Are Routine Cranial Ultrasounds Necessary

in Premature Infants Greater Than 30 Weeks

Gestation?
N. Ja’Neice Harris, M.D.,1 Diana Palacio, M.D.,2 Andrew Ginzel,3
C. Joan Richardson, M.D.,1 and Leonard Swischuk2

ABSTRACT

The purpose of this study was to validate the recommendation of the American
Academy of Neurology and the Child Neurology Society that screening cranial ultra-
sonography be performed routinely on all infants of less than 30 weeks gestation at 7 to
14 days of age and again between 36 and 40 weeks postmenstrual age, and, by using this
practice parameter, to determine the number of babies with a clinically significant abnormal
screening cranial ultrasound (US) who would otherwise have been missed. A retrospective
study of 486 infants of 30 to 33 weeks gestation born January 1, 1999 to June 30, 2004 was
done. All had screening cranial ultrasounds. Grade III and/or grade IV intraventricular
hemorrhage (IVH) occurred in 4 (0.8%) infants of 30 to 31 weeks gestation. Infants with
significant IVH had either risk factors for brain injury or symptoms that would eventually
warrant US during their hospitalization. Seven (1.4%) infants had periventricular leuko-
malacia (PVL). All infants with a final diagnosis of PVL had pre- and/or perinatal risk
factors associated with PVL. There was a significant trend toward fewer abnormal cranial
ultrasounds from 30 to 33 weeks gestation (p ¼ 0.04). Our study supports the recom-
mendation by the American Academy of Neurology and the Child Neurology Society that
screening US can be limited but suggests that the gestational age cut off should be 30 weeks
or less.

KEYWORDS: Head ultrasounds, intraventricular hemorrhage, premature infants

C ranial ultrasounds (US) are an established
modality for detection of brain injury in very low birth-
weight infants. In the practice guideline for neuroimag-
ing of the neonate, the American Academy of Neurology
and Child Neurology Society states that routine US are
unnecessary in premature infants of 30 weeks gestation
or greater. Their recommendation is that routine screen-
ing cranial US be performed on all infants of less than 30
weeks gestation at 7 to 14 days of age and again between
36 and 40 weeks postmenstrual age.1 The recommen-
dation was based on review of several studies which
showed that the significant abnormalities (grade III/IV
intraventricular hemorrhage [IVH], periventricular leu-
komalacia [PVL], and ventriculomegaly [VM]) that
would alter treatment or provide prognostic information
were more common (20 to 25%) in infants with a
gestational age of less than 30 weeks.2–5
It has been our institutional practice to obtain a
screening US at 7 to 10 days of age on infants 33 weeks
gestation or less and/or birthweight less than 1600 g and
to repeat the US at 40 weeks postmenstrual age for
infants 30 weeks gestation or less and/or birthweight less

