Chemother 2008;61:976e994.
48. Diekema DJ, Climo M. Preventing MRSA infections: finding it
is not enough. J Am Med Assoc 2008;299:1190e1192.
49. Gould IM. Control of methicillin-resistant Staphylococcus
aureus in the UK. Eur J Clin Microbiol Infect Dis 2005;24:
789e793.
50. Dancer SJ. Considering the introduction of universal MRSA
screening. J Hosp Infect 2008;69:315e320.
51. Tacconelli E. Methicillin-resistant Staphylococcus aureus:
risk assessment and infection control policies. Clin
Microbiol Infect 2008;14:407e410.
Journal of Hospital Infection (2009) 73, 378e385
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

The role of environmental cleaning in the control

of hospital-acquired infection
S.J. Dancer*

Department of Microbiology, Hairmyres Hospital, East Kilbride, UK

Available online 1 September 2009

KEYWORDS Summary Increasing numbers of hospital-acquired infections have gener-

Acinetobacter; ated much attention over the last decade. The public has linked the so-
Environmental called ‘superbugs’ with their experience of dirty hospitals but the precise
cleaning;
role of environmental cleaning in the control of these organisms remains
C. difficile;
unknown. Until cleaning becomes an evidence-based science, with
Hospital-associated
infection; established methods for assessment, the importance of a clean environment
Infection control; is likely to remain speculative. This review will examine the links between
MRSA; the hospital environment and various pathogens, including meticillin-resis-
Norovirus; tant Staphylococcus aureus, vancomycin-resistant enterococci, norovirus,
VRE Clostridium difficile and acinetobacter. These organisms may be able to sur-
vive in healthcare environments but there is evidence to support their vulner-
ability to the cleaning process. Removal with, or without, disinfectants,
appears to be associated with reduced infection rates for patients. Unfortu-
nately, cleaning is often delivered as part of an overall infection control pack-
age in response to an outbreak and the importance of cleaning as a single
intervention remains controversial. Recent work has shown that hand-touch
sites are habitually contaminated by hospital pathogens, which are then de-
livered to patients on hands. It is possible that prioritising the cleaning of
these sites might offer a useful adjunct to the current preoccupation with
hand hygiene, since hand-touch sites comprise the less well-studied side of
the hand-touch site equation. In addition, using proposed standards for
hospital hygiene could provide further evidence that cleaning is a cost-effec-
tive intervention for controlling hospital-acquired infection.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

Introduction
* Address: Department of Microbiology, Hairmyres Hospital,
Eaglesham Road, East Kilbride G75 8RG, UK. Tel.: þ44 1355
585000; fax: þ44 1355 584350. There has been much debate over hospital clean-
E-mail address: stephanie.dancer@lanarkshire.scot.nhs.uk liness and increasing numbers of hospital-acquired
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.03.030
Environmental cleaning and HAI
infections (HAIs). The public have linked their
visual experience of dirty hospitals with the risk
of HAI but there is little evidence to support this at
present.1 Cleaning has never been regarded as an
evidence-based science and consequently receives
little attention from the scientific community.
Since there are no scientific standards to measure
the effect of an individual cleaner, or assess en-
vironmental cleanliness, finding the evidence for
benefit in the control of infection is further ham-
pered.2 There are always basic aesthetic consider-
ations that cannot be disregarded; a perception of
cleanliness, however defined, is expected for
patients, their relatives and staff from healthcare
environments.
Cleaning is routinely monitored by visual audit
in the UK. Looking to see if a ward is clean may
fulfil aesthetic obligations but it does not provide
a reliable assessment of the infection risk for an
individual patient on that ward.3 The organisms
that cause infection are invisible to the naked
eye and their existence is not necessarily associ-
ated with the presence of visual dirt. Furthermore,
the impression of cleanliness is confounded by
clutter, and fabric and maintenance deficits.4
Visual assessment will inevitably be subject to
bias under these circumstances. It is more difficult
to clean a crowded, cluttered environment,
perhaps related to a cleaner’s incentive, when
confronted with peeling plaster, cracked tiles or
worn floor coverings.4
Sites that are frequently touched by hands are
thought to provide the greatest risk for patients,
and those situated right beside patients provide
the biggest risk of all.5e7 The responsibility for
cleaning near-patient hand-touch sites does not
always rest with the ward cleaners, however, since
beds, drip stands, lockers and overbed tables are
more usually cleaned by nurses.7,8 Nurses are
also responsible for the decontamination of more
delicate clinical equipment. This overlapping of
cleaning responsibilities has created some confu-
sion; it has also meant that cleaning opportunities
of some items are missed or abandoned.9,10
The microbial pathogens that cause HAI have
two special properties: first, they are recognised
as hospital pathogens; second, they have an innate
ability to survive on surfaces in the hospital
environment for long periods of time.1 They in-
clude organisms such as meticillin-resistant Staph-
ylococcus aureus (MRSA), Clostridium difficile,
vancomycin-resistant enterococci (VRE), Acineto-
bacter spp. and norovirus.1,6 This mini-review
will summarise the evidence for the presence
and survival of these organisms in the clinical en-
vironment as well as support for cleaning as a valid
379
infection control intervention for patients. There
will also be some discussion on the measurement
of ‘cleanliness’ of the healthcare environment
and why this is important for future work evaluat-
ing the role of hospital cleaning and HAI.
Meticillin-resistant Staphylococcus
aureus
MRSA resists desiccation and can survive in hospital
dust for up to a year.11 It is found throughout the
hospital environment, particularly around patients
known to be colonised or infected with the bac-
terium. Molecular fingerprinting of these strains
shows that MRSA-positive patients tend to shed
their own strain of MRSA into the near-patient
environment.12 If staff enter a room containing
an MRSA patient, two-thirds of them will acquire
the patient’s strain on gloved hands or apron.12
Even if they do not touch the patient directly,
four in ten will still exit the room carrying the
patient’s strain of MRSA on hands or apron.12
MRSA can be found on general surfaces such as
floors and radiators, furniture such as beds and
lockers, and clinical equipment.7e10,12e14 Some
sites, e.g. linen, curtains, beds, lockers and
overbed tables, tend to harbour MRSA more
frequently than others.7,8,13 It is thought that
contamination of near-patient hand-touch sites
provides the biggest risk of MRSA acquisition for
patients.2,6 In addition, there is a small but signif-
icant increase in the risk of acquiring MRSA if a pa-
tient is admitted into a room previously occupied
by carrier patients.15
There is some evidence that cleaning removes
MRSA from the ward environment with benefit for
patients.6,16 An outbreak of MRSA lingered for sev-
eral months on a urological ward, resisting all the
usual infection control interventions such as pro-
motion of hand hygiene and isolation of patients.16
The investigating team found the outbreak strain
of MRSA scattered throughout the ward environ-
ment and doubled the number of domestic clean-
ing hours from 60 per week to 120. Following
this, there was no further isolation of the outbreak
strain from the environment and the number of pa-
tients affected decreased immediately. The clean-
ing intervention was thought to have played
a significant role in the termination of the out-
break and was estimated to have saved at least
£28,000.16
Another outbreak of glycopeptide-intermediate
S. aureus (GISA) in an intensive therapy unit
proved difficult to control until a wave of further
control measures, including enhanced cleaning,
380
was introduced.17 The outbreak encompassed two
patient clusters, although genotyping showed that
all cases were caused by the same strain. The
second cluster occurred despite the institution of
maximum contact-isolation procedures. This di-
rected attention towards the inanimate environ-
ment as the major source of contamination, since
it was thought that re-emergence of the strain
could be explained by its ability to survive on inert
surfaces. The meticulous cleaning procedures
finally implemented probably helped to stop the
outbreak, although it was not possible to deter-
mine the relative roles of barrier precautions and
environmental decontamination in eradicating
the strain.17
Outwith the outbreak situation, another study
examined the effects of enhanced cleaning on two
matched surgical wards in a controlled cross-over
trial for two six-month periods.18 There were nine
ward-acquired MRSA infections during routine
cleaning periods, but only four when the wards re-
ceived extra cleaning, notably targeting hand-
touch sites and clinical equipment. More MRSA pa-
tient-days during the enhanced cleaning periods
predicted at least thirteen new cases instead of
the four that actually occurred. The study con-
cluded that targeted cleaning using detergent
wipes and water could be a cost-effective mech-
anism of reducing MRSA infection on a surgical
ward.18
Vancomycin-resistant enterococci
VRE are not credited with the same degree of
pathogenicity as MRSA, but they may still cause
infections in vulnerable patients, including out-
breaks that are difficult to control. In addition, the
vanA gene has been shown to transfer to MRSA,
making the latter even more difficult to treat.19
Part of the problem of controlling VRE in hospitals
is due to their extreme longevity in the hospital
environment and their resistance to routine clean-
ing.1 Even bleach-based cleaning sometimes fails
to eradicate these hardy organisms.20 Contact
with contaminated surfaces in the rooms of colon-
ised patients results in transfer of VRE to gloved
hands, despite cleaning with disinfectants.5,21,22
In addition, patients admitted into rooms previ-
ously accommodating patients with VRE are them-
selves more likely to acquire the organism.15
It has already been suggested that environmen-
tal cleaning might be important in the control of
VRE.23 Additional work has highlighted its presence
on hand-touch sites near the patient, just as for
MRSA.10,22 A recent paper describes the impact of
S.J. Dancer
improved environmental cleaning on the spread
of VRE in a medical intensive care unit (ICU),
with and without promotion of hand hygiene com-
pliance.24 In this study, there were four periods:
the first served as a baseline period without any
interventions; the second began after a 30 day
period during which there was an education and
intensified observation programme to improve
cleaning; the third served as a ‘washout’ period
without further interventions; and the fourth in-
cluded an intensive hand hygiene campaign. The
study found that enforcing cleaning measures
along with encouraging hand hygiene was associ-
ated with less surface contamination with VRE,
cleaner healthcare worker hands and a significant
reduction in VRE cross-transmission among
patients. The authors concluded that decreasing
environmental contamination might help to
control the spread of VRE in hospitals.24
Clostridium difficile
C. difficile is a spore-forming anaerobe that has
been recovered in abundance from the environ-
ment of symptomatic patients.1,25 The more
patients there are with C. difficile on the ward,
the more likely it is that other patients will also ac-
quire the organism.26 Environmental contamina-
tion with spores is now well accepted as a risk
factor for the acquisition of C. difficile.27 Further-
more, as the level of environmental contamination
increases, so does the amount of C. difficile on the
hands of healthcare workers, and once again near-
patient hand-touch sites are regarded as a particu-
lar risk.25,28
Having established that there is a dynamic
transmission cycle for C. difficile similar to that
for MRSA, there is additional evidence to support
the value of cleaning in its control.1,29e31 Follow-
ing a rise in C. difficile cases, one hospital intro-
duced enhanced cleaning with hypochlorite into
two ICUs.31 One of the ICUs applied the extra
cleaning to all areas, including rooms used solely
by staff, and sensitive clinical equipment was
wiped over twice daily using hypochlorite-impreg-
nated cloths. The other unit introduced the inten-
sive hypochlorite clean into isolation rooms
housing patients already infected with C. difficile.
Rates of infection decreased in both units over sev-
eral months and appeared to be maintained at
a lower rate for at least two years after the clean-
ing intervention, despite some relaxation of the
initial regimen.31
Infection control teams do not question the
importance of thorough environmental cleaning
Environmental cleaning and HAI
for controlling C. difficile, although whether
cleaners should use disinfectants or detergents
for routine cleaning continues to be hotly de-
bated.25,29,30 Although bleach cleaning certainly
appears to have an effect on both environmental
contamination and patient acquisition of C. diffi-
cile, it could be the increased physical effort of
enhanced cleaning that contributes towards the
overall result rather than a direct effect of the
bleach itself.30 It is known that cleaning of toilets
and other sites using appropriate disinfectants
does not totally remove all traces of C. diffi-
cile.27,30,32,33 It is likely that agreement on the
best way of removing C. difficile from the environ-
ment will elude us for some time to come.
Acinetobacter
Acinetobacter can also be recovered from the
hospital environment with ease, including inani-
mate hand-touch sites near the patient.34 Seeding
clinical and environmental strains onto Formica
surfaces demonstrated survival of between one
and two weeks, although some strains are known
to survive for much longer.35 The importance of
cleaning in controlling outbreaks of Acinetobacter
spp. has been emphasised in previous studies.36,37
One of these describes an outbreak caused by mul-
tiply resistant strains of A. baumannii involving
more than thirty patients in two ICUs.37 Environ-
mental contamination was recognised as an impor-
tant reservoir of the epidemic strains and the
outbreak ceased only after both ICUs were closed
for terminal cleaning and disinfection.37
Another study examined the levels of environmental
contamination with acinetobacter in a neurosurgical
ICU during a prolonged outbreak.38 As with MRSA
and C. difficile, there were many near-patient
hand-touch sites that yielded the epidemic strain.
This study also demonstrated a significant associa-
tion between the amount of environmental con-
tamination and patient colonisation. The
conclusion was that high standards of cleaning
play an integral role in controlling outbreaks of
acinetobacter in the intensive care setting, al-
though once again, little is known about the best
way to clean in non-outbreak settings.38
A further study describes what happened fol-
lowing the introduction of bedside computers in
a paediatric burns ward.39 There was a sudden
increase in the number of patients acquiring acine-
tobacter and environmental screening demon-
strated the organism on various surfaces in the
patients’ rooms, especially the plastic covers
over the bedside computer keyboards. Targeted
381
infection control measures that included the use
of gloves before using the computer and thorough
disinfection of the plastic covers effectively termi-
nated the outbreak. Up until the outbreak oc-
curred, no one had thought to include the
computer keyboards in a routine cleaning specifi-
cation.39 Similarly, during a period of high en-
demicity of acinetobacter in another burns unit,
the endemic strain was identified from surfaces
close to the patient, including bed linen, medica-
tion table and the display surface of an overbed
monitor.40
Norovirus
Norovirus can be found on a huge variety of surfaces
both in hospitals and in the community.41e43 Several
studies cite the association of norovirus with hand-
touch sites such as toilet taps, door-handles, hospi-
tal equipment, elevator and microwave buttons,
switches and telephones.41,42,44 When fingers
come into contact with virus-contaminated ma-
terial, norovirus is consistently transferred to
typical hand-touch sites.44
The importance of environmental cleaning in
the control of outbreaks of norovirus is widely
accepted.41,45 This includes clinical equipment, as
well as floors, toilets and general surfaces. Cur-
tains should be removed and sent to the laundry,
and other soft furnishings either washed down or,
preferably, steam-cleaned. All general cleaning,
especially the toilets and bathroom areas, should
be with a chlorine-containing disinfectant or
bleach at a specified concentration. Cleaning poli-
cies should always include the use of these disin-
fectants since detergent-based cleaning often
fails to eradicate the virus from the environ-
ment.44 Without scrupulous attention to the en-
vironment, outbreaks not only continue, but will
resume within a short space of time. This is
because norovirus survives in the warm clinical
environment with ease while retaining its infectivity,
readily transferring to patients from hands that
have just touched a contaminated site.
Outside hospitals, norovirus outbreaks can be
devastating in closed or semi-closed communi-
ties.46 These include sudden and extensive out-
breaks in hotels or cruise-liners, but outbreaks can
also occur in nursing and residential homes, prisons
and schools. An outbreak reported recently in a pri-
mary school involved 79 pupils and 24 members of
staff.47 Subsequent investigation of the outbreak
showed that person-to-person contact was a major
factor in the transmission of the virus, but there
was evidence that poorly cleaned computer
382
equipment was also implicated. A strain of noro-
virus, identical by RNA sequencing to two strains
retrieved from patients, was isolated from one com-
puter keyboard and mouse in one particular class-
room. This occurred even after instituting
a bleach clean the previous day. Public health offi-
cials recommended good hand-washing practices,
exclusion of symptomatic persons and thorough en-
vironmental disinfection with a diluted (1:50 con-
centration) household bleach solution, to include
sites that are shared but not commonly cleaned.47
Discussion
There is plenty of evidence supporting the role of
domestic cleaning in hospitals as an important
intervention in the control of HAI. Unfortunately,
it often constitutes part of an overall infection
control package in response to an outbreak and its
importance as a stand-alone activity remains
controversial. This does not encourage on-going
managerial support for cleaning services in the
hospital, particularly if resources are limited. The
situation is hampered by the lack of scientific
standards for hospital cleaning, rather than the
subjective visual assessment practised at present.
The following three sections argue for a measured,
targeted and methodological approach to domes-
tic cleaning in hospitals.
What is ‘clean’?
If we state that a hospital is clean, we assume that
it looks clean and that it is safe for patients.
Unfortunately, the microbes responsible for HAI
are invisible to the naked eye. This means that
visual assessment is insufficient for defining clean-
liness, nor will it accurately predict the infection
risk for patients.2 Cleanliness should not actually
be defined without indicating how it is assessed.
A recent study compared visual assessment against
both biochemical (ATP bioluminescence) and mi-
crobiological screening of the hospital environ-
ment.3 The results showed that whereas most
surfaces looked clean, less than a quarter were
free from organic soil (ATP) and less than half
were microbiologically clean.3 Given the risk of ac-
quiring hospital pathogens from a hospital ward,
visual assessment is outdated, inadequate and sci-
entifically obsolete. The only benefit from a visual
inspection is to appease aesthetic obligations.
There has been suggestion that hospitals would
benefit from cleaning standards emulating those
implemented in the food industry.2,48 Food prep-
aration surfaces are subjected to routine sampling
S.J. Dancer
using a range of techniques in a coordinated and in-
tegrated approach.48 Any isolation of pathogens, or
pathogen indicators, causes concern and warrants
immediate action. By contrast, environmental sur-
face sampling in hospitals usually only takes place
in response to an outbreak e and then only if the
infection control team responsible has the motive,
means and interest to initiate environmental screen-
ing.48 It is time that high risk surfaces were subjected
to routine screening in order to monitor overall
levels of microbial dirt and the results used to gener-
ate increased or targeted cleaning before a hospital
outbreak occurs. It is false economy to wait until
an outbreak occurs before the clinical environment
receives the attention it deserves.
There is no reason why cleaning should not
become an evidence-based science, particularly
when we know that lack of it is associated with
human infection. Microbiological standards have
been proposed, using long-established principles
from the food industry as well as from standards
governing media such as air and water.2 Since the
pathogens of interest are widely scattered in time
and space, these standards make use of quantita-
tive and qualitative microbial indicators rather
than focus on trying to find a discrete pathogen.
Practical applications for high risk surfaces in hospi-
tals would not be difficult to institute although the
finer details related to risk, and site, require further
evaluation.2,49 Surfaces in outpatient corridors do
not present a similar risk of infection as hand-touch
sites might, situated beside a ventilated patient in
an ICU. Attempts have already been made to com-
pare benchmark visual, ATP bioluminescence and
microbiological values against each other, and one
study has attempted to evaluate microbiological
standards against infection risk in an ICU.8,50
Where to clean?
There is increasing evidence regarding the impor-
tance of hand-touch sites in the transmission of
pathogens to healthy persons, as well as to
patients.51,52 It is also becoming apparent that
the sites closer to the patient are more likely to
furnish an infection risk than those situated fur-
ther away.7,8 The role of these near-patient
hand-touch sites in MRSA transmission and, indeed,
other hospital pathogens, has not been given the
priority that it deserves. Ward cleaners work to
a set specification that encompasses general
surfaces and bathrooms, with emphasis on the
cleaning of floors and toilets.53 These are not nec-
essarily near-patient hand-touch sites. Examples
of the latter include bed rails, bedside lockers,
infusion pumps, door handles and various switches,
Environmental cleaning and HAI
including the nurse-call button, and they rarely
feature in the domestic specification.2,7,8
It is already known that traditional sites such as
toilets, general surfaces and sinks tend to attract
high rates of cleaning but that hand-touch sites,
which are more likely to harbour and transmit
microbial pathogens, are only poorly cleaned.10,54
The responsibility for cleaning many hand-touch
sites usually rests with the ward nurses, who are
often very busy and almost permanently under-
staffed in many hospitals. Two recent studies in
ICUs have demonstrated an increased risk of infec-
tion following periods of inadequate nurse staffing,
or conversely, excess workload.55,56 It may be that
concentrating available cleaning resources on
high-risk hand-touch sites would be the most
cost-effective cleaning strategy at the present
time.18 A recent study has shown that it is possible
to improve environmental cleaning following an
educational campaign and feedback on adequacy
of discharge cleaning.10
How to clean?
Most of the studies describing the benefits from
cleaning in this review used disinfectants to clean
the hospital environment. Virtually all were re-
ported as part of the response to an outbreak. Only
a few UK-based studies used detergent and water,
and even fewer reported cleaning benefits in the
absence of an outbreak.18,30 It appears that when
reviewing the evidence for the role of cleaning in
the control of HAI, there are several issues which
still require clarification. First, is there any differ-
ence between the quantity, quality and methods
for routine cleaning compared with what is needed
in the event of an outbreak; and second, is it suf-
ficient to proclaim the benefits from cleaning with
disinfectants without establishing what can be
achieved using soap and water alone? These ques-
tions require an evidence-based approach before
we can set the best specification for cleaning in
our hospitals. In addition, no one has yet modelled
different cleaning methods against the infection
risk for patients, their degree of vulnerability and
the clinical area in which they are exposed.
Most hospitals have their own domestic specifi-
cations for wards, operating theatres, outpatient
and non-clinical areas. More countries are pro-
ducing national standards for environmental clean-
ing. These set a necessary and valuable precedent
but they are not based on sufficient scientific
evidence to justify their contents. At the beginning
of the twenty-first century, we simply do not know
how to clean our hospitals in order to create the
safest environment for patient care e hence
383
a veritable blunderbuss approach to cleaning in
the event of an outbreak and managerial reluc-
tance to protect, let alone prioritise, routine
standards of cleaning outwith the outbreak situa-
tion. There is a lot of work still to do to establish
cleaning as an evidence-based science, and to
translate the evidence into cleaning practices in
healthcare environments. Given the importance of
controlling HAI, it makes sense to support current
cleaning practises until such time as sufficient
evidence is forthcoming.6
There is plenty of evidence to support basic
cleaning in hospitals. More and better-directed
cleaning reduces the risk of acquisition of a variety
of hospital organisms. It might also affect the
number of patients with these so-called ‘super-
bugs’ in the community, since hospital acquisition
invariably leads to patients taking them home.57
More interest in basic hygiene is warranted be-
cause microbes are becoming increasingly resis-
tant due to the inappropriate, inadequate and
uncontrolled use of antibiotics and there are fewer
and fewer agents in the developmental pipeline.58
A century ago, people died from trivial wounds
because there were no antibiotics.
Simple hygiene could be our only defence when
Darwinian evolution finally terminates the antibi-
otic era.59 We need to raise the level of awareness
regarding hygiene, and its importance, throughout
society and it is hospitals that should lead the way.
If they at least start by cleaning up the way they
could, then this would at least offer some sem-
blance of safety to patients.
Conflict of interest statement
The author has received research funding from
UNISON for studies on hospital cleaning.
Funding sources
None.
References
1. Dancer SJ. Mopping up hospital infection. J Hosp Infect
1999;43:85e100.
2. Dancer SJ. How do we assess hospital cleaning? A proposal
for microbiological standards for surface hygiene in hospi-
tals. J Hosp Infect 2004;56:10e15.
3. Griffith CJ, Cooper RA, Gilmore J, Davies C, Lewis M. An
evaluation of hospital cleaning regimes and standards. J
Hosp Infect 2000;45:19e28.
4. Maurer IM. Hospital hygiene. 3rd edn. Bristol: Wright PSG;
1985.
5. Bhalla A, Pultz NJ, Gries DM, et al. Acquisition of nosoco-
mial pathogens on hands after contact with environmental
surfaces near hospitalised patients. Infect Control Hosp
Epidemiol 2004;25:164e167.
384
6. Dancer SJ. Importance of the environment in meticillin-
resistant Staphylococcus aureus acquisition: the case for
hospital cleaning. Lancet Infect Dis 2008;8:101e113.
7. Dancer SJ, White L, Robertson C. Monitoring environmental
cleanliness on two surgical wards. Int J Environ Health Res
2008;18:357e364.
8. White L, Dancer SJ, Robertson C, McDonald J. Are hygiene
standards useful in assessing infection risk? Am J Infect
Control 2008;36:381e384.
9. Blythe D, Keenlyside D, Dawson SJ, Galloway A. Environ-
mental contamination due to methicillin-resistant Staphy-
lococcus aureus (MRSA). J Hosp Infect 1998;38:67e70.
10. Goodman ER, Platt R, Bass R, Onderdonk AB, Yokoe DS,
Huang SS. Impact of an environmental cleaning intervention
on the presence of methicillin-resistant Staphylococcus au-
reus and vancomycin-resistant enterococci on surfaces in
intensive care unit rooms. Infect Control Hosp Epidemiol
2008;29:593e599.
11. Wagenvoort JH, Sluijsmans W, Penders RJ. Better environ-
mental survival of outbreak vs. sporadic MRSA isolates. J
Hosp Infect 2000;45:231e234.
12. Boyce JM, Potter-Bynoe G, Chenevert C, King T. Environ-
mental contamination due to methicillin-resistant Staphy-
lococcus aureus: possible infection control implications.
Infect Control Hosp Epidemiol 1997;18:622e627.
13. Lemmen SW, Hafner H, Zolldan D, Stanzel S, Lutticken R.
Distribution of multi-resistant Gram-negative versus Gram-
positive bacteria in the hospital inanimate environment. J
Hosp Infect 2004;56:191e197.
14. Hardy KJ, Oppenheim BA, Gossain S, Gao F, Hawkey PM. A
study of the relationship between environmental contami-
nation with methicillin-resistant Staphylococcus aureus
(MRSA) and patients’ acquisition of MRSA. Infect Control
Hosp Epidemiol 2006;27:127e132.
15. Huang SS, Datta R, Platt R. Risk of acquiring antibiotic-resis-
tant bacteria from prior room occupants. Arch Intern Med
2006;166:1945e1951.
16. Rampling A, Wiseman S, Davis L, et al. Evidence that hospital
hygiene is important in the control of methicillin-resistant
Staphylococcus aureus. J Hosp Infect 2001;49:109e116.
17. de Lassence A, Hidri N, Timsit JF, et al. Control and out-
come of a large outbreak of colonization and infection with
glycopeptide-intermediate Staphylococcus aureus in an in-
tensive care unit. Clin Infect Dis 2006;42:170e178.
18. Dancer SJ, White LF, Lamb J, Girvan EK, Robertson C. Mea-
suring the effect of enhanced cleaning in a UK hospital:
a prospective cross-over study. BMC Infect Med 2009;7:28.
19. Sievert DM, Rudrik JT, Patel JB, McDonald LC, Wilkins MJ,
Hageman JC. Vancomycin-resistant Staphylococcus aureus
in the United States, 2002e2006. Clin Infect Dis 2008;46:
668e674.
20. Noble MA, Isaac-Renton JL, Bryce EA, et al. The toilet as
a transmission vector of vancomycin-resistant enterococci.
J Hosp Infect 1998;40:237e241.
21. Ray AJ, Hoyen CK, Das SM, Taub TF, Eckstein EC, Donskey CJ.
Nosocomial transmission of vancomycin-resistant enterococci
from surfaces. J Am Med Assoc 2002;287:1400e1401.
22. Hayden MK, Blom DW, Lyle EA, Moore CG, Weinstein RA.
Risk of hand or glove contamination after contact with pa-
tients colonized with vancomycin-resistant enterococcus or
the colonized patients’ environment. Infect Control Hosp
Epidemiol 2008;29:149e154.
23. Martinez JA, Ruthazer R, Hansjosten K, Barefoot L,
Snydman DR. Role of environmental contamination as a risk
factor for acquisition of vancomycin-resistant enterococci
in patients treated in a medical intensive care unit. Arch
Intern Med 2003;163:1905e1912.
S.J. Dancer
24. Hayden MK, Bonten MJM, Blom DW, Lyle EA, van de
Vijver DAMC, Weinstein RA. Reduction in acquisition of van-
comycin-resistant enterococcus after enforcement of rou-
tine environmental cleaning measures. Clin Infect Dis
2006;42:1552e1560.
25. Verity P, Wilcox MH, Fawley W, Parnell P. Prospective evalu-
ation of environmental contamination by Clostridium diffi-
cile in isolation side rooms. J Hosp Infect 2001;49:204e209.
26. Dubberke ER, Reske KA, Olsen MA, et al. Evaluation of Clos-
tridium difficile-associated disease pressure as a risk factor
for C. difficile-associated disease. Arch Intern Med 2007;
167:1092e1097.
27. Kaatz GW, Gitlin SD, Schaberg DR, et al. Acquisition of Clos-
tridium difficile from the hospital environment. Am J Epi-
demiol 1998;127:1289e1294.
28. Samore MH, Venkatamaran L, DeGirolami PC, Arbeit RD,
Karchmer AW. Clinical and molecular epidemiology of spor-
adic and clustered cases of nosocomial Clostridium difficile
diarrhoea. Am J Med 1996;100:32e40.
29. Mayfield JL, Leet T, Miller J, Mundy LM. Environmental con-
trol to reduce transmission of Clostridium difficile. Clin
Infect Dis 2000;31:995e1000.
30. Wilcox MH, Fawley WN, Wrigglesworth N, Parnell P, Verity P,
Freeman J. Comparison of the effect of detergent versus
hypochlorite cleaning on environmental contamination
and incidence of Clostridium difficile infection. J Hosp
Infect 2003;54:109e114.
31. McMullen KM, Zack J, Coopersmith CM, Kollef M,
Dubberke E, Warren DK. Use of hypochlorite solution to de-
crease rates of Clostridium difficile-associated diarrhoea.
Infect Control Hosp Epidemiol 2007;28:205e207.
32. Eckstein BC, Adams DA, Eckstein EC, et al. Reduction of Clos-
tridium difficile and vancomycin-resistant Enterococcus con-
tamination of environmental surfaces after an intervention
to improve cleaning methods. BMC Infect Dis 2007;7:61.
33. Alfa MJ, Dueck C, Olson N, et al. UV-visible marker confirms
that environmental persistence of Clostridium difficile
spores in toilets of patients with C. difficile-associated diar-
rhea is associated with lack of compliance with cleaning
protocol. BMC Infect Dis 2008;8:64e70.
34. Getchell-White SI, Donowitz LJ, Groschel DH. The inani-
mate environment of an intensive care unit as a potential
source of nosocomial bacteria: evidence for long survival
of Acinetobacter calcoaceticus. Infect Control Hosp Epi-
demiol 1989;10:402e407.
35. Wendt C, Dietze B, Dietz E, Ruden H. Survival of Acineto-
bacter baumannii on dry surfaces. J Clin Microbiol 1997;
35:1394e1397.
36. Scerpella EG, Wanger AR, Armitige L, Anderlini P,
Ericsson CD. Nosocomial outbreak caused by a multiresistant
clone of Acinetobacter baumannii: results of the caseecon-
trol and molecular epidemiologic investigations. Infect Con-
trol Hosp Epidemiol 1995;16:92e97.
37. Tankovic J, Legrand P, de Gatines G, Chemineau V, Brun-
Buisson C, Duval J. Characterisation of a hospital outbreak
of imipenem-resistant Acinetobacter baumannii by pheno-
typic and genotypic typing methods. J Clin Microbiol
1994;32:2677e2681.
38. Denton M, Wilcox MH, Parnell P, et al. Role of environmen-
tal cleaning in controlling an outbreak of Acinetobacter
baumannii on a neurosurgical intensive care unit. J Hosp In-
fect 2004;56:106e110.
39. Neely A, Maley MP, Warden GD. Computer keyboards as res-
ervoirs for Acinetobacter baumannii in a burn hospital. Clin
Infect Dis 1999;29:1358e1359.
40. Seifert H, Boullion B, Schulze A, Pulverer G. Plasmid DNA
profiles of Acinetobacter baumannii: clinical application
Environmental cleaning and HAI
in a complex endemic setting. Infect Control Hosp Epi-
demiol 1994;15:520e528.
41. Wu HM, Fornek M, Schwab KJ, et al. A norovirus outbreak at
a long-term-care facility: the role of environmental surface
contamination. Infect Control Hosp Epidemiol 2005;26:
802e810.
42. Gallimore CI, Taylor C, Gennery AR, et al. Environmental
monitoring for gastroenteric viruses in a pediatric primary
immunodeficiency unit. J Clin Microbiol 2006;44:395e399.
43. Evans MR, Meldrum R, Lane W, et al. An outbreak of viral
gastroenteritis following environmental contamination at
a concert hall. Epidemiol Infect 2002;129:355e360.
44. Barker J, Vipond IB, Bloomfield SF. Effects of cleaning and dis-
infection in reducing the spread of norovirus contamination
via environmental surfaces. J Hosp Infect 2004;58:42e49.
45. Green J, Wright PA, Gallimore CI, Mitchell O, Morgan-
Capner P, Brown DWG. The role of environmental contami-
nation with small round structured viruses in a hospital out-
break investigated by reverse-transcriptase polymerase
chain reaction assay. J Hosp Infect 1998;39:39e45.
46. Love SS, Jiang X, Barrett E, Farkas T, Kelly S. A large hotel
outbreak of Norwalk-like virus gastroenteritis among three
large groups of guests and hotel employees in Virginia. Epi-
demiol Infect 2002;129:127e132.
47. CDC. Norovirus outbreak in an elementary school e District
of Columbia, February 2007. Morb Mortal Wkly Rep 2008;
56:1340e1343.
48. Griffith CJ. Hospital-acquired infection: are there lessons
from the food industry? Biomed Sci 2006 August:697e699.
49. Al-Hamad A, Maxwell S. How clean is clean? Proposed
methods for hospital cleaning assessment. J Hosp Infect
2008;70:328e334.
385
50. Lewis T, Griffith C, Gallo M, Weinbren M. A modified ATP
benchmark for evaluating the cleaning of some hospital
environmental surfaces. J Hosp Infect 2008;69:156e163.
51. Oelberg DG, Joyner SE, Jiang X, Laborde D, Islam MP,
Pickering LK. Detection of pathogen transmission in neona-
tal nurseries using DNA markers as surrogate indicators.
Pediatrics 2000;105:311e315.
52. Rheinbaben F, Schunemann S, Gross T, Wolff H. Transmis-
sion of viruses via contact in a household setting: experi-
ments using bacteriophage straight phiX174 as a mode
virus. J Hosp Infect 2000;46:61e66.
53. Anonymous. New model cleaning contract. London: NHS
Estates; December 2004.
54. Carling PC, Briggs JL, Perkins J, Highlander D. Improved
cleaning of patient rooms using a new targeting method.
Clin Infect Dis 2006;42:385e388.
55. Dancer SJ, Coyne M, Speekenbrink A, Samavedam S,
Kennedy J, Wallace PGM. Methicillin-resistant Staphylococ-
cus aureus (MRSA) acquisition in an intensive care unit
(ICU). Am J Infect Control 2006;34:10e17.
56. Hugonnet S, Chevrolet J-C, Pittet D. The effect of workload
on infection risk in critically ill patients. Crit Care Med
2007;35:76e81.
57. Scanvic A, Denic L, Gaillon S, Giry P, Andremont A,
Lucet JC. Duration of colonisation by methicillin-resistant
Staphylococcus aureus after hospital discharge and risk fac-
tors for prolonged carriage. Clin Infect Dis 2001;32:
1393e1398.
58. Budd R. Penicillin: triumph and tragedy. Oxford: Oxford
University Press; 2007.
59. Dancer SJ. Back to cleanliness. Rapid response. Br Med J
2008;336.
Journal of Hospital Infection (2009) 73, 386e391
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Controversies in infection: infection control or

antibiotic stewardship to control
healthcare-acquired infection?
I.M. Gould*

Department of Medical Microbiology, Aberdeen Royal Infirmary, Foresterhill, Aberdeen AB25 2ZN, UK

Available online 10 July 2009

KEYWORDS Summary Despite record resource being devoted to the control of health-
Antibiotic stewardship; care-acquired infection (HCAI), rates have never been higher. Although the
Healthcare-acquired discovery of the contagiousness of puerperal sepsis by Alexander Gordon
infection;
heralded the golden era of bacteriology and antibiotics, this led to a belief
Infection control
that infection was beaten. This in its turn may well have led us into a false
sense of security and an over-reliance on antibiotics. Modern medicine has
built many of its advances on a need for antibiotics, but their very success
has led to huge over-use and resulting problems of resistance. Compounded
by the absence of a good antibiotic pipeline we are now being forced to ad-
dress the paradox of antibiotics; namely that they may actually be causing
many HCAIs. Not only Clostridium difficile infection, but many others such
as those caused by meticillin-resistant Staphylococcus aureus, are more or
less completely contingent on antibiotic prescribing. Control of prescribing
would probably be just as effective a measure in our fight against HCAI as
conventional infection control measures. Arguably, traditional infection
control is akin to fire-fighting and antibiotic stewardship to prevention.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

Introduction infection (HCAI) seemingly uncontrollable, it is

pertinent to assess whether current strategies to
In a time of unparalleled resource devoted to contain HCAI are failing and, if so, why?
infection control, and healthcare-acquired Certainly, the public perception is of hospitals
as the dangerous places of Florence Nightingale’s
time.1 Meticillin-resistant Staphylococcus aureus
* Tel.: þ44 1224 554954; fax: þ44 1224 550632. (MRSA) rates continue to rise in most countries,
E-mail address: i.m.gould@abdn.ac.uk as do most of the multi-resistant Gram-negative
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.02.023
Control of healthcare-acquired infection
bacteria (see the European Antimicrobial Resis-
tance Surveillance System website) and in many
countries Clostridium difficile seems to be an in-
creasing problem, not withstanding its ascertain-
ment difficulties.2 With MRSA and C. difficile, at
least, it is quite clear that those infections
are an additional burden, with rates of meticillin-
susceptible S. aureus (MSSA) infection remaining
stable or even increasing.3
What is also notable is that the majority of day-
to-day infection control problems in the hospital
are due to increasingly resistant bacteria. In this
bicentenary year of Charles Darwin’s birth it is
pertinent to reflect on the role that antibiotic use
may have as a selecting force for evolution of
these ‘superbugs’.
In this short review I will reflect on our over-
reliance on antibiotics in modern medicine and try
to put antibiotic stewardship in historical perspec-
tive alongside infection control, at least as an
equal partner in our fight against HCAI.
Are infection control policies failing?
Ignaz Semmelweis is often cited as the father of
modern infection control, although such concepts
are actually more than 2000 years older!4
Marcus Terentius Varro (116e27 BC), a Roman
soldier and Director of the Imperial Library in
Rome, warned against the building of homes near
swamps ‘because there are bred certain minute
creatures which cannot be seen by the eyes, which
float in the air and enter the body through the
mouth and nose and there cause serious diseases’.
Girolamo Fracastoro (1478e1553) was perhaps
the first to formulate a recognisable ‘modern
theory’ of contagion. He stated that it could occur
in three ways; by direct contact with the infected,
by contact with their clothes, or through the air.
He described the agents of contagion as germs or
seeds, small imperceptible particles, each specific
for a different disease.
It took the invention of the microscope for
theories to advance. Robert Hooke (1635e1703),
Athanasius Kircher (1602e1680) and Antonie
van Leeuwenhoek (1632e1723) described micro-
organisms in numerous settings using microscopes.
Hooke even described ‘little worms’ in the blood
of victims of bubonic plague in 1656! By 1764, Sir
John Pringle (1707e1782) had established the
basic principles of sanitation and ventilation of
hospital wards and military quarters. He uses the
word ‘antiseptic’ in the context of Kircher’s
theory of contagion by microscopic germs
(‘animalcula’).
387
But it is to the obstetrician Alexander Gordon of
Aberdeen (1752e1798) that we must look for the
next big breakthrough.5 Based on detailed diaries
he kept of an outbreak of puerperal fever which
ran its course from 1789e1793 in Aberdeen, Gordon
deduced the relationship between erysipelas and
puerperal fever, its cause as being transferred
from the postmortem room to the lying-in room
by midwives and its prevention by adequate
hand hygiene and disinfection of clothes. His
treatise, first published in 1795, predates Sem-
melweis by more than 50 years but despite being
reprinted several times over the next century, in
London and the USA, he has always been over-
shadowed by Semmelweis.6 Interestingly, like
Semmelweis, Gordon made himself unpopular
among his colleagues with his observations. Soon
after, he left Aberdeen and died of tuberculosis.
Gordon was cited by Oliver Wendell Holmes but
not by Semmelweis, so it is interesting to specu-
late on whether Semmelweis was actually aware
of Gordon’s work.7 Possibly he was and probably
he should have been, as the treatise had been
published so many times. Also, Scottish medicine
was well renowned during that period and com-
munication with and travel to Europe frequent.
On the other hand, Semmelweis was said not to
have a good grasp of the medical literature, un-
like Gordon.8
Further advance awaited Pasteur’s refutation of
spontaneous generation of life, Koch’s recognition
of the first pathogenic organism (anthrax), Lister’s
implementation of Pasteur’s findings in the form of
antiseptic surgery and its development into aseptic
surgery.9 In the meantime, Florence Nightingale
had introduced better sanitation and hand washing
at the Barrack hospital in Scutari, although her
role in this has recently been debated.1
It is well documented that antiseptic and sub-
sequently aseptic surgery and other procedures
were rapidly taken up by virtually all informed
medical practitioners over the following years and
popular perception is that they were well adhered
to, with the traditional matron overseeing good
practice.9 But does this still generally apply or has
poor practice crept in, albeit perhaps just outside
the operating theatre? Certainly this seems to be
the case where hand hygiene adherence rates of
20e40% are considered normal.10 If this be so,
may it be due directly or indirectly to the great
success of, and perhaps over-reliance on, anti-
biotics e with a belief that antibiotics can negate
the need for good hygiene, asepsis and cleanliness?
Furthermore, could antibiotic use actually be
harmful in some instances, actually increasing
the number of infections? Increasingly we hear
388
calls for zero tolerance in failure to adhere to hand
hygiene and other infection control policies. But
perhaps some of the policies of recent years have
been misplaced. Take universal precautions
(UPs), for example. Such a blind faith in their
broad applicability has, I believe, been one of
the main reasons for the failure to control
MRSA.11 Over-reliance on UPs to the exclusion of
specific policies targeting the biology of problem
organisms such as MRSA is counterproductive,
even if UPs could be successfully implemented. A
problem organism such as MRSA should be targeted
specifically according to its method of spread and
biological weaknesses.12 In the case of MRSA this
means active surveillance cultures, admission cul-
tures, isolation, decolonisation and decontamina-
tion.13 This has been clear for at least three
decades but with a few exceptions, for example
in Scandinavia and Holland, such policies have
been very poorly implemented.
The antibiotic era
It is hard to imagine hospital medicine in the pre-
antibiotic era, but contemporary reports of the
difference that antibiotics made to individual cases
well describe the fantastic advance that they
were.14 Indeed it is the case that many of the great
advances of modern medicine and surgery would
not have been possible without the ability to rely
on antibiotics to cure secondary infection. But at
what cost? The average hospital now uses, on aver-
age, levels approaching 100 defined daily doses per
100 occupied bed-days, a value equivalent to eco-
logical saturation.15 This is the equivalent of all pa-
tients receiving a full daily dose of antibiotic from
the day of admission until discharge. Of course,
on average, only less than half of inpatients will ac-
tually receive antibiotics at any one time, but those
who do receive them are often on double doses or
combination therapy. Many guidelines advocate
prolonged antibiotic prophylaxis, particularly for
the ever-increasing number of immunosuppressed
patients, and often these uses are justified by a rea-
sonable evidence base.16 On many other occasions,
however, prescribing is definitely inappropriate e
‘just in case’ or on the basis of a poor quality sever-
ity assessment or misdiagnosis. Current policies to
shorten length of stay and curtail costs also
encourage empiric use, often of unnecessarily
broad-spectrum antibiotics. Combination therapy
is often used for a number of reasons, good
and bad, including broadening spectrum to accom-
modate increasing antibiotic resistance, thus
completing the spiralling circle of therapeutic
empiricism.17
I.M. Gould
Much antibiotic over-use may seem justifiable
for the individual patient when viewed in isolation;
nevertheless antibiotic use, unlike any other drug
category, cannot be viewed in such isolation.
Antibiotic use as a cause of HCAI
Each antibiotic prescription has an environmental,
ecological consequence.18 Only over the past
decade has this become widely accepted, although
Fleming warned of this at an early stage. But the
downside of antibiotics goes well beyond develop-
ment of resistance as already mentioned with
MRSA and C. difficile. Indeed, not only does anti-
biotic use select for and maintain antibiotic resis-
tance, but it also enhances its spread.19,20 This
latter point is crucial to the control of modern
HCAI and illustrates how potentially critical anti-
biotic stewardship is in the control of most HCAIs.
Possibly it also illustrates why traditional infec-
tion control policies have not been as successful
as we would have hoped.
Increased transmission of tetracycline-resistant
organisms during tetracycline therapy was demon-
strated as long ago as 1960.21 In 2008 increased
numbers of MRSA were demonstrated in the noses
of carriers receiving quinolones or b-lactams (to
which the MRSA strains were resistant) compared
with controls.22 Presumably this is due to the com-
petitive advantage that was achieved by the anti-
biotic administration, ablating the normal
protective commensal flora (including MSSA), al-
lowing multiplication of the MRSA with increased
potential for contaminating the environment. It
is easy to imagine adjacent patients being more
susceptible to acquisition of MRSA if they are also
on an appropriate ‘MRSA selecting’ antibiotic. Fur-
thermore, antibiotics such as the quinolones and
cephalosporins are well known to increase expres-
sion of fibronectin adhesins, facilitating adherence
and ability to colonise, and also many other viru-
lence factors such as toxins which might cause col-
onisation to develop into infection.19,20 With the
increasing trend to admission screening for MRSA,
such knowledge should be put to good use by
avoiding prescription of agents to which the
MRSA is resistant, both in MRSA carriers and their
contacts. This is not always an easy task when
dealing with multiply resistant MRSA.
In 1970 Sleigh et al. described inability to
control an outbreak of multidrug-resistant (MDR)
Klebsiella pneumoniae in a neurosurgery unit.23
All conventional control measures had failed, so
the authors took the unusual, and probably never
to be repeated, step of banning prescription of
all antibiotics. Not only did the outbreak strain
Control of healthcare-acquired infection
disappear (actually it may have been declining
anyway) but, more interestingly, the total number
of infections decreased dramatically and no one
died of uncontrolled infection during the period
of embargo on antibiotic prescriptions.
So we have to ask ourselves, could antibiotic use
be increasing the incidence of most HCAIs due to
MDR bacteria, and not just MRSA and C. difficile?
Perhaps it is not just resistance that is caused by
antibiotics but also HCAI? This is a frightening
thought on one hand, as we have become so reliant
on antibiotics. On the other hand, this reliance has
probably become over-reliance, leading to poor
quality infection control in the belief that infec-
tion has been beaten by antibiotics. But it is also
a good time to take stock, as the antibiotic pipe-
line is dry and there are no significant develop-
ments expected for another 10e20 years.24 So
the sooner we are more cautious in our use of
antibiotics the better, needing to preserve those
that we have.
Returning to MRSA as an example, there are
many caseecontrol studies demonstrating prior
antibiotic exposure, particularly with cephalo-
sporins and quinolones as a risk factor for both
MRSA colonisation and infection.25 The same
applies, of course, to many other MDR bacteria
including most of the MDR Gram-negatives. For
MRSA there are also numerous ecological studies
demonstrating such a relationship at a community
and hospital level.14 There are even a dozen or so
studies in the literature suggesting that MRSA rates
were decreased by antibiotic policies that reduced
use of cephalosporins and quinolones.26 Usually,
however, such studies demonstrate the concept
of ‘squeezing the balloon’ whereby reduction in
use of one drug is mirrored by increasing use of an-
other. Usually this will have its own resistance
problems, although it may be that the cephalo-
sporins and quinolones, in particular, are more
prone to resistance developing than the penicil-
lins, for example.
Antibiotic stewardship
Unfortunately, reducing total antibiotic use is
even more difficult than modulating class use
and there are very few robust studies in the
literature that demonstrate a reversal of resis-
tance.27 There are complex molecular reasons
why this might be the case, particularly the
mobile gene cassette that can integrate on
chromosomes to encode multiple resistance
determinents. Use of any one agent thus encoded
can provide selection advantage for maintaining
389
all such encoded resistance determinants. Also,
any loss of fitness can often be compensated for
by mutations in the bacteria. So it may be that
proper stewardship can only halt the current de-
velopment of resistance, although there are
enough examples of reversals to give us some
hope.28
One of the main problems is how to reduce
overall antibiotic use, as modern medical devel-
opments seem to know no end to the immunosup-
pression of the patient. Currently, our efforts on
stewardship (which mainly revolve around formu-
laries and guidelines) achieve only uniformity of
prescribing with adherence to policies, guidelines
and formularies. Paradoxically this may actually
be more harmful, as the best defence against
resistance is probably diversity of prescribing.29
Little effort is actually put into reducing prescrib-
ing with the possible exception of shortening dose
duration. What is really required is better diagno-
sis (of course including better diagnostics), better
severity assessment and, where antibiotics are in-
dicated, better knowledge of pharmacokinetic/
pharmacodynamic dosing schedules to optimise
outcome and delay development of resistance.
These are difficult things to address in the middle
of the night when relatively junior doctors are
making the prescribing decision, often trying
to protect themselves from the threat of under-
treating (and perhaps of litigation) and trying to
do their best for the patient. More is better or let’s
prescribe ‘just in case’!
The example of recent guidelines from the
British and American Thoracic Societies for treat-
ment of community-acquired pneumonia (CAP) are
a good case in point.30,31 They make recommenda-
tions for various severities of illness but the natu-
ral inclination of most clinicians is always to
‘cover’ themselves or their patients with the rec-
ommendation for treatment of severe cases. Argu-
ably, such resulting over-use of cephalosporins,
quinolones and macrolides may have triggered
the MRSA outbreaks in North America and the
UK.32 The guidelines themselves are not without
fault. While they purport to be evidence-based,
the antibiotic recommendations are definitely
not, leaning heavily to combination therapy and
the perceived need to cover all aetiological
agents. The Scandinavian countries, which by and
large have retained penicillin monotherapy as
their treatment of choice for CAP, have not
published worse outcomes and have many fewer
problems with antibiotic resistance.
Gone are the days when we can aspire to ‘cover’
100% in the empiric treatment of every severe
infection.33 Paradoxically, this is at a time when it
390
has never been clearer that immediate appropri-
ate empiric therapy is beneficial to the septic
patient. Difficult decisions have to be made. For
instance, do we prescribe a carbapenem and
glycopeptide for all septic patients if we have an
ESBL and MRSA problem? At what ‘level’ is MRSA
or ESBL a problem that should impact on empiric
choice? What happens when carbapenem resis-
tance or glycopeptide resistance become problem-
atic, as they already have in some parts of the
world? Colistin, tigecycline and fosfomycin are
possible alternatives but not so universally appli-
cable as carbapenems. Outcome with them or
with glycopeptides, even in susceptible organisms,
may be suboptimal, making decisions on empiric
therapy even more difficult. While re-evaluation
and possible step-down are often an option within
the next 48 h, this will be too late for some
patients and in the majority there will not be any
positive cultures to guide follow-on therapy.
Much more strategic thought needs to be given to
these important issues in the absence of new
antibiotics to help us out of this dilemma.
In conclusion, new thinking is needed in our use
of antibiotics, not only to delay the development
of further antibiotic resistance but to help in the
control of HCAI. This will require widespread
consultation and the time just might be now.
Conflict of interest statement
None declared.
Funding sources
None.
References
1. Williams K. Reappraising Florence Nightingale. Br Med J
2008;337:1461e1466.
2. Planche T, Aghaizu A, Holliman R, et al. Diagnosis of Clos-
tridium difficile infection by toxin detection kits: a system-
atic review. Lancet Infect Dis 2008;8:777e784.
3. Gould IM. Costs of hospital-acquired methicillin-resistant
Staphylococcus aureus (MRSA) and its control. Int J Antimi-
crob Agents 2006;28:379e384.
4. Dallas J. Little worms which propagate plague. J R Coll
Physicians Edinb 2008;38:376e377.
5. Porter IA. Alexander Gordon, M.D. of Aberdeen 1752e1799.
Edinburgh: University of Aberdeen/Oliver & Boyd; 1959.
6. Gordon A. Treatise on the epidemic puerperal fever of
Aberdeen. London; 1795.
7. Holmes OW. The contagiousness of puerperal fever. N Engl
Q J Med Surg 1843;i:503.
8. Loudon I. The tragedy of childbed fever. London: Oxford
University Press; 2000.
9. Bulloch W. The history of bacteriology. London: Oxford Uni-
versity Press; 1960.
10. Cepeda JA, Whitehouse T, Cooper B, et al. Isolation of pa-
tients in single rooms or cohorts to reduce spread of MRSA
I.M. Gould
in intensive-care units: prospective two-centre study. Lan-
cet 2005;365:295e304.
11. Gould IM. Can we control MRSA? Scott Med J 2007;52:3e4.
12. Lindsay JA, Holden MT. Understanding the rise of the super-
bug: investigation of the evolution and genomic variation of
Staphylococcus aureus. Funct Integr Genomics 2006;6:
186e201.
13. Gould IM, MacKenzie FM, MacLennan G, Pacitti D, Watson EJ,
Noble DW. Topical antimicrobials in combination with admis-
sion screening and barrier precautions to control endemic
methicillin-resistant Staphylococcus aureus in an intensive
care unit. Int J Antimicrob Agents 2007;29:536e543.
14. Jeffrey JS, Thomson S. Penicillin in battle casualties.
Br Med J 1944;2(4356):1e4.
15. MacKenzie FM, Bruce J, Struelens MJ, Goossens H,
Mollison J, Gould IM, ARPAC Steering Group. Antimicrobial
drug use and infection control practices associated with
the prevalence of methicillin-resistant Staphylococcus au-
reus in European hospitals. Clin Microbiol Infect 2007;13:
269e276.
16. Gafter-Gvili A, Paul M, Fraser A, Leibovici L. Effect of quino-
lone prophylaxis in afebrile neutropenic patients on micro-
bial resistance: systemic review and meta-analysis.
J Antimicrob Chemother 2007;59:5e22.
17. Kim JH, Gallis HA. Observations on spiraling empiricism: its
causes, allure, and perils, with particular reference to anti-
biotic therapy. Am J Med 1989;87:201e206.
18. Sarkar P, Gould IM. Antimicrobial agents are societal drugs:
how should this influence prescribing? Drugs 2006;66:
893e901.
19. Gould IM. Antibiotic policies to control hospital-acquired in-
fection. J Antimicrob Chemother 2008;61:763e765.
20. Dancer SJ. The effect of antibiotics on methicillin-resistant
Staphylococcus aureus. J Antimicrob Chemother 2008;61:
246e253.
21. Berntsen C, McDermott W. Increased transmissibility of
staphylococci to patients receiving an antimicrobial drug.
N Engl J Med 1960;262:637e642.
22. Cheng VCC, Li IWS, Wu AKL, et al. Effect of antibiotics
on the bacterial load of meticillin-resistant Staphylococcus
aureus colonization in anterior nares. J Hosp Infect 2008;
70:27e34.
23. Price DJE, Sleigh JD. Control of infection due to Klebsiella
aerogenes in a neurosurgical unit by withdrawal of all anti-
biotics. Lancet 1970;2:1213e1215.
24. Gould IM. Antimicrobials: an endangered species? Int J Anti-
microb Agents 2007;30:383e384.
25. Tacconelli E. Does antibiotic exposure increase the risk of
methicillin-resistant Staphylococcus aureus (MRSA) isola-
tion? A systematic review and meta-analysis. J Antimicrob
Chemother 2008;61:26e38.
26. Gould IM. Comment on: Interventions to control MRSA: high
time for time-series analysis? J Antimicrob Chemother
2009;63:224.
27. Davey P, Brown E, Fenelon L, et al. Systematic review of
antimicrobial drug prescribing in hospitals. Emerg Infect
Dis 2006;12:211e216.
28. Gould IM. The epidemiology of antibiotic resistance. Int J
Antimicrob Agents 2008;32:S2eS9.
29. Sandiumenge A, Diaz E, Rodriguez A, et al. Impact of di-
versity of antibiotic use on the development of antimicro-
bial resistance. J Antimicrob Chemother 2006;57:
1197e1204.
30. Anonymous. Guidelines for the management of adults with
hospital-acquired, ventilator-associated, and healthcare-
associated pneumonia. Am J Respir Crit Care Med 2005;
171:388e416.
Control of healthcare-acquired infection 391

31. British Thoracic Society Standards of Care Committee. BTS
guidelines for the management of community acquired
pneumonia in adults. Thorax 2001;56:iv1e64.
32. Monnet DL, MacKenzie FM, López-Lozano JM, et al. Antimi-
crobial drug use and methicillin-resistant Staphylococcus
aureus, Aberdeen, 1996e2000. Emerg Infect Dis 2004;10:
1432e1441.
33. Masterton RG. The new treatment paradigm and the role
of carbapenems. Int J Antimicrob Agents 2009;33:
105e110.
Journal of Hospital Infection (2009) 73, 392e396
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Where does infection control fit into a hospital

management structure?
E.T. Brannigan a, E. Murray a,b, A. Holmes a,b,*
a
Imperial College Healthcare NHS Trust, London, UK
b
Imperial College, London, UK

Available online 20 August 2009

KEYWORDS Summary To be effective, infection prevention and control must be

Hospital-acquired integrated into the complex and multiple interlinking systems within a hos-
infection; pital’s management structure. Each of the systems must consider how
Organisational change;
activity associated with it can be optimised to minimise infection risk to
Patient safety;
Quality
patients. The components of an organisational structure to achieve these
quality assurance and patient safety aims are discussed. The use of perfor-
mance management tools in relation to infection control metrics is re-
viewed, and the use of hospital-acquired infection as a proxy indicator
for deficiencies of system management is considered. Infection prevention
and control cannot be the role and responsibility of a single individual or
a small dedicated team; rather it should be a priority at all levels and
integrated within all management systems, including the research and
educational agendas.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

Introduction
Where does infection control fit into a hospital
management structure? The question should first
be asked what a hospital management structure
should provide in relation to infection prevention
* Corresponding author. Address: Imperial College London,
Hammersmith Hospital Campus, Du Cane Road, London W12
0NN, UK.
E-mail address: alison.holmes@imperial.ac.uk
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.03.031
and control. Any such structure should set its sights
beyond merely addressing any legislative and stat-
utory requirements, and should provide a frame-
work for organisational change as well as the
systems required to support this change. Key ele-
ments of this framework include a belief in a shared
purpose and aspiration to excellence, supporting
reinforcement systems, assurance that staff have
the skills required for change and that staff have
consistent and multiple role models within the
organisation. These elements could almost be
Hospital management and infection
considered together as an ‘organisational change
bundle’. The structure should also have the ability
to foster reliability and resilience.1,2
Hospital management must consider infection
control primarily as a core aspect of patient
safety, and as an indicator of quality of care. It
must be acknowledged that achieving quality
patient care and safety is entirely dependent on
implementation of best practice in individual
clinical care, and that this delivery requires
expert input at both operational and strategic
levels.
Role and interplay of multiple
management systems
Infection prevention and control relies upon the
successful interplay and management of multiple
management systems, integrating infection pre-
vention and control into management at all levels
and forcing specific consideration of infection risk
for patients throughout multiple management
systems. To achieve optimally lowered infection
risk for patients, action is required in a wide
variety of domains. Examples of the diverse areas
and aspects of a hospital infrastructure which must
be considered as having a role in infection pre-
vention and control include human resources (HR),
staff:patient ratios, bed management, patient
pathways, staff training, information and informa-
tion technology, contract management, procure-
ment, estates and facilities, capital planning,
building, design, performance monitoring, anti-
biotic stewardship, organisational learning, adop-
tion of innovation, risk management, governance,
priority setting, resource allocation, communica-
tions, and business planning among others. Given
the complexity of assuring high quality standards
across such diverse groups and the contingency of
each of these systems one on the other, achieving
effective infection prevention cannot therefore
be devolved to an individual specialist team or
to an isolated group of committed experts.
Instead a comprehensive whole-scale organisa-
tional approach is required. Infection risk needs
to be considered in almost all areas of hospital
management, patient pathways and delivery of
care. Awareness of infection prevention priorities
must be embedded and be considered a prerequi-
site in decision-making at all levels. Infection
control must be fully integrated into the health-
care structure and systems, into performance and
patient outcome metrics and into the hospital
culture including in research and education
strategy.
393
Impact of human resources management
on infection control
Areas in which HR management has a direct impact
on infection control include occupational health,
health clearance and vaccination against prevent-
able infection. However, manpower details such
as working patterns, team formation and skill-mix
are also inextricably linked to infection control
outcomes. Infection risk has been shown to be
strongly related to staffing levels and staff train-
ing. These are underpinned by other HR activities
including recruitment and retention, staff ap-
praisal, job descriptions, induction, leave manage-
ment, agency staff and shift patterns. HR
management has been shown to have particular
impact in the context of critical care areas,
especially intensive care units (ICUs). Staffing
levels among nursing staff especially have been
shown significantly to affect hospital-acquired
infection, both in neonatal and in adult ICUs.3,4
Physician staffing levels have also been studied;
infection-related outcomes are not addressed sep-
arately, but findings relate to the overall adverse
impact on LOS and mortality.5
An organisation-wide approach
A comprehensive organisational approach should
be adopted, one which recognises existing man-
agement structures and facilitates development of
a model that fully integrates infection control into
hospital management and the quality agenda.
Infection control must be a core part of gover-
nance and must maintain a high clinical profile
alongside a high management profile.
Bridging the gap between managers and
clinicians
Infection control provides an opportunity, agenda
and framework for such a bridging strategy. If
infection control sits solely within a separate
service or team, unlinked to management strat-
egy, or with little clinical and managerial influ-
ence, it has limited effectiveness and impact.
Strategic decision-makers without knowledge or
without systems to ensure consideration of in-
fection control put patients at risk. Managers
need to be aware of the direct impact, any
knock-on effects and potential collateral damage
that could occur as a result of any decision-
making, whether it be about capital planning,
394
pathways, service reconfiguration, etc. Avoiding
any compromise in infection prevention and pa-
tient safety depends on continuous awareness
which can be achieved by improved dialogue and
working to a common agenda.6e9 Skills and training
needs of management and clinical staff must be
realised, and there must be clear lines of account-
ability which follow funding structures, and which
are supported by agreed performance indicators.10
These indicators should be recognised and incor-
porated in any balanced scorecards, dashboards
or organisational performance monitoring tools.
Balanced scorecard for infection
prevention and patient safety
The use of the balanced scorecard provides
a framework to measure performance beyond
finances in private industry. It allows alignment
of performance measures with an organisation’s
strategic mission and goals, measuring perfor-
mance, and factors driving performance. In es-
sence it is a tool that can provide a basis for
executing good strategy well and managing
change successfully. There are caveats with such
an approach inasmuch as you get what you
measure, and use of a scorecard is known to
skew activity. The scorecard is not fixed; instead
it needs regular refreshing and updating as prior-
ities of the institution evolve.11 Many trusts use
the balanced scorecard approach for internal per-
formance managing. Indicators used must have
credibility with frontline staff and have shared
agreement of their value. Leaders must select
outcome or process measures that represent the
interests of frontline healthcare workers, must
ensure that it has local face validity, and that per-
formance on the chosen measure will help focus
quality improvement.12
Organisational learning and culture
This requires creation of an environment which
permits continuous learning and quality improve-
ment shared across professional bodies, and across
directorates and units. Systems-based approaches
embedded in the framework of the organisation
must be adopted for sustainability. These systems
must be shaped and supported by clinical and
managerial expertise, within a culture that sup-
ports and reinforces infection control as a corpor-
ate priority, continually reinforcing the required
behaviour and values. Interventions must be sup-
ported by regular feedback, monitoring systems
E.T. Brannigan et al.
and metrics which drive continuous learning and
improvement. A critical component in all of these
is strong corporate accountability and hospital
leadership. ‘Achieving safety requires more
than individual carefulness. It is a corporate
responsibility.’13,14
Leadership
Within a hospital structure leadership must come
directly from the Chief Executive Officer (CEO)
who must have demonstrable commitment to
infection prevention and control and leader-
ship.15 The Director of Infection Prevention and
Control (DIPC) is the person appointed by each
trust who has direct line of accountability to
the CEO and this role provides opportunities for
specialist leadership and composite roles.16 The
DIPC leads and champions infection prevention
and control at multiple levels within the organi-
sation, ensuring that a consistent message is
embedded in directorates, groups, teams, and
networks. Consistent role models are key to
organisational change.1 The DIPC must have
a profile and recognition of this role by peer
groups and the organisation and provide a bridg-
ing role between managers and clinicians. The
multiple layered approach allows the DIPC to
address issues pertinent to patients and to public
concerns first and foremost, to communicate
a clear message to staff which states and pro-
motes the view of infection control as a shared
goal involving all staff, across the multidisciplin-
ary spectrum. This includes delivering a strong
message to all prescribers relating to prudent
prescribing and ensuring that antibiotic steward-
ship is fully addressed by the organisation. The
DIPC role must deliver internal reinforcement,
leadership, board-to-ward engagement, exper-
tise and organisational commitment, as well as
provide role modelling and exemplify corporate
responsibility.
Lessons learned from approaches and
processes from outside healthcare
Like many other areas of hospital management and
patient safety, much can be learned from arenas
outside healthcare. It is also worth noting that the
significant reductions of MRSA bacteraemias made
by many NHS trusts were due not to new technol-
ogy or scientific advances but through better
application of existing knowledge using a variety
of managerial approaches. These include the risk
Hospital management and infection 395

reduction and hazard management approaches HCAI as an indicator of complex systems

from high risk industries. management and organisational
Organisational approaches, as used in other
resilience
complex industries, can be applied to acute trusts
to embed infection control into the running of
There is pressure for greater use of quality in-
hospitals and delivery of clinical care, and organisa-
dicators and measures in healthcare, and infection
tional change management tools can be used to
rates and infection prevention is a major indicator
drive this. These include leadership and the
of quality in clinical care and quality in manage-
development of a ‘board-to-ward’ approach of
ment.21,22 Infection control can serve as a marker
engagement, and performance indicators with
of an organisation’s capacity to manage multiple,
scorecards or dashboards while maintaining aware-
complex systems, and HCAI rates can therefore
ness of their limitations. In addition to trusts
be proxy indicators of levels of staffing, levels
addressing internal reinforcement systems, exter-
of training, organisational stress, management
nal reinforcement provides the scrutiny and
failure, inadequate systems, reliability, and
incentives to achieve goals, and has been seen
resilience.
with public information campaigns, the introduc-
A resilient system is one that quickly recovers
tion of mandatory reporting, the intense media
stability after an unexpected event or in the face
interest and increased patient choice. Strong
of continuous significant stresses in a complex,
external reinforcement of CEO accountability for
dynamic environment; the stability is provided by
healthcare-associated infections (HCAIs) was pro-
constant change rather than continuous repeti-
vided not just by mandatory data submission in
tion.23 The challenge to the system is that it must
England, but also by the mandatory requirement
facilitate reliability while developing resilience to
that CEOs review and sign off data monthly. Finally
both internal and external threats and stressors,
the introduction of legislation made the delivery of
e.g. mergers, leadership change, financial deficit,
infection prevention and control and compliance
infection outbreaks, global recession, major policy
with core duties legal requirements for every NHS
shift, environmental incident and technology
trust.
change among others.
Valuable lessons from the safety arena include
Common to both the Maidstone and Stoke
the use of quality improvement tools and the
Mandeville Health Care Commission reports re-
development and adoption of ‘care bundles’ to
lating to C. difficile were themes related to organ-
ensure best practice and minimising risk in partic-
isational stressors; recent mergers, preoccupation
ular clinical interventions. Local improvement
with financial situation, service reconfigurations
campaigns can learn form advertising and politics,
and high bed occupancy levels (>90%). This was
as have the national patient safety and quality
in addition to the more obviously directly related
improvement campaigns such as those of the
issues such as poor antibiotic stewardship, issues
Institute of Health Improvement’s ‘1000 lives’
relating to isolation facilities and the state of the
campaign, the ‘5 million lives’ campaign, and the
hospital environment.24,25
DoH ‘Saving Lives’ campaign. These national cam-
Resilient organisations hold to key principles:
paigns also provided additional external reinforce-
acknowledging even small failings, resisting over-
ment to local programmes.
simplified accounts of events, remaining sensitive
The effective use of teams has been recognised
and responsive to front-line operations, maintain-
as a key aspect of delivering quality performance
ing reserves and capability in anticipation of
in other industries. This involves better use of
stressors, and are adaptable to shifting locations
team structures with an acknowledgement of in-
of expertise.23 Each of these principles should be
dividual skills, but with staff understanding that
addressed by a trust in its overarching approach
teamwork is valued rather than individual inter-
to infection prevention and control.
ventions. This relies on leadership, adaptability, as
well as mutual performance monitoring and sup-
port.17,18 The team approach has been particularly
valuable in reducing line-associated bacteraemia Discussion
in ICUs. Staff can be further supported by encour-
agement of use of decision aids, systems which Infection prevention and control should have
make the desired action the default, and clear de- a presence and profile throughout the hospital
lineation of processes and pathways, all of which management structure. From the top and down
seek to decrease the human factor and increase through every level of decision-making and care,
the system reliability.19,20 supported by clear leadership, systems, structures
396
and expert input, an organisational approach to
infection prevention is needed, involving manage-
ment structures and systems to change behaviour
and culture, to drive quality improvement and
support sustainable best practice in a resilient
framework. In this context training needs must be
addressed, and further applied research is needed
to determine which healthcare models best deliver
effective, reliable and resilient infection
prevention.26,27
Acknowledgements
We are grateful to the support from the National
Institute of Health Research and the Biomedical
Research Centre at Imperial College.
Conflict of interest statement
None declared.
Funding sources
Support from the NIHR.
References
1. Lawson E, Price C. Psychology of change management.
McKinsey Q 2003;2:30e41.
2. Holmes A. Working smarter: an organisational approach to
infection prevention. Bull R Coll Pathol 2008;142:106e109.
3. Harbarth S, Sudre P, Dharan S, Cadenas M, Pittet D. Out-
break of Enterobacter cloacae related to understaffing,
overcrowding, and poor hygiene practices. Infect Control
Hosp Epidemiol 1999;20:598e603.
4. Hugonnet S, Chevrolet JC, Pittet D. The effects of workload
on infection risk in critically ill patients. Crit Care Med
2007;35:296e298.
5. Pronovost PJ, Angus DC, Dorman T, Robinson KA,
Dremsizov TT, Young TL. Physician staffing patterns and
clinical outcomes in critically ill patients: a systematic re-
view. J Am Med Assoc 2002;288:2151e2162.
6. Silversin J, Kornacki M. Leading physicians through change:
how to achieve and sustain results. Tampa, FL: American
College of Physician Executives; 2000.
7. Shortell SM, Waters TM, Clarke KWB, Budetti PP. Physicians
as double agents: maintaining trust in an era of multiple
accountabilities. J Am Med Assoc 1998;280:1102e1108.
E.T. Brannigan et al.
8. Edwards N. Doctors and managers: building a new relation-
ship. Clin Med 2005;5:577e579.
9. Edwards N, Kornacki MJ, Silversin J. Unhappy doctors: what
are the causes and what can be done? Br Med J 2002;324:
835e838.
10. Holmes A, Dinneen M, Murray E. Creating a new culture.
Publ Serv Rev: Health 2006;8:19e22.
11. Kaplan RS, Norton DP. Using the balanced scorecard as
a strategic management system. Harv Bus Rev 1996;1:
75e85.
12. Pronovost PJ, Berenholtz SM, Needham DM. A framework for
health care organizations to develop and evaluate a safety
scorecard. J Am Med Assoc 2007;298:2063e2065.
13. Leape L, Epstein AM, Hamel MB. A series on patient safety.
N Engl J Med 2002;347:1272e1274.
14. Annas GJ. The patient’s right to safety e improving the
quality of care through litigation against hospitals. N Engl
J Med 2006;354:2063e2066.
15. Institute of Medicine. In: Kohn LT, Corrigan J, Donaldson MS,
editors. To err is human: building a safer health system.
Washington, DC: National Academy Press; 1999.
16. Department of Health. Winning ways: working together to
reduce healthcare associated infection in England. London:
DoH; 2003.
17. Baker DP, Day R, Salas E. Teamwork as an essential compo-
nent of high-reliability organizations. Health Serv Res 2006;
41:1576e1598.
18. Wenzel RP, Edmond MB. Team-based prevention of cath-
eter-related infections. N Engl J Med 2006;355:2781e2783.
19. Resar R. Making noncatastrophic health care processes re-
liable: learning to walk before running in creating
high-reliability organizations. Health Serv Res 2006;41:
1677e1689.
20. Pronovost PJ. Creating high reliability in health care organi-
zations. Health Serv Res 2006;41:1599e1617.
21. Darzi A. High quality care for all. NHS next stage review
final report. London: Department of Health; 2008.
22. Vincent C, Aylin P, Dean Franklin B, et al. Is health care
getting safer? Br Med J 2008;337:1205e1207.
23. Weick K, Sutcliffe K. Managing the unexpected: assuring
high performance in an age of complexity. San Francisco:
Jossey-Bass; 2001.
24. Healthcare Commission. Investigation into outbreaks of
Clostridium difficile at Maidstone and Tunbridge Wells
NHS trust. London: HC; 2007.
25. Healthcare Commission. Investigation into outbreaks of
Clostridium difficile at Stoke Mandeville hospital, Bucking-
hamshire hospitals NHS trust. London: HC; 2006.
26. McDonnell A, Wilson R, Goodacre S. Evaluating and imple-
menting new services. Br Med J 2006;332:109e112.
27. UK Clinical Research Collaboration. Translational infection
research initiative. London: Medical Research Council;
2008.
Journal of Hospital Infection (2009) 73, 397e399
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Hand hygiene and infection in hospitals: what do

the public know; what should the public know?
M. Fletcher*

National Patient Safety Agency, London, UK

Available online 26 September 2009

KEYWORDS Summary Healthcare-associated infection (HCAI) is a topic of increasing

Health care associated public interest, particularly to users of health services. At the same time,
infections; there is a move towards greater openness and transparency across the
Patients;
whole healthcare sector. Thus we see public reporting of performance in
Perception;
relation to rates of HCAI and infection control practices is increasingly well
Public;
Awareness; established in the NHS in England. So does any of this make a difference?
Information And how embedded is the ‘public right to know’? In this paper it is argued
that, although the public right to know about rates of HCAIs is well recog-
nised, the evidence base about the impact of such information is limited.
The paper suggests actions which can be taken by boards and senior
leaders in healthcare organisations to increase impact. Furthermore the
example of hand hygiene suggests that we have some way to go in creating
an environment in which patients feel empowered to ask questions that
may reduce their own vulnerability to infection.
ª 2009 Published by Elsevier Ltd on behalf of The Hospital Infection Society.

Introduction hospitalised.2 At the same time, there is a move

Healthcare-associated infection (HCAI) is a topic of
increasing public interest, particularly to users of
the National Health Service (NHS) in England.1 This
is accompanied by genuine public concern about
the risk of becoming infected as a result of being
* Address: National Patient Safety Agency, 4-8 Maple Street,
London WIT 5HD, UK. Tel.: þ44 20 7927 9527.
E-mail address: martin.fletcher@npsa.nhs.uk
0195-6701/$ - see front matter ª 2009 Published by Elsevier Ltd on behalf of The Hospital Infection Society.
doi:10.1016/j.jhin.2009.06.029
towards greater openness and transparency across
the whole healthcare sector.3
Arguably, the ‘public right to know’ is already
an accepted fact by the NHS in England. Public
reporting of performance in relation to rates of
HCAI and infection control practices is increasingly
well established. There is a large e and growing e
amount of information about HCAI in the public
domain. For example, many NHS trusts’ websites
routinely display data on their infection control
programme and performance. Such information is
398
also increasingly common at the entrance to
hospitals and clinical areas. More widely, the
Health Protection Agency routinely publishes
data on national infection rates.4
So does any of this make a difference? And how
embedded is the ‘public right to know’?
In this paper it is argued that, although the
public right to know about rates of HCAIs is well
recognised, the evidence base about the impact of
such information is limited. The paper suggests
actions which can be taken by boards and senior
leaders in healthcare organisations to increase
impact. Furthermore the example of hand hygiene
suggests that we have some way to go in creating
an environment in which patients feel empowered
to ask questions that may reduce their own
vulnerability to infection.
The evidence on public reporting
Broader than the field of infection control, the
research literature suggests a number of drivers
for publishing patient care performance data,
including5:
e as a means of improving the quality of care
through greater accountability and
transparency;
e as a means of informing selection by patients or
their intermediaries in the context of select-
ing, rewarding or incentivising better perform-
ing providers;
e to help providers identify areas in which they
are underperforming.
There is little evidence about the effect of
publication both in the context of HCAI and wider
health service performance.6 The growth in public
reporting has not been matched by a growth of re-
search on impact. There are mixed signals on the
effect of public reporting on patient outcomes.
Publication of performance data can be an impor-
tant stimulus for quality improvement. This sug-
gests that the organisation which is publishing
data may itself be one of the key audiences. There
is emerging evidence on consumer selection of
health plans in an American context with some
suggestion of unintended consequences such as
adverse selection. As Fung et al. conclude7:
Evidence is scant particularly about individual pro-
viders and practices. Rigorous evaluation of many
major public reporting systems is lacking. Evidence
suggests that publicly releasing performance data
stimulates quality improvement activity at the hos-
pital level. The effect of public reporting on
M. Fletcher
effectiveness, safety and patient-centeredness re-
mains uncertain.
Boards and senior leaders in healthcare organi-
sations have a key role to play in building a culture
of safety and openness in their organisations.8 How
can they make the ‘public right to know’ more
meaningful?
Three actions are suggested. First, boards and
senior leaders need to pay close attention to the
purpose, design and implementation of informa-
tion that is publicly reported. A key question to ask
is ‘What do we want to achieve by publishing
patient care performance information?’ Second,
the public need to be involved in developing and
designing measures on which to report. A key
question to ask is ‘How have we involved key
audiences in the design of the measures?’ Third,
the presentation of information needs to be
improved. A key question to ask is ‘How can we
present information in ways that are easy to use
and understand?’
The example of hand hygiene
At its most basic, it could be argued that the
‘public right to know’ should commence with
a patient being able to ask healthcare workers
(HCWs) whether they have cleaned their hands. In
practice, this does not appear to happen often,
and what appears simple is, in fact, much more
complex.
Surveys with HCWs undertaken by the National
Patient Safety Agency’s ‘cleanyourhands’ cam-
paign suggest that nearly two-thirds of these staff
feel that being prompted by patients could help to
improve their hand hygiene behaviour.9 At the
same time, around one-quarter of these staff felt
that such a prompt would create tension between
HCWs and patients.
Surveys with hospital inpatients and members
of the public indicate that around half of the
respondents said that they were unlikely ever to
ask nurses or doctors whether they had cleaned
their hands. In practice, many fewer respondents
had actually asked this question of a doctor or
nurseethis is despite the fact that patient
involvement has been an important feature of
the cleanyourhands campaign in England and
Wales from its launch in 2004. Indeed, more
than 96% of trusts say that they involve patients
in their work to improve infection control.9 The
Chief Medical Officer for England has also drawn
attention to the importance of patient involve-
ment in influencing the hand hygiene behaviour
of staff.10
Hospital hygiene and infection: what should the public know?
Discussion
Despite the growth in publication of performance
data on HCAI, and public awareness of the risks of
infection, it appears that many members of the
public are not comfortable to ask HCWs whether
they have cleaned their hands before touching
them; further, that a significant proportion of
HCWs feel uncomfortable about being asked this
question. This would suggest that the ‘public right
to know’ does not yet extend to this most basic
information. In short, awareness and information
do not appear to be leading to change.
What lies behind this? Three factors are
suggested.
First, patients want to believe that their HCWs
are doing the right thing and do not want to
think that they may not be getting anything less
than the best possible care.11 Second, the ques-
tioning of a HCW often feels for both parties
like a criticism of the clinical skills of the HCW
rather than a helpful part of ensuring the safest,
cleanest care. This reflects broader findings on
patient involvement in patient safety.12 Third,
many healthcare organisations have not created
an environment in which both patients and
HCWs feel that it is indeed ‘OK to ask’. In short,
for many patients, the risk of perceived harm of
such a question appears to outweigh the benefit
of action. This is despite public concern about
HCAIs.
In conclusion, the public do have a right to know
about the infection rate in their hospital. More and
more information is being put into the public
domain. However, the availability of such infor-
mation is not enough on its own. Patient involve-
ment in hand hygiene is a good example of the
complexity and challenge. The fact that many
patients do not feel able to ask HCWs whether
they have cleaned their handseand that HCWs feel
uncomfortable being asked e suggests that the
399
‘public right to know’ has not yet fully reached the
bedside.
Conflict of interest statement
None declared.
Funding sources
None.
References
1. Gould DJ, Drey NS, Millar M, Wilks M, Chamney M. Patients
and the public: knowledge, sources of information and per-
ceptions about healthcare-associated infection. J Hosp
Infect 2009;72:1e8.
2. Gill J, Kumar R, Todd J, et al. Meticillin-resistant Staphylo-
coccus aureus: awareness and perceptions. J Hosp Infect
2006;62:333e337.
3. Department of Health. Handbook to the NHS Constitution.
London: DoH; 2009.
4. Health Protection Agency. Surveillance of healthcare
associated infections report. London: HPA; 2008.
5. Berwick DM, James B, Coye MJ. Connections between
quality measurement and improvement. Med Care
2003;41:30e38.
6. Shekelle PG, Yee-Wei L, Mattke S, Damberg C. Does public
release of performance results improve quality of care? A
systematic review. London: Health Foundation; 2008.
7. Fung CH, Lim Y-W, Mattke S, Damberg C, Shekelle PG.
Systematic review: the evidence that publishing patient
care performance data improves quality of care. Ann Intern
Med 2008;148:111e123.
8. Pittet D, Donaldson L. Clean care is safer care: the first
global challenge of the WHO World Alliance for Patient
Safety. Am J Infect Control 2005;33:476e479.
9. National Patient Safety Agency. It’s OK to ask: empowering
the hospital patient to ask about hand hygiene. 2009
[Unpublished observations].
10. Chief Medical Officer. On the state of public health: annual
report of the Chief Medical Officer. London: Department of
Health; 2006.
11. Nuffield Trust. Involving people in public disclosure of
clinical data. Nuffield Trust; 2003.
12. Davis RE, Koutantji M, Vincent CA. How willing are patients
to question healthcare staff on issues related to the quality
and safety of their healthcare? An exploratory study. Qual
Saf Health Care 2008;17:90e96.
Journal of Hospital Infection (2009) 73, 400e407
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

What are the drivers of the UK media coverage of

meticillin-resistant Staphylococcus aureus, the
inter-relationships and relative influences?
T. Boyce a,*, E. Murray b, A. Holmes b
a
The King’s Fund, London, UK
b
Imperial College Healthcare NHS Trust, Imperial College, London, UK

Available online 20 August 2009

KEYWORDS Summary Meticillin-resistant Staphylococcus aureus (MRSA) infection in

Hospital-acquired the UK receives intense media attention. The nature of coverage, political
infection; responses and solutions offered has been questioned and the relationship
Hospital cleaning;
between health professionals, the media and government policy needs
Mass media;
Meticillin-resistant
greater understanding. We identified 2880 articles on MRSA published in
Staphylococcus 12 UK newspapers between 1994 and 2005, compared with 21 articles in
aureus; six major US newspapers. To investigate the relative influences and rela-
News; tionships further, 68 weeks of coverage from 1990 to 2004 were analysed.
Risk communication The dates were selected based on publication dates of the ten most
frequently cited articles on MRSA according the ISI Web of Science portal
of Department of Health press releases on MRSA since 1997. Within this
period, 351 news articles were published with members of the public and
politicians representing 60% of sources quoted. Scientific articles, even
those with the highest number of citations, have negligible influence on
newspaper coverage. Simple solutions quoted in the newspaper articles
focused almost exclusively on cleaning. The UK press exhibits a high interest
in MRSA compared with that of the USA. Healthcare workers, experts and
professional bodies have criticised the nature of media reporting, but have
had little influence or involvement in the press. This may facilitate journal-
ists, celebrities, the public and politicians to drive these stories unchecked
and allow politics to address only the simplistic solutions generated.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

* Corresponding author. Address: Research Fellow in Public Health, The King’s Fund, 11e13 Cavendish Square, London W1G 0AN,
UK. Tel.: þ44 20 7307 2663; fax: þ44 20 7307 2809.
E-mail address: t.boyce@kingsfund.org.uk

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.05.022
UK media coverage of MRSA
Introduction
In a 2008 UK-wide British Broadcasting Corporation
(BBC) poll on the National Health Service (NHS),
acquiring an infection while in hospital ranked as
the main public concern about inpatient hospital
care (BBC News online: Infections ‘the biggest NHS
fear’; 30 June 2008). Concerns about infection
received a much higher rating (40%) than clinical
issues such as poor quality treatment (8%) or
insufficient staff to meet patients’ needs (10%).1
Meticillin-resistant Staphylococcus aureus (MRSA)
has become a major preoccupation of the UK
media, with tabloid headlines promoting mops
and matrons as an overarching simple solution,
but with more complex factors such as antibiotic
stewardship, bed management, isolation capacity,
staffing, behaviour and practice receiving little
coverage.2 This article explores these issues and
analyses the interactions and relative influences
of scientists, clinicians, policymakers and public
opinion on the media coverage of MRSA in the UK.
Since the early 1960s Western nations have
witnessed increasing numbers of infections associ-
ated with MRSA.3,4 In 2004 the UK National Audit
Office reported that from 2001 to 2004, there had
been an 8% increase in the number of reported
S. aureus bacteraemias (from 17 933 to 19 311) and
an increase of 5% in the number of cases attributed
to MRSA (from 7250 to 7647).5 The issue of MRSA in
England is unique for three reasons. First, in Octo-
ber 2000, the UK government announced that each
acute hospital trust in England would monitor and
report numbers of MRSA bacteraemias. The first
hospital-acquired infection (HAI) for mandatory re-
porting was to be MRSA bloodstream infection (BSI).
Second, a reduction target was set by the govern-
ment of 50% in three years. Third, alongside the
increase in MRSA BSI cases that was then being
seen in the UK, media coverage increased and inten-
sified, with MRSA often at the top of the UK news
agenda. This was a level of coverage that appeared
to be unprecedented in other national media.
Media representations of health issues are sig-
nificant as key sources of information for the public
both as patients making health decisions and as an
electorate. Media coverage can also drive health
policy.6,7 The role of media in driving change and
providing external reinforcement to drive change
must also be considered and understood, particu-
larly in light of the MRSA BSI reduction achieved in
England as reported in September 2008.8e10
UK national newspaper coverage of MRSA was
examined over a 13 year period, covering 30 key
dates when influential pieces of peer-reviewed
401
research were published and when the UK De-
partment of Health (DoH) issued press releases
about MRSA. Selecting these key dates when these
different sources might have influenced press
coverage allowed us to explore the relationship
and influence between the government, media and
published medical research.
Methods
Twelve quality and popular newspapers with the
highest circulation in the UK (Daily Telegraph,
Sunday Telegraph, Guardian, Observer, Daily
Mail, Mail on Sunday, Mirror, Sunday Mirror, Sun,
News of the World, The Times, Sunday Times)
and six major US newspapers: New York Times,
Los Angeles Times, Washington Post, USA Today,
Boston Globe and Chicago Tribune (widely
regarded as the elite of the US newspaper market
and highly influential on other media outlets) were
examined for their coverage of MRSA including the
number of articles published.
Selection of time periods for further analysis
of UK press
From 1990 to 2004, a total of 30 separate dates in
12 national UK newspapers were studied. 1990 is
the start of the sample as this is the earliest date
when most UK newspapers appear on the Lexis
Nexis subsription service database, which is
a searchable archive of international newspapers,
magazines, legal documents and other printed
sources. The dates are based on the activities of
two key sources of material on MRSA: medical and
scientific journals and the UK Department of
Health (DoH). The top ten cited articles published
between 1990 and 2004 with MRSA as a subject
were identified on the ISI Web of Science portal.
The publication dates of these papers were then
used as dates for analysis. ‘Most-cited’ status
may take several months, if not several years, to
be achieved; these articles are proxies to show
the type of coverage science research receives.
Tabloids and broadsheets can and do cover compli-
cated science stories, for example, coverage of
stem cell science demonstrates that tabloids do
report complicated science stories.11,12
All press releases about MRSA published on the
UK DoH website since 1997 provided an additional
19 dates. The Health Protection Agency (HPA) also
issues press releases related to MRSA; however,
most concern the rates of MRSA, some of which
appear on the DoH website. They issue their own
402
press releases, and two press releases in our
sample originated as HPA press releases. The HPA’s
own press releases are not included in the sample
as journalists primarily turn to the HPA for statis-
tics on MRSA rates. They are not widely regarded
as key media agenda-setting bodies.12
An additional date when scientists proactively
tried to engage the media about MRSA was also
selected. This was a press conference held in July
2004 at the Science Media Centre.
Each date was examined for a three-week
period; from one week before the press release
or journal article was published and the two weeks
afterwards. This period was used to include cover-
age before official press releases are issued (due to
unofficial leaks) and also to gauge how coverage
evolves after information from science and gov-
ernment sources has been released. Some dates
overlap, particularly in the summer and winter of
2004 when a number of DoH press releases were
published. Accounting for overlapping dates, a to-
tal of 68 weeks were analysed over 13 years. The
search term ‘MRSA’ was used, every article with
this term was analysed (duplicates articles were
omitted). Whereas both MRSA and M.R.S.A. are
used by journalists in both countries, for consis-
tency, MRSA was used as the search term e it is
Table I Newspaper articles identified on 30 key dates (pertaining to publication of Department of Health or
journal articles) from 1990 to 2004 containing the term ‘MRSA’ where Department of Health or journal articles
were cited
Documents analysed
press release is mentioned
Science journal articles:
Levine et al. (1991)13
Panlilio et al. (1992)14
Kreiswirth et al. (1993)15
Mulligan et al. (1993)16
Voss et al. (1994)17
Hiramatsu et al. (1997)18
Herold et al. (1998)19
Enright et al. (2000)20
Pittet et al. (2000)21
Kuroda et al. (2001)22
Science Media Centre press briefing
DoH press release dates:
12.07.04, 19.10.04, 04.11.04, 07.12.04
12.06.00, 09.06.03, 05.12.03, 15.10.04
08.02.02, 14.07.04
22.11.99, 21.07.04, 05.11.04
01.12.04
15.12.04
23.04.98, 03.09.98, 17.12.98
16.10.00
MRSA, meticillin-resistant Staphylococcus aureus; DoH, Department of Health.
T. Boyce et al.
more commonly used by journalists and there are
a larger number of articles using MRSA. This
selection process resulted in a total of 233 news
articles which mentioned the DoH press releases or
journal articles (Table I).13e22
The articles on MRSA were examined to identify:
(i) story leads; (ii) quoted sources; (iii) reported
causes of the problem; and (iv) reported solutions.
Results
MRSA reporting in UK versus USA
In 2004, a total of 737 articles about MRSA were
published in major UK national newspapers and by
2005 this had increased to 1514, a total of 2251
articles in just two years. By contrast, reporting in
the USA was significantly lower. Only 21 articles
were published in six major US newspapers from
1994 to 2005 (Figure 1). There were signs that this
was changing, with a marked increase from five
articles in 2005 to 125 articles in 2007. The main
themes of HAI media coverage differed. In the
UK coverage was primarily concerned with hospi-
tal-acquired infections, the NHS and governments
being blamed for poor response or lack of action
No. of articles where Main theme
0 Medical research
2 Medical research
3 Medical research
34 Cleaning
21 Solutions
18 Rates
15 Targets
3 Products
1 Science
0 Report
0 Monitoring
UK media coverage of MRSA 403

1600

1400

1200
No. of articles on MRSA

1000

800

600

400

200

0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Year

Figure 1 Meticillin-resistant Staphylococcus aureus (MRSA)-related news reporting by newspapers in the UK (dashed
line) compared with the USA (solid line) from 1995 to 2007.

to reduce HAI. By contrast, US coverage focused on
community-acquired infections, as there is no cen-
tral healthcare system; newspapers focused on in-
fections found in schools and sporting facilities,
and blamed individuals.
Analysis of UK press coverage over key dates
Between 1986 and 1994, MRSA received sporadic
coverage in the UK media. By 2000, MRSA reporting
had attained a momentum with celebrities de-
scribing their MRSA-related experiences. TV actress
Lesley Ash had meticillin-susceptible S. aureus
(MSSA) but the media widely reported that she
had MRSA. Agony aunt and former nurse Claire
Rayner, also president of the Patients Association,
frequently appeared in the media as a vocal vic-
tim of MRSA. MRSA was also seeping into fictional
media. In the opening episode of the 2006 Dr Who
series, a race of nun-like catwomen ran a hospital
with thousands of humans infected with a disease
secretly hidden in the basement. Screenwriter
Russell T. Davies said that when he was writing
the episode ‘(e)veryday MRSA was in the newspa-
pers.’22,23 Since 2002, MRSA has received substan-
tial coverage in the UK national press, particularly
in the popular newspapers.c A total of 233 articles
containing the term ‘MRSA’ appeared in the 12
newspapers.
Only five newspaper articles refer to scientific
journal articles: two cited Kuroda et al. (2001)22
(published in The Lancet) and referred to a Science
Media Centre briefing. By contrast, 95 articles
(41%) included information from a DoH press re-
lease (Table I).
Story leads
The lead of most stories was not about science and
data, but instead about victims or political state-
ments about MRSA. This emphasis on personal
narratives and political statements demonstrates
the difficulty for scientists or scientific research to
have a significant impact on how MRSA is covered.
Instead of the story being about systematic prob-
lems or public health challenges, it becomes

c
Traditionally UK newspapers were classified by size and hence quality: ‘tabloids’ were aimed at a ‘popular’ (mainly working class)
market whereas ‘broadsheets’ were ‘quality’ newspapers aimed at a more educated and wealthy market. In 2003 the first UK broad-
sheet changed its size to tabloid and many others soon followed. Hence the more common terms now used to describe UK newspapers
are ‘quality’ and ‘popular’.
404 T. Boyce et al.

a human interest story. Personal stories were more
common in the popular newspapers, where 44
stories led with this theme, compared with only
nine stories in the quality papers. Personal stories
put a ‘human face’ to medical or scientific stories
and are viewed by journalists as a way to capture
the audience’s attention.24 Scientific research
made up 15% of all leads, but what makes this sta-
tistic even more significant is that these stories
were chosen during periods when sources (the
DoH and scientists) were trying to control or con-
tribute to the news agenda.
Sources for quotations by journalists
An analysis of sources quoted directly by journal-
ists in their stories revealed that health profes-
sionals were absent from much of the coverage of
MRSA. The most common sources were members of
the public affected by MRSA, suggesting that the
dominant discourse in stories about MRSA is not
scientific, but rather a more general and access-
ible discourse (Table II).
Organisations such as the Hospital Infection Soci-
ety, the Infection Control Nurses Association (sub-
sequently known as the Infection Prevention
Society), the Royal Colleges (of Physicians, Surgeons,
Pathologists and Nursing) and the British Medical
Association are largely absent from the coverage.
The public was the most common source: the
‘common-sense’ solution of cleaning accounted for
69% of the solutions reported. This apparently
simple solution to MRSA offered by the public and
then reported by journalists is in contradiction to
those more complex solutions offered by health
professionals.
Reported causes
When the causes of MRSA were reported, jour-
nalists used simple discourses and rarely reported
the findings of scientific evidence or referred to
possible problems with NHS systems or structures.
Cleaners and cleanliness was the leading re-
ported cause and accounted for 102 (35.9%) of
all causes discussed. More complex ideas such as
the challenges of the poor isolation capacity in
NHS hospitals were mentioned in only 4.9% of
articles. Other ideas such as staffing levels, use of
agency staff, screening policies, or intravenous
line care were not mentioned at all (Table III).
Thirty-eight percent of stories contained no ref-
erences to the causes of MRSA, leaving a large
gap in terms of opportunities to build public
awareness.
Reported solutions
With the emphasis on cleaning as the primary
cause of MRSA, it is therefore no surprise that
the majority of reported solutions (Table IV) also
related to cleaning. Although environmental
cleaning has a role in contributing to the control
of HAI levels, the politically determined target
was a reduction in a specific infection caused by
a specific organism: hospital-acquired bacteraemia

Table II Newspaper articles identified on 30 key dates (pertaining to publication of Department of Health or
journal articles) from 1990 to 2004 containing the term ‘MRSA’ by sources directly quoted in all articles
Sources directly quoted in all articles No. of appearances % total sources
Public affected by MRSA 66 16.3
Politician (Labour Party) 52 12.9
Scientist/doctor 51 12.6
Politician other 42 10.4
DoH/HPA 41 10.2
Hospital representatives 36 8.9
Public e other 22 5.5
Pressure group 21 5.2
Other (e.g. lawyer, union, private health care, celebrity) 33 8.2
Health e other 12 3
Nurse 9 2.2
BMA/Royal Colleges 7 1.7
Infection Control Nurses Association 6 1.5
Hospital Infection Society 3 0.7
Cleaners 3 0.7
Total 404
MRSA, meticillin-resistant Staphylococcus aureus; DoH, Department of Health; HPA, Health Protection Agency; BMA, British
Medical Association.
UK media coverage of MRSA 405

Table III Newspaper articles identified on 30 key dates (pertaining to publication of Department of Health or
journal articles) from 1990 to 2004 containing the term ‘MRSA’ by reported causes of MRSA
Reported causes of MRSAa No. %
Cleaners/cleanliness 102 35.9
Hospital staff’s own hygiene/lack of knowledge 63 22.2
Other (e.g. government policy, visitors, shortage of consultants) 33 11.6
Government health targets 22 7.7
Invasive procedures 17 6
Over-prescription of antibiotics 16 5.6
Bed occupancy levels 14 4.9
Lack of isolation rooms 14 4.9
Shortage of consultants with infectious disease control expertise 2 0.7
Pressure on health professionals to prescribe antibiotics 1 0.4
Total 284
Articles not citing causes 88
MRSA, meticillin-resistant Staphylococcus aureus.
a
More than one cause can be included in a story.

caused by MRSA. It was this MRSA BSI data, sup-
plied regularly by the mandatory reporting
scheme, that was being monitored. The near
exclusive focus on environmental cleaning was
unhelpful in that it suggested that hospital cleanli-
ness alone contributed to bacteraemia prevention
and controlling antibiotic-resistant infection, and
that cleaning could ‘solve’ completely the problem
of increasing MRSA BSI. Some research also con-
cluded that there was ‘no direct link’ between
hospital cleanliness and the risk of MRSA bacterae-
mia.25 The way in which causes of the MRSA prob-
lem and the potential solutions are reported is
important because the media ‘can affect the types
of policy solutions the public and policymakers
consider’.26
Considering the level of information, or how the
problem is defined, it is apparent that although the
media did include some ‘facts’ about MRSA,
the government cannot depend on the media to
do the job of informing the public about MRSA,
especially the likely causes and potential solutions
to the problem. This responsibility lies with health
professionals.
Discussion
When considering the paucity of stories resulting
from the ten most highly cited scientific and
medical journal articles on MRSA, it is clear that
science did not influence media coverage of MRSA.
Medical journals do not have a significant impact

Table IV Newspaper articles identified on 30 key dates (pertaining to publication of Department of Health or
journal articles) from 1990 to 2004 containing the term ‘MRSA’ by reported solutions of MRSA
Reported solutions of MRSA No. %
Cleaner wards/hospital 99 38.1
Other (e.g. research, private healthcare, products, suing) 42 16.2
Matron/inspections 37 14.2
Cleaning product (or self-cleaning product) 18 6.9
New drug 15 5.8
Hospital guidelines 14 5.4
Patients asking staff to wash hands 10 3.8
Isolation units 7 2.7
Reduce drug use 6 2.3
Related to pets 5 1.9
Ringfencing 3 1.2
Screening for MRSA 2 0.8
Publishing MRSA rates 2 0.8
Total 260
MRSA, meticillin-resistant Staphylococcus aureus.
406
on how the MRSA story is reported.2,25 In the time
frame we examined, the coverage of MRSA demon-
strates that the media can drive policy away from
scientific evidence and towards popular, common-
sense solutions and continues to do so. Indeed,
media coverage ‘can be an integral part of how
policy is formed and re-formed as well as how it
is packaged in reports and then framed by press
releases’.27 For example, at the UK Labour Party
conference in September 2007, the Secretary of
State for Health, Alan Johnson, announced that
each hospital would undergo a ‘deep clean’. But
are the media the only ones to blame? Medical
journals are also responsible for failing to influ-
ence media coverage e less than half of the
medical journals promoted their research with
a press release. The combination of key govern-
ment sources promoting cleaning as a solution
and scientific sources not promoting their own
peer-reviewed research resulted in the story being
reported in a particular way.
Health professionals might also argue that the
UK government’s agenda is incorrect or that MRSA
is being unnecessarily politicised. They might
believe that the problem is more about poor
management or lack of resources and therefore
feel that contributing their opinions to the media
is redundant because the problem is not rooted in
‘medicine’ or ‘science’. There is little published
research to verify these views, although lack of
training has been identified as the reason that
scientists failed to participate in media cover-
age.28 If health professionals fail to engage with
the media or challenge the frames that the gov-
ernment or more ‘media-savvy’ sources try to
set, then they are failing to participate in a very
significant way in which the public learn and
inform themselves about science and health is-
sues. The media are more influential and impor-
tant when reporting new scientific stories, such
as MRSA.12,29,30
This is a health story which receives a great
deal of coverage and can be regarded as an
opportunity for scientists and health professionals
to engage with the public, to challenge the overt
politicisation of MRSA, to promote basic knowl-
edge of infection and risk and a more comprehen-
sive understanding of HAIs and the threat of
antibiotic resistance. At the same time, this is
a story that recognises the ability of the media to
drive change and improvement regardless of sci-
entific evidence. Some lessons can be learned
from the MMR (measlesemumpserubella) vaccine
story where many doctors did not get involved
because they believed that the science was in-
correct and thus did not require serious
T. Boyce et al.
refutations. Unfortunately health professionals
did not see the value of engaging with the media
to ensure public understanding until it was too
late, when vaccination rates had fallen signifi-
cantly and mumps and measles returned to schools
across the UK.12 In terms of addressing MRSA and
all HAIs, as a core aspect of patient safety and
quality of care it is not too late to engage with
the media and the public, and ultimately, to influ-
ence policy.8
Conflict of interest
None declared.
Funding sources
The authors acknowledge the support from the
National Institute for Health Research (NIHR)
Biomedical Research Centre Funding Scheme at
Imperial College and support from the Centre
for Patient Safety and Service Quality (CPSSQ)
funded by the NIHR.
References
1. Jolley S. Assessing patients’ knowledge and fears about
MRSA infection. Nurs Times 2008;104:32e33.
2. Anonymous. MRSA: Moving beyond mops and matrons.
Lancet 2007;370:362.
3. Reacher MH, Shah A, Livermore DM, et al. Bacteraemia and
antibiotic resistance of its pathogens reported in England
and Wales between 1990 and 1998: trend analysis. Br Med
J 2000;320:213e216.
4. Grundmann H, Ires-de-Sousa M, Boyce J, Tiemersma E.
Emergence and resurgence of methicillin-resistant Staphy-
lococcus aureus as a public-health threat. Lancet 2006;
368:874e885.
5. National Audit Office. Improving patient care by reducing
the risk of hospital acquired infection: a progress report.
Report No: HC876:3. London: Stationery Office; 2004.
6. Miller D, Kitzinger J, Beharrell P. The circuit of mass com-
munication: media strategies, representation and audience
reception in the AIDS crisis. London: Sage Publications;
1998. pp. 128e149.
7. Entwistle V, Sheldon T. The picture of health? Media cover-
age of the health service. In: Franklin B, editor. Social
policy, the media and misrepresentation. 1st ed. London:
Routledge; 1999. p. 118e134.
8. Holmes AH. Can organisational change reduce hospital-
acquired infections? J Hosp Infect 2007;65:191e192.
9. Lawson E, Price C. The psychology of change management.
McKinsey Q 2003;2:30e41.
10. Health Protection Agency. Health protection report,
September 19. Wkly Rep 2008;2:38.
11. Williams C, Kitzinger J, Henderson L. Envisaging the embryo
in stem cell research: discursive strategies and media re-
porting of the ethical debates. Sociol Health Illn 2003;25:
793e814.
12. Boyce T. Health, risk and news: the MMR vaccine and the
media. London: Peter Lang; 2007. pp. 186e188.
UK media coverage of MRSA
13. Levine D. Slow response to vancomycin or vancomycin plus
rifampin in methicillin-resistant Staphylococcus aureus
endocarditis. Ann Intern Med 1991;115:674e680.
14. Panlilio A. Methicillin-resistant Staphylococcus aureus in
United States hospitals, 1975e1991. Infect Control Hosp
Epidemiol 1992;13:582e586.
15. Kreiswirth B. Evidence for a clonal origin of methicillin
resistance in Staphylococcus aureus. Science 1993;259:
227e230.
16. Mulligan M, Murray-Leisure KA, Ribner BS, et al. Methicillin-
resistant Staphylococcus aureus e a consensus review of
the microbiology, pathogenesis, and epidemiology with im-
plications for prevention and management. Am J Med 1993;
94:313e328.
17. Voss A, Milatovic D, Wallrauch-Schwarz C, Rosdahl V,
Braveny I. Methicillin-resistant Staphylococcus aureus in
Europe. Eur J Clin Microbiol Infect Dis 1994;13:50e55.
18. Hiramatsu K, Aritaka N, Hanaki H, et al. Dissemination in
Japanese hospitals of strains of Staphylococcus aureus
heterogeneously resistant to vancomycin. Lancet 1997;
350:1670e1673.
19. Herold BC, Immergluck LC, Maranan MC, et al. Community-
acquired methicillin-resistant Staphylococcus aureus in
children with no identified predisposing risk. J Am Med
Assoc 1998;279:593e598.
20. Enright MC, Day NPJ, Davies CE, Peacock SJ, Spratt BG.
Multi-locus sequence typing for characterization of meth-
icillin-resistant and methicillin-susceptible clones of
Staphylococcus aureus. J Clin Microbiol 2000;38:
1008e1015.
407
21. Pittet D, Hugonnet S, Harbarth S, et al. Effectiveness of
a hospital-wide programme to improve compliance with
hand hygiene. Lancet 2000;356:1307e1312.
22. Kuroda M, Ohta T, Uchiyama I, et al. Whole genome
sequencing of methicillin-resistant Staphylococcus aureus.
Lancet 2001;357:1225e1240.
23. Byrne C. Doctor who takes on flesh-eating bugs and the
Almighty. The Independent March 2006;30:7.
24. Henderson L, Kitzinger J. The human drama of genetics:
‘hard’ and ‘soft’ media representations of inherited breast
cancer. Sociol Health Illn 1999;21:560e578.
25. Green D. Does hospital cleanliness correlate with meticillin-
resistant Staphylococcus aureus bacteraemia rates? J Hosp
Infect 2006;64:184e186.
26. Brown JD, Walsh-Childers K. Effects of media on personal
and public health. In: Bryant J, Zillmann D, Oliver MB, edi-
tors. Media effects: advances in theory and research. 2nd
ed. London: Routledge; 2002. p. 453e488.
27. Davidson R. Radical blueprint for social change? Media
representations of new labour’s policies on public health.
Sociol Health Illn 2003;25:532e552.
28. Gascoigne T, Metcalfe J. Incentives and impediments to
scientists communicating through the media. Sci Comm
1997;18:265e282.
29. Kitzinger J. Framing abuse: media influence and public
understandings of sexual violence against children. London:
Pluto; 2004. pp. 30e39.
30. Hornig-Priest S. Structuring public debate on biotechnology
media frames and public response. Sci Comm 1994;16:
166e179.
Journal of Hospital Infection (2009) 73, 408e413
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Are national targets the right way to improve

infection control practice?
M. Millar*

Barts and The London NHS Trust, Pathology and Pharmacy Building, 80 Newark Street,
London E1 2ES, UK

Available online 21 August 2009

KEYWORDS Summary The ‘right way to improve infection control practice’ should be
Healthcare-acquired cost-effective and lead to a fair distribution of infection control resources.
infection; Cost-effectiveness is a measure of aggregate ‘good’, and fairness empha-
Risk perception;
sises similar treatment for individuals under similar circumstances. The
Targets
UK national meticillin-resistant Staphylococcus aureus (MRSA) bloodstream
infection (BSI) target encourages National Health Service trusts to
prioritise strategies aimed at MRSA BSI prevention. Under resource-limited
conditions, the MRSA BSI target inevitably encourages deprioritisation of
patients at risk of non-target healthcare-associated infection (HCAI), some
of which are associated with an equal or larger burden of adverse outcome.
Established healthcare improvement strategies, such as the Plan, Do,
Study, Act (PDSA) cycle advocated by the Health Foundation, require the
setting of aims (or targets). If we are to improve infection control practice
then we need to decide on what to measure, how to measure it, and what
the improvement (target) should be. In selecting targets for infection pre-
vention, account should be taken of the contribution of HCAI to adverse
health outcomes overall. Human risk compensation behaviour and
microbial adaptation may both counteract the overall benefit of infection
targets isolated from overall outcomes. Risk taking is part of a healthy
healthcare system. We must be careful not to isolate HCAI outcomes from
overall outcomes or to isolate ‘risk takers’ from ‘risk controllers’. We must
try to limit the scope for human risk compensation and we must watch out
for microbial adaptation. Targets should be set locally, taking account of
fairness and cost-effectiveness. Locally relevant information is key;
positive incentives work best.
ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.

* Tel.: þ44 020 3246 0296; fax: þ44 020 3246 0325.
E-mail address: michael.millar@bartsandthelondon.nhs.uk

0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.02.022
National targets for infection control
Introduction
Leatherman and Sutherland trace the use of man-
agement targets back to the work of Drucker and
the development of the goal-setting theory of mo-
tivation.1e3 The Health of the Nation places health
targets within the context of the management of
the UK National Health Service.4 Some of the cri-
teria for the selection of targets specified in The
Health of the Nation include that the area chosen
should be a major cause of disability or avoidable
death defined using mortality, morbidity and qual-
ity of life data; that effective interventions are
available; and that the problem is quantifiable.
Criteria and proposals for the selection of targets
were also developed through the 1990s in a number
of other countries, for example Australia.5
Benefits of health targets include the consolida-
tion of higher level strategic objectives into local
policy and practice, consensus building (by opening
dialogue about priorities and expectations), facili-
tation of the monitoring of progress (by identifying
target measures), and the provision of a basis for
performance contracting and incentives. In review-
ing the evidence that regulation has improved the
quality of healthcare, Leatherman and Sutherland
point out that attainment of targets has been
greatest in areas amenable to central command
and control performance management, but ‘less
successful in those areas where intersectoral policy
and actions are needed or where complex social
behavioural change is required’.6 One of the prob-
lems with assessment of the impact of health tar-
gets is that target measures do not necessarily
provide information on the impact of the target
overall. This point is taken up in a systematic review
undertaken on behalf of the Health Foundation en-
titled ‘Does public release of performance results
improve quality of care?’7 This report concludes
that public release of information stimulates
change at the level of the hospital but that there
is little evidence that public release improves pa-
tient outcomes and that it may cause consumers
(on occasions) to make decisions that are inconsis-
tent with their own health goals. This discussion is
primarily focused on how targets are selected, but
it is important to remember that the way in which
targets that have been selected are then used is
another important consideration.
Current infection targets
Current UK National Health Service Healthcare
Associated Infection (HCAI) targets are specified
in the Better Care for All Public Service agreement
409
covering the period 2008/9 to 2010/11.8 Two
targets are specified: (i) the annual numbers of
MRSA bacteraemias should be <50% of the 2003/4
level for each year covered by the agreement
and (ii) the rates of Clostridium difficile in 2010/11
should be <30% of the 2007/8 number.
Potential dangers associated with
‘top-down’ targets
Bevan and Hood state that regulation by targets
assumes that priorities can be targeted; that the
part that can be measured stands for the whole;
and that what is omitted doesn’t matter.9 This per-
spective applies generally to targets but has par-
ticular resonance for the infection targets which
focus on a very limited range of infections and
at-risk populations. Millar et al. argue that the cur-
rent UK National Health Service (NHS) MRSA bac-
teraemia target forces NHS trusts to prioritise
control of MRSA bacteraemia, which accounts for
a small proportion (2%) of NHS HCAI.10 NHS trusts
are inevitably resource-limited, so the MRSA target
inevitably leads to deprioritisation of patients at
risk of other types of HCAI. In that sense the tar-
gets are not fair and, by deprioritising, the larger
proportion of non-target HCAIs may also not be
cost-effective. A recent paper from the USA
describes the focusing of resources on a single
antibiotic-resistant pathogen (such as MRSA) as
an ‘inherently flawed approach’ and goes on to
show that resources would be better spent on con-
trol of bloodstream infections overall rather than
MRSA bloodstream infections if we are trying to
maximise numbers of lives saved for a given expen-
diture.11 Some additional concerns relating to
national infection targets are that:
e they distract focus and constrain flexibility in
response to non-target and emerging
infections;
e infection outcomes are dissociated from other
(arguably more important) patient outcomes;
e if they are too narrowly focused, they leave
scope for human compensation behaviour and
for microbial adaptation.
It is almost impossible to predict future trends
in infectious diseases. This is well illustrated by
the over-optimism about the control of infectious
diseases during the 1960s, epitomised in the
statement by William Stewart, US Secretary
General in 1967, ‘It is time to close the book on
infectious diseases.’ What we do know is that
agents of infection are adaptable, that future
410
trends in infectious diseases are likely to have
a significant impact on humans, but that we cannot
predict why, where and in what form; so alertness
and ability to respond flexibly are key to the future
control of infectious diseases and in this respect
narrowly focused targets are associated with
a potential danger of distracting alertness to new
and emerging infectious diseases.
A narrowly focused infection target could be
achieved while the overall outcomes remain un-
changed or deteriorate, so there are also potential
dangers associated with the dissociation of in-
fection outcomes from overall outcomes. Arguably
preventable infection is important through the
contribution that preventable infection makes to
overall outcomes.
Risk perception and risk compensation
behaviour
Both individuals and institutions have to balance
risks and benefits, and there have been many
studies which describe individual and institutional
behaviour in response to perceived risks and
benefits associated with particular areas of
activity.12,13 That there are benefits as well as risks
associated with healthcare activities is acknowl-
edged within recent NHS guidelines and policy
statements; for example, the NHS Operating
Framework 2008/9 in section 2.16 states that ‘no
healthcare system can be entirely risk-free’ and
‘Best Practice Risk’ describes ‘an aspiration
towards positive risk management which takes
account of the beneficial as well as harmful con-
sequences of taking risks’.14,15 Even though the
concept that risks and benefits are associated
with risk-taking behaviour is established in the
NHS literature, nevertheless the Health Act 2006
Part 2 states that ‘managers in NHS organisations
Propensity to take risks
Balancing behaviour
Perception of risk of
adverse event
(such as infection)
‘Cultural filter’
Figure 1 Balancing risks and benefits (modified from p. 15 of Adams13).
M. Millar
are to ensure that patients are cared for in a clean
environment where the risks of healthcare associ-
ated infection are kept as low as possible’.16 For
the risks of infection to be kept as low as possible
would require that we do not take risks, imposing
many constraints and limits to the real and poten-
tial benefits of risk taking, such as might be associ-
ated with the everyday activities of NHS operating
theatres.
Individual perception of danger is a major
psychological determinant of our response to
risks.12,13,17 Adams has done much to develop the
ideas and arguments underpinning our understand-
ing of risk compensation and describes the rela-
tionship between perceived danger, rewards from
taking risks, adverse events from taking risks,
and our propensity to take risks (see Figure 1).
It is important that those on the front line who
provide healthcare benefits (such as those involved
in surgery) share the goals and objectives of those
whose jobs it is to work to minimise the risks of
infection (infection control professionals). In prac-
tice, risk controllers (such as infection control pro-
fessionals) tend to be geographically, managerially
and operationally separated from the risk takers who
provide healthcare. Insistence on the primacy of
overall outcomes (rather than narrowly focused in-
fection outcomes) has the potential to facilitate the
development of common goals between the ‘risk
takers’ who provide healthcare and the ‘risk control-
lers’ who manage (infection) risks. Infection targets
potentially bring conflict to this relationship by
emphasising narrowly focused infection targets
rather than overall outcomes for patients.
Risk perception
Slovic reports primary research and draws together
much of the literature on the perception of risk.12
Rewards (from taking risks)
Experience of adverse event(s)
National targets for infection control
Several studies have quantified the relative risks
and benefits of different types of activity. For
example, risks and benefits can be quantified for
motor vehicle driving by looking at how much we
spend on motor vehicles and the risk of death asso-
ciated with motor vehicles, with surgery, and with
vaccinations. Another way of quantifying percep-
tions of risk is to ask people to rank risks and
benefits associated with particular activities. Gen-
erally speaking, perceptions of risks and benefits
tend to show some positive correlation. Where
there is a lot of benefit associated with a particular
activity, people generally are prepared to carry
a larger degree of risk than when there is less ben-
efit. Some of the findings from research on risk
perception are shown in Table I.
Risk perception is a major determinant of
human risk behaviour; people often misperceive
risks, and the perception of the scale of risk can be
amplified by social factors such as the media. That
the perception of risk is an important determinant
of human behaviour is well illustrated by a graph of
road accident deaths in Sweden in the 1960s.13 In
September 1967 the laws were changed concerning
the side of the road on which people were allowed
to drive, and in association with this change there
was a dramatic decline in road accident deaths. In-
dividual motor vehicle drivers perceived a risk that
others might forget which side of the road to drive
on, so they changed their own behaviour e resulting
in a dramatic reduction in road deaths.
Risk compensation occurs when people react to
a safety initiative by acting less safely e
frequently because of a change in the perception
of danger (risk of an adverse consequence). The
question arises: if we have narrowly focused in-
fection targets, then does the human response to
those targets generate compensatory increases in
risk outside of the target area(s)? There are many
instances where safety interventions have not
generated the expected benefits, for example
Table I Risk perception
e Risk perception is a major determinant of human
‘risk’ behaviour
e Individuals have a perception of risk associated
with a certain activity (healthcare) or lifestyle
e Individuals accept a certain level of risk and
benefit associated with a certain activity (health-
care) or lifestyle
e Perception of the scale of risks can be amplified by
social factors such as the media
e The scale of risk may be misperceived
411
specially designed children’s seats for cars, cycling
helmets and speed cameras. Risk compensation
has also been described as the Achilles’ heel of
innovations in human immunodeficiency virus (HIV)
prevention. For example, improving treatment of
HIV reduces people’s perception of risk and may
have the adverse consequence of leading to more
risk of HIV acquisition.18
What perceptions are associated with the HCAI
targets that might lead to risk compensation? Some
relevant perceptions are given in Table II.
Most of those with experience of the manage-
ment of patients with infection are aware of the
potential for microbial adaptation, but there are
also dangers associated with separating infection
outcomes from overall outcomes because of
human risk compensation behaviour. Humans can
and often do change their behaviour in response to
safety measures, including (probably) HCAI control
interventions.
Have things really got better with
respect to HCAI overall?
Many millions of pounds have been spent in the UK
on the control of target infections since 2000. If we
accept that microbes can adapt to changes in human
behaviour and activities, and that humans can
change their behaviour in response to perceptions
of risk, then we might also ask whether the NHS
infection targets have produced an overall benefit.
The National Prevalence Study of HCAI showed
that in 1993/94, 9% of inpatients at a given time
were under treatment for an HCAI and that this
figure had fallen to 8.2% in 2006.19 It is difficult to
interpret this information because the duration of
hospital stay has steadily reduced over the last ten
years, so it is not clear that the differences in
prevalence and in the results from different types
of hospital could not be accounted for simply by
Table II Targets and perception of risk
Perceptions:
e the infection targets are the main priority
e failure to achieve the target is a threat (to the
individual and/or institution)
e non-target infections do not matter (on a personal
or institutional level)
e HCAI is less of a problem (now) because HCAI
control is on target
e ‘others’ are doing something about HCAI
HCAI, healthcare-associated infection.
412
changes in the denominator e those patients who
stay in hospital long enough to acquire a hospi-
tal-acquired infection. The prevalence studies
have been based on all of the inpatients present
on the day of the study, whereas it is known that
only a small and decreasing proportion of patients
admitted to hospital stay for >24 h. For example,
in one London teaching trust, of the 100 000 admis-
sions per year, 52 000 stay for less than a day,
15 000 stay for 24e48 h and only 9000 stay for
>48 h. Between 1995/6 and 2003/4 the mean
length of stay for NHS patients fell from 10 days
to 7.1 days. It is hard to know therefore if the
small reduction in prevalence of HCAI is attribut-
able to improvements in HCAI prevention or in-
stead reflects changes in the denominator
(resulting from shorter hospital stay).
Problems with reporting bias also make it diffi-
cult to assess whether HCAIs are being better
controlled now than before the instigation of
targets. Although MRSA deaths mentioned on death
certificates between 1993 and 2006 increased
dramatically, this may reflect reporting bias.
Unpublished data (derived from London teaching
trusts) suggest that MRSA bloodstream infection
rates fell between 1998 and 2008, while rates of
bloodstream infection with other types of bacteria
have been increasing. In one London teaching
hospital, increases in E. coli bacteraemia have bal-
anced the decreases in MRSA bloodstream infection
(Figure 2) during that period, so overall the rates of
bacteraemia caused by significant pathogens
(E. coli, Klebsiella spp., Pseudomonas aeruginosa,
meticillin-sensitive and -resistant Staphylococcus
aureus) are largely unchanged.
90
80
70
Patient episodes
60
50
40
30
20
10
0
99
02
05
08
19
20
20
20
Quarter
Figure 2 Changes in the frequency of Escherichia coli
and Staphylococcus aureus bacteraemia in a London
Teaching Hospital 1999e2008. Solid line: MRSA þ MSSA;
dashed line: E. coli.
M. Millar
UK Health Protection Agency data for 2003e2007
show an increase in the reports of bacteraemia
caused by Klebsiella and Enterobacter spp. against
a decline in the numbers of MRSA bacteraemias.
There were w4500 Klebsiella spp. bacteraemia
reports to the HPA in 2003 and >6000 in 2007
(>30% increase). This compares with 7698 MRSA
bacteraemia reports between April 2003 and March
2004 and 4926 reports in 2007 (36% reduction). That
the targets have led to changes in human behaviour
is well illustrated by the observation that most
trusts have an active process of root cause analysis
for MRSA bloodstream infections; on the other
hand, few NHS trusts routinely carry out root cause
analyses for other causes of bacteraemia.
How do we improve infection control
practice?
If we are to improve infection control practice
then we need to decide what to measure, how to
measure it and the scale of the improvement that
we are looking for. The national objective is to
improve people’s overall life expectancy and
reduce health inequalities.14 I would argue that
the objective of healthcare is to improve the
quality of life of patients, and that the objective
of infection control professionals is to control
infection risks in a manner that best serves to
improve the overall well-being and interests
(including quality of life) of patients. Some
proposed ideal properties of infection targets
are given in Table III.
Assurance of the quality of HCAI control
requires measures of both outcomes and process.
Table III Ideal properties of infection targets
e Relevant (to local priorities)
e Transparent to stakeholders
e Fair (do not add to inequalities in health outcomes
or resource distribution)
e Linked with cost-effective interventions/actions
e Measurable (reliable data retrieval, definable
parameters)
e Sensitive to changes in local policy and practice
e Do not separate ‘risk takers’ from ‘risk controllers’
e Take account of the possibility of risk compensation
e Do not constrain alertness and responsiveness to
non-target HCAIs
HCAI, healthcare-associated infection.
National targets for infection control
HCAI varies from place to place and the contribu-
tion or otherwise of HCAI to overall outcomes also
varies. HCAI performance can only be assessed
within a specific context, taking account of the
benefits as well as the risks of various healthcare
activities. Local targets encourage local owner-
ship, have the potential to encourage (if selected
on the basis of overall outcomes) engagement of
‘risk takers’ with ‘risk controllers’, and are more
likely to be locally relevant. An ethical argument
for local decision-making with respect to health-
care resource allocation is made by Ashcroft.20 The
Health Foundation states that the right targets,
used in an appropriate way, can improve health
quality and that it is important that systems and
services are designed in a way that ensures achiev-
ing targets in one area does not lead to worsening
standards in others.21 Our Safer Patient Initiative
demonstrates how much can be achieved when
hospitals set their own targets and are given the
right technical support.
The Plan, Do, Study, Act plan cycles proposed by
the Health Foundation are a simple approach that
anyone can apply. They reduce risk by starting
small; they can be used to plan, develop and
implement change; and they are highly effective.
Any government has responsibility to ensure
that taxpayers’ money is well used, but the
question is: ‘Are national targets the best way to
ensure that taxpayers’ money is well used?’ I have
argued that there are potential pitfalls with
national infection (control) targets. Wilde, who
has done much to develop our understanding of
human psychology with respect to risk compensa-
tion, describes four tactics to improve risk man-
agement, and these are based on perceptions of
risk.17 He suggests that we can increase the per-
ceived benefit of cautious behaviour, decrease the
perceived cost of cautious behaviour, increase
the perceived cost of risky behaviour and decrease
the perceived benefit of risky behaviour. He states
that, of all the accident countermeasures that are
currently available, those affecting people’s moti-
vation seem to be most successful, and those that
reward people for accident-free performance
seem to be the most promising.
In conclusion, I have argued that risk taking is
part of a healthy healthcare system, that we must
be careful not to isolate HCAI outcomes from
overall outcomes or to isolate ‘risk takers’ from
‘risk controllers’. We must try to limit the scope
for risk compensation and we must watch out for
microbial adaptation. Targets should be set
locally, taking account of fairness and cost-
effectiveness. Locally relevant information is
key, positive incentives work best.
413
Conflict of interest statement
None declared.
Funding sources
None.
References
1. Leatherman S, Sutherland K. The quest for quality: refining
the NHS reforms (1998e2008). London: The Nuffield Trust;
2008.
2. Drucker P. Management by objectives. New York: Harper
Collins; 1993 (first published 1954).
3. Mitchell TR, Daniels D. Motivation. In: Borman WE, Ilgen DR,
Klimoski RJ, editors. Handbook of psychology. Industrial
organizational psychology, vol. 12. New York: Wiley & Sons;
2003.
4. Anonymous. The health of the nation e a strategy for
England. London: Department of Health; 1992.
5. Nutbeam D, Wise M, Bauman A, Harris E, Leeder S. Goals and
targets for Australia’s health in the year 2000 and beyond.
Canberra: Australian Government Publishing Service; 1993.
6. Sutherland K, Leatherman S. Regulation and quality
improvement: a review of the evidence. London: The
Health Foundation; 2006.
7. Shekelle PG, Yee-Wei L, Mattke S, Damberg C. Does public
release of performance results improve quality of care?
London: The Health Foundation; 2008.
8. Anonymous. Better care for all public service agreement
2008/9e2010/11 objectives indicators. London: UK Depart-
ment of Health Strategic Framework; July 2008.
9. Bevan G, Hood C. Have targets improved performance in
the English NHS? Br Med J 2006;332:419e422.
10. Millar M, Coast J, Ashcroft R. Are meticillin-resistant Staph-
ylococcus aureus bloodstream infection targets fair to those
with other types of healthcare associated infection or cost-
effective? J Hosp Infect 2008;69:1e5.
11. Wenzel RP, Bearman G, Edmond MB. Screening for MRSA:
a flawed hospital infection control intervention. Infect
Control Hosp Epidemiol 2008;29:1012e1018.
12. Slovic P. The perception of risk. Virgina: Earthscan; 2000.
13. Adams J. Risk. London: Routledge; 2001.
14. Anonymous. NHS operating framework 2008/9. London:
Department of Health; 2007.
15. Anonymous. Best practice in managing risk. London:
Department of Health; June 2007.
16. Anonymous. The Health Act 2006. Code of practice for the
prevention and control of healthcare associated infections.
London: Department of Health; January 2008.
17. Wilde GJS. Target risk 2: a new psychology of safety and
health. Toronto: PDE Publications; 2001.
18. Cassell MM, Halperin DT, Shelton JD, Stanton D. Risk
compensation: the Achilles’ heel of innovations in HIV
prevention. Br Med J 2006;332:605e607.
19. Smythe ETM, McIlvenny G, Enstone JE, et al. Four country
healthcare associated infection prevalence study 2006:
overview of the results. J Hosp Infect 2008;69:230e248.
20. Ashcroft R. Fair rationing is essentially local: an argument
for postcode prescribing. Health Care Anal 2006;14:
135e144.
21. Anonymous. What impact does regulation have on the
quality of healthcare? Viewpoint. London: The Health
Foundation; 2007.
Journal of Hospital Infection (2009) 73, 414e417
Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

REVIEW

Responsibility for managing healthcare-associated

infections: where does the buck stop?
B.I. Duerden*

Department of Health, London, UK

Available online 17 September 2009

KEYWORDS Summary The prevention and control of healthcare-associated infections

Bacteraemia; (HCAIs) requires a tripartite partnership between clinicians and carers,
HCAI; managers and government/Department of Health (DoH) across the whole
Legislation;
health and social care community. Mandatory surveillance of meticillin-
Performance
resistant Staphylococcus aureus bacteraemia and Clostridium difficile in-
management;
Surveillance fection has shown a significant fall from peak numbers in 2003/04 and
2006, respectively, and there is now a zero tolerance approach to prevent-
able infections and poor practice. Success so far has been based on senior
management commitment, enhanced real-time surveillance, implementa-
tion of clinical protocols (high impact interventions, prudent prescribing),
improved hand hygiene and environmental cleaning, and training and
audit, backed up by a heightened performance management focus through
targets and legislation (Code of Practice). DoH improvement teams have
supported National Health Service trusts in implementing change. Respon-
sibility for managing HCAI is a combination of managerial responsibility
based upon compliance assurance that procedures and protocols are being
implemented and personal professional responsibility of all clinicians and
other healthcare workers.
Crown Copyright ª 2009 Published by Elsevier Ltd on behalf of The Hospital
Infection Society. All rights reserved.

Infection is different from all other medical spe-
cialties e it spreads! It creates a fascinating in-
terplay of biology between microbial populations
* Address: Department of Health, Wellington House, 133-155
Waterloo Road, London SE1 8UG, UK. Tel.: þ44 0207 972
4567; fax: þ44 0207 972 1001.
E-mail address: brian.duerden@dh.gsi.gov.uk
0195-6701/$ - see front matter Crown Copyright ª 2009 Published by Elsevier Ltd on behalf of The Hospital Infection Society. All rights reserved.
doi:10.1016/j.jhin.2009.06.027
and human populations, all of it heavily influenced
by human behaviour. Infection knows no bound-
aries across health and social care as bacteria travel
with their human hosts and it is the behaviour of
patients and healthcare professionals that deter-
mines the risk of healthcare-associated infection
(HCAI) and is the key to its prevention and control.
HCAI affects all health and social care settings
HCAI: where does the buck stop?
whether in the National Health Service (NHS) or
the independent sector. This is a wide community
embracing all parts of the NHS [Department of
Health (DoH), strategic health authorities with
their performance management responsibilities,
primary care trusts (PCTs) as commissioners and
providers of services, acute NHS trusts and foun-
dation trusts providing secondary and tertiary
care, mental health trusts and ambulance trusts]
and the independent sector (hospitals, indepen-
dent treatment centres and the great majority of
nursing and care homes). The Health Protection
Agency (HPA) has a key role in supporting infection
prevention and control activity across this wider
community.
The HCAI challenge for the NHS in England is
exemplified by the results of mandatory surveil-
lance of meticillin-resistant Staphylococcus aureus
(MRSA) bacteraemia (introduced in 2001) and Clos-
tridium difficile infection (introduced in 2004 for
patients aged >65 years and extended in 2007 to
those aged between 2 and 64 years). The MRSA
bacteraemia reports peaked in 2003/04 at 7700
cases and the C. difficile infections were 55 681
in patients aged >65 years in 2006 (which repre-
sented 75% of cases). Why had these numbers risen
so high? My personal view is that between 1970 and
2000 a dichotomy had developed between infec-
tion prevention and control and the modern medical
practice that had delivered increased life expec-
tancy, better treatment for many conditions such
as cancer and heart disease, increasingly complex
surgery and better management of many chronic
illnesses. Infection was considered to be a nuis-
ance, but we had antibiotics and vaccines. For
those in microbiology and infection control, this
was a fascinating period in an interesting specialty
with much to do. New infections were discovered
and described at a rate of one per year in that
period, including C. difficile infection in 1978,
but the subject had become the province of the
infection specialists rather than mainstream
medicine. Reducing HCAI needed a change of the
corporate mindset from one which created a sys-
tem to deliver specialist clinical care and, some-
times, as a secondary issue, took measures to
prevent infection, to one that creates a safe
environment for patient care and then delivers
specialist clinical care within that environment.
Responsibility for HCAI control is a tripartite
partnership between clinicians and carers who are
responsible for safe patient care and the diagnosis,
treatment, prevention and control of infection in
their patients; boards, chief executive and
managers who are responsible for providing the
corporate environment for good prevention and
415
control of HCAI, and for having systems in place to
ensure that it happens; and the government/DH
that sets standards, ensures priority for HCAI
prevention and control, sets targets and monitors
outcome, and uses the systems of performance
management to ensure that it happens throughout
the NHS. Infection prevention and control became
a statutory responsibility in the NHS with the
Health Act 2006 which introduced the Code of
Practice (CoP) for HCAI that applied to all NHS
bodies.1 Compliance was assessed by the Health-
care Commission through its annual health checks,
improvement notices and annual specialist inspec-
tions against the Code (commonly known as the
‘Hygiene Code’). The Health and Social Care Act
2008 (coming into force in April 2009) built upon
the 2006 requirements by establishing a single
regulator, the Care Quality Commission, and
extending the CoP to the independent sector and
all care settings (independent hospitals, nursing
homes and care homes).2 Registration, which is
already a requirement for the independent sector,
is extended to all NHS bodies and compliance with
the CoP will be a requirement for registration.
The current DoH HCAI programme was launched
with the Winning ways strategy published in
December 2003.3 In July 2004, the action plan
Towards cleaner hospitals and lower rates of
infection was published and this set a target for
a 50% reduction in MRSA bacteraemias in the NHS
in England by 2008.4 The NHS deserves to be con-
gratulated on achieving this target and the first
three-quarters of 2008/09 showed a reduction of
>60% from the quarterly average in 2003/04. The
target and performance management system
backed by legislation has ensured that infection
prevention and control are a top priority, but the
NHS cannot be complacent and must continue to
improve. The next phase of the strategy was set
out in Clean, safe care: reducing infections and
saving lives in January 2008.5 The target is a ceiling
and individual NHS trusts and the NHS generally
must aim to get as far below it as possible and to
maintain year-on-year reductions. This is led by
a zero tolerance approach to aim for a minimum
number of infections. Zero tolerance does not
mean that ‘there will be no infections’ as this is
biologically implausible. However, there should
be no tolerance towards preventable or avoidable
infections or of poor clinical practice, whether this
be in hand hygiene compliance, aseptic proce-
dures or imprudent antibiotic prescribing. The
aim is to get things right every time.
A new element of the programme is the decision
to screen all patients admitted to NHS hospitals for
MRSA colonisation. This requirement applies to all
416
elective admissions (with some exceptions for very
low risk procedures) which must be effected by
April 2009 and extended to all admissions by
2010/11 at the latest. The principle behind the
screening programme is that colonisation generally
precedes infection, and that a colonised patient is
at risk of developing infection themselves and is
a possible source of transmission to others. If
possible, MRSA-positive patients should be iso-
lated, but in any case, a positive screen should
be followed by decolonisation (alternatively
known as suppressive treatment) to reduce the
risk to the individual and the risk of transmission.
Decolonisation immediately reduces the bioburden
of MRSA to cover the period of highest risk and of
clinical interventions; it needs to be done as close
to the clinical treatment as possible. However, it
also needs to be recognised that re-emergence or
recolonisation can occur over a period of weeks or
months. An effective screening programme needs
partnership between hospitals and primary/com-
munity care settings to deliver the best service to
patients.
A similar partnership is needed to control C. dif-
ficile infection. The unacceptably high levels
reached in 2006 started to come down in 2007/08
but there is still much to be done. The government
set a target to reduce C. difficile infection by 30%
by 2010/11 and the indications are that this will be
achieved with significant reductions occurring
during 2008/09. However, the aim must be to drive
these numbers down as far as possible, and a new
and updated guidance document, Clostridium
difficile infection: how to deal with the problem,
was published in January 2009.6 The target is
based on PCT populations with a subset applied
to the acute hospital settings, so the two must
work together to achieve reductions. The numbers
for death certificates recording C. difficile infec-
tion as the underlying cause or a contributory
factor to deaths reminds us that this is an infection
with significant mortality, especially in elderly and
vulnerable people. In 2007, C. difficile was men-
tioned 8324 times on death certificates, with
4056 (49%) of these being given as the underlying
cause. By this stage, the overall number of cases
had started to fall but the awareness among
doctors of the contribution of C. difficile infection
to deaths was still leading to greater numbers
recorded on death certificates.
These improvements have been brought about
by markedly changed practice across all NHS
bodies from boardroom to ward and clinic. There
has been a sea-change in high level management
involvement with an emphasis throughout NHS
organisations on the requirements for infection
B.I. Duerden
prevention and control. The HPA has supported the
DoH with enhanced surveillance systems for MRSA
and C. difficile. At the clinicianepatient interface
there has been a strong focus on improved clinical
practice protocols through the Saving lives and
Essential steps packages.7,8 These comprise self-
assessment tools for NHS trusts and primary care
organisations that reflect the CoP core duties,
and sets of ‘high impact interventions’ (care bun-
dles) for clinical procedures, backed up by several
guidance notes. The care bundless help staff to
perform procedures properly, on every occasion,
through sets of headline instructions and a simple
audit tool focused on preventing the spread of
infection and performing invasive procedures
properly. There has also been a major emphasis
on cleanliness and hygiene. The ‘cleanyourhands’
campaign from the National Patient Safety Agency
continues to drive improved hand hygiene and
there has been a revitalisation of environmental
cleaning, particularly for areas where there are
or have been infected patients.9 All of this has
been backed up by improved training and a height-
ened performance management focus through the
targets and legislation. Management priority and
responsibility has recognised that HCAI prevention
and control is not just the responsibility of the in-
fection control teams but extends from the board
and chief executive to all staff with clinical owner-
ship being key to delivery. All NHS bodies were
required to appoint a director of infection preven-
tion and control, a senior person with clinical and
managerial responsibility and authority who could
ensure that infection prevention and control
programmes were properly implemented.
The measurement provided by the enhanced
surveillance systems was necessary to manage the
improvement programme. A national surveillance
programme provides consistent definitions and is
important for performance management and target
monitoring. Data from enhanced surveillance
identifies risk factors. However, the key to local
delivery is having local surveillance data provided in
a realistic and timely manner to all wards, units,
divisions, and specialties so that they can identify
their problem areas, detect outbreaks and episodes
of increased incidence, and monitor progress.
These local data need to be reviewed at all
management levels on a regular and frequent basis.
The care bundles focused on key invasive
clinical procedures that carried high risk of in-
troducing infection e the insertion and care of
central venous catheters, peripheral cannulae and
renal dialysis catheters, management of surgical
wounds, urinary catheters, patients who are
incubated and ventilated, and those who have
HCAI: where does the buck stop?
C. difficile infection. The supporting guidance
helps NHS trusts to implement MRSA screening pro-
cedures, ensure that their blood culture protocols
were giving reliable results, guided trusts in devel-
oping their antimicrobial prescribing framework
and identified the appropriate approaches to isola-
tion and cohorting of infected patients.
The programme has been promoted and sup-
ported through DoH improvement teams that have
visited most acute NHS trusts in England to help
identify areas for action and support the imple-
mentation of improved practice and procedures.
This has shown clearly that the prevention and
control of HCAI requires continued management
attention, including those commissioning and mon-
itoring health and social care services, improved
clinical care, antimicrobial stewardship to support
prudent antibiotic prescribing, audit of protocols
and procedures, and training for all health and
social care staff. Making this happen requires
a partnership between management responsibility,
which requires compliance assurance from board
to ward, and personal responsibility of all staff
involved, which should be set out in job descrip-
tions and job plans, be supported by mandatory
training and continued professional development,
be included in appraisal in individual performance
review, and at times backed up by disciplinary
measures if there is continued failure to im-
plement the required measures. The compliance
assurance requires a combination of surveillance
data to identify cases of infection, outbreaks and
deaths, together with audit results of key activities
such as hand hygiene, implementation of clinical
protocols, isolation protocols, appropriate anti-
biotic prescribing and cleanliness inspections. All
of these should be reviewed at all management
levels from unit and ward to directorate, division
and board. Those responsible for commissioning
services and monitoring delivery need to include
a requirement of infection prevention and control
in all their commissioned service and their con-
tract monitoring should include a review of such
provision in terms of target numbers of infections
and process monitoring.
In implementing the HCAI programme the govern-
ment/DoH has used the political imperative,
measurement through surveillance, target setting,
professional support, performance management
and legislation, all with the aim of changing human
behaviour (clinical and managerial) to overcome
the challenge of HCAI. This has been a wake-up
417
call for all involved in providing NHS services. It is
no longer acceptable for infections to be regarded
as normal. Patients can be very ill, some may die,
many will stay in hospital longer and some will
need major surgery as a result of infections. The
treatment of these infections consumes significant
NHS resources that can be better used. Against this
background, zero tolerance of HCAI is a profes-
sional obligation for everyone, be they clinicians
and healthcare support staff in all units, wards,
directorates, specialties, executive directors and
board members in acute NHS trusts, PCTs and
strategic health authorities, both commissioners
and providers and extending to DoH and govern-
ment. It requires a partnership across the whole of
the health and social care community to provide
an unbroken chain of care in order to break the
chain of infection.
Conflict of interest statement
None.
Funding sources
This article describes the work of the Depart-
ment of Health/NHS HCAI programme of which
the author is Clinical Director. There is no
external funding of the programme.
References
1. Health Act 2006eCode of practice for the prevention and
control of healthcare associated infections. London: DoH;
2006.
2. Health and Social Care Act 2008eCode of practice for the
prevention and control of healthcare associated infections.
London: DoH; 2009.
3. Department of Health. Winning ways: working together to
reduce healthcare associated infection in England. Report
from the Chief Medical Officer. London: DoH; 2003.
4. Department of Health. Towards cleaner hospitals and
lower rates of infection: a summary of action. London:
DoH; 2004.
5. Department of Health. Clean, safe care: reducing infection
and saving lives. London: DoH; 2008.
6. Department of Health. Clostridium difficile infection: how
to deal with the problem. London: DoH; 2009.
7. Department of Health. Saving lives: reducing infection,
delivering clean and safe care. London: DoH; 2007.
8. Department of Health. Essential steps to safe, clean care.
London: DoH; 2006.
9. National Patient Safety Agency. Ready, steady, go! The full
guide to implementing the cleanyourhands campaign in your
Trust. London: NPSA; 2004.
Journal of Hospital Infection (2009) 73, 418e419
Available online at www.sciencedirect.com
LETTER TO THE EDITOR
Review of Lancet conference on healthcare-
associated infections
Madam,
I attended this Lancet conference at the Queen
Elizabeth II Conference Centre in London on 11
and 12 December 2008 as a final year specialist reg-
istrar in medical microbiology and virology. Before
attending the conference, I reviewed the pro-
gramme for the two day conference and saw that
it covered many of the infection control issues faced
by infection control professionals in UK hospitals.
The presentations were relevant to current
problems in hospitals, such as epidemiology and
management of meticillin-resistant Staphylococcus
aureus (MRSA) and Clostridium difficile infection,
specific pathogens such as norovirus, glycopep-
tide-resistant enterococci (GRE), extended-
spectrum b-lactamase-producing coliforms (ESBLs)
and specific clinical problems such as ventilator-
associated pneumonia (VAP), catheter-related
bloodstream infections (CRBSIs) and surgical site
infections (SSIs). In addition, the speakers were
from various countries in Europe and the USA, which
I found interesting as this would give a flavour of
what was happening outside the UK.
In the first couple of talks e the history and
international epidemiology of healthcare-associ-
ated infection (HAI) were presented, which I
thought was a good starting point. The idea of
developing a surveillance system in the UK and USA
for multiresistant organisms and information tech-
nology systems that can have an impact on reducing
HAI and antimicrobial resistance was useful. The
next talk entitled ‘International comparison of the
social, media, and political context of HAIs’ was
very different and the speaker pointed out the
enormous amount of media interest, particularly in
issues relating to MRSA and C. difficile, and espe-
cially in the UK. There was an interesting debate
on the extent and frequency of doctors’ or scien-
tists’ involvement with the media to prevent
0195-6701/$ - see front matter ª 2009 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2009.03.029
www.elsevierhealth.com/journals/jhin
misinformation being published and to keep
the public better informed. It left me with the opin-
ion that microbiologists and other infection special-
ists should perhaps take time to actively engage
themselves with the media more often.
The programme discussing specific pathogens
was well chosen e GRE, Pseudomonas aeruginosa,
ESBL-producing coliforms and norovirus. I picked
up some interesting facts from each of these; for ex-
ample, the reasons for the propensity for
Enterococcal Clonal Complex 17 to cause hospital
infections and the presence of specific virulence
factors such as enterococcal surface protein, which
is expressed on the cell surface and is responsible
for biofilm formation; also the fact that ampicillin-
resistant entercocci were the precedents of VRE in
hospitals. The presentation on norovirus by Marion
Koopmans from The Netherlands addressed the
causes for the current epidemics of this virus. Her
data demonstrated that the norovirus GII4 genotype
has dominated since 2002 and she outlined the dif-
ferences in the epidemiology between GII4 and
the non-GII4 isolates. The hypothesis proposed for
the current prevalence of GII4 and its variants is
that GII4 strains differ in their host binding proper-
ties and therefore the population segment targeted
and level of immunity in that group (which was
based on molecular epidemiological data and math-
ematical modelling). I would have liked to hear
more on molecular epidemiology and newer
methods of diagnosis and management of some of
these organisms.
The next category of talks about common HAIs
(SSI, VAP and CRBSI) which are burning issues in UK
hospitals today was extremely relevant. Particu-
larly useful were the debate regarding the need
for an agreed surveillance methodology allowing
for the complexity of SSI and the data on post-
discharge diagnosis, as well as the current evi-
dence for the prevention of VAP and CRBSI.
The last set of presentations for the day carried
several important messages regarding basic infec-
tion control practices: the use of alcohol hand
rub in developing countries; the concept of
Letter to the Editor
‘hand-touch sites’; and the newer molecular
methods available for diagnosing specific patho-
gens and identifying antibiotic resistance with
application of these in infection control practice,
e.g. multiplex polymerase chain reaction, IDI
MRSA assay for screening for MRSA, and finger-
printing techniques (pulsed-field gel electrophore-
sis and variable number tandem repeat typing).
The latter might better have been placed earlier in
the day, perhaps before lunch!
Day 2 started with a talk by Brian Duerden entitled
‘Responsibility for managing HAIs: where does the
buck stop?’ He presented the UK Department of
Health’s current HAI programme with focus on MRSA
bacteraemias and C. difficile infection. Key themes
in his talk were ‘zero tolerance’ and ‘everyone’s
responsibility’. This was followed by a session on
specific pathogens, and this time the topics were
hospital- and community-acquired MRSA, C. diffi-
cile and acinetobacter. The mathematical model-
ling showing transmission of MRSA between
hospital and community was novel and interesting.
The afternoon session was on controversies. The
talk on the importance of infection control versus
antibiotic manipulation was truly controversial
and the contrast was made between ‘infection
control and firefighting’ and ‘antibiotic stewardship
and prevention’. The difficulties arising from the
absence of a standardised acceptable method for
measuring antibiotic use were noteworthy. I was
419
amazed when I heard the talk about the Dutch
infection control experience, and although much
of it is unthinkable in the UK or other countries, it
was fascinating to hear how aggressively MRSA was
managed in The Netherlands.
On the whole, although there was some repeti-
tion of material that was presented especially on
epidemiology of MRSA and C. difficile in the UK, and
although some of the material was probably aimed
at the wider audience which included the media,
administrators, regulators and policymakers as
well as representatives from commerce and indus-
try, I found this conference both useful and refresh-
ing. The conference was especially helpful to
trainees planning to take their final Royal College
of Pathologists exams e certainly the sessions that
covered specific pathogens and problems, and all
the presentations are freely available on the
conference website (http://www.hai.thelancetcon-
ferences.com/App/homepage.cfm?moduleid=4646
53&appname=100585).
M. Meda
St George’s Hospital NHS Trust, London, UK
Tel.: þ44 1483 766950.
E-mail address: manjulasudheer@yahoo.com
Available online 9 July 2009