1
Division of Neonatology, 2Department of Radiology, 3Senior Medical
Student, University of Texas Medical Branch, Galveston, Texas.
Address for correspondence and reprint requests: Dr. N. Ja’Neice
Harris, 301 University Blvd, Route 0526, Galveston, TX 77555.
Am J Perinatol 2007;24:17–22. Copyright # 2007 by Thieme
Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001,
USA. Tel: +1(212) 584-4662.
Accepted: July 3, 2006. Published online: November 8, 2006.
DOI 10.1055/s-2006-954960. ISSN 0735-1631.
17
18 AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 24, NUMBER 1
than 1250 g. To validate the practice parameter sug-
gested by the American Academy of Neurology, a
retrospective study was performed on infants 30 to 33
weeks gestation admitted to our neonatal intensive care
unit (NICU) over a 5.5-year period. Our objective is to
determine the number of infants with a clinically sig-
nificant abnormal US who would have been missed had
we used this practice parameter.
METHODS
Appropriate subjects for this retrospective study were
identified using our computerized patient database.
Medical records were reviewed on all preterm infants
30 to 33 weeks gestation admitted to the NICU at the
University of Texas Medical Branch from January 1,
1999, to June 30, 2004. Screening US was routinely
obtained at 7 to 10 days of age on all infants, or earlier,
if clinically indicated. A second US was obtained at
40 weeks postmenstrual age or at discharge on infants
30 weeks gestation or less and/or birthweight less than
1250 g. The ultrasound examinations were performed
using a Logic-Book Ultrasound Scanner (General Elec-
tric, Milwaukee, WI) with an 8-MHz convex trans-
ducer. Examinations were performed using criteria
established by the American Institute of Ultrasound in
Medicine, which use a minimum of 14 images obtained
through the anterior fontanel. All ultrasound findings
were reviewed and finalized by a faculty pediatric radi-
ologist. IVH was described and reported using Papile’s
classification.6 In infants with varying degrees of IVH,
the higher grade of IVH was used for graphing and
statistical purposes. Infants excluded from statistical
analysis were those with known genetic anomalies, in
utero diagnoses of cerebral abnormalities, or gestational
age not between 30 to 33 weeks. Comparisons of intra-
Figure 1 Ultrasound results.
2007
cranial abnormalities among each gestation age group
were analyzed with the x2 for linear trend test.
RESULTS
A total of 486 infants 30 to 33 weeks gestation had
screening US performed from January 1, 1999, to June
30, 2004. Ninety-one percent (n ¼ 442) of the study
population had a normal US. Forty-four had an abnormal
US. The abnormalities included IVH, PVL, VM, mid-
parietal hemorrhage, colpocephaly, lissencephaly, agen-
esis of the corpus callosum, absent cavum septum pellu-
cidum, and choroid plexus cyst. IVH occurred in 4.3% of
patients (n ¼ 21; Fig. 1). Four infants had grade III or IV
IVH. The incidence of clinically significant IVH in our
population was 0.8%, compared with an average inci-
dence of 6.9% in the studies used as the basis for the
published practice parameter.7–10 Of the four infants with
significant IVH (grade III or IV), one was 30 weeks,
three were 31 weeks gestation, and all four were inborn
patients (Fig. 2). All four infants with significant IVH
had either risk factors or clinical symptoms that would
have warranted a cranial US during their hospital course
(Table 1). Those risk factors and clinical symptoms
included placental abruption, hypoxic ischemic encephal-
opathy, seizure activity, severe respiratory distress
requiring prolonged ventilation, or increased head cir-
cumference with widened sutures. Infant 1 had a screen-
ing US on day of life (DOL) 7, which showed a left grade
II and right grade III hemorrhage, and required ventila-
tor support for the first week of life. Despite serial
ventricular taps, infant 1 developed hydrocephalus and
required a ventriculoperitoneal shunt.11 Infant 2 required
ventilator support for the first 10 days of life, developed
seizures, and had an US on DOL 5, which showed left
grade III and right grade I IVH. Infant 3 had a screening
 ROUTINE CRANIAL ULTRASOUNDS IN PREMATURE INFANTS/HARRIS ET AL 19

Figure 2 Grades of intraventricular hemorrhage.

US on DOL 7, which demonstrated a right grade III and
left grade IV IVH. Infant 3 also required a ventriculo-
peritoneal shunt for hydrocephalus. Infant 4 had a low
biophysical profile (2/10), no fetal movement for 1 day
prior to delivery, low Apgar scores, seizure activity at
birth, and hypotension requiring pressor support. Early
US showed right grade III IVH and left IV IVH
extending into the periventricular parenchyma.
PVL was seen in 1.4% (n ¼ 7) of the screening
cranial US. The majority of the infants were 31 weeks
gestation and greater than 1500 g. Only one infant was
born outside of the institution. Four of the seven infants
with PVL on screening US had no evidence of PVL on
the US obtained at 40 weeks postmenstrual age. All
infants with a final diagnosis of PVL were inborn
patients, and had one or more pre- and/or perinatal
risk factors associated with PVL3,12 (Table 2).
Ventriculomegaly was seen in 2.3% (n ¼ 11) of
the patients. The grading system for ventricular dila-
tion is mild (0.5 to 1 cm), moderate (1 to 1.5 cm), and
severe (> 1.5 cm).5 All subjects had mild ventricular
dilation. Five infants had other US abnormalities:
colpocephaly, agenesis of the corpus callosum, absent
cavum septum pellucidum, choroid plexus cyst, and
mid-parietal hemorrhage. One infant had colpocephaly
and lissencephaly. This patient was later identified to
have a terminal deletion of chromosome 17p13.3
(Miller-Dieker syndrome).
Of the 486 study patients, only 10 (2%) had
abnormal screening US that were of clinical significance,
defined as any abnormality that could negatively affect
the neurodevelopmental outcome of an infant. These
included grade III/IV IVH, PVL, and moderate to
severe VM. A x2 value for linear trend test was used to
compare the percentages of clinically important abnor-
mal US (Table 3). There was a statistically significant
trend toward fewer abnormal US from 30 to 33 weeks
(p ¼ 0.04).
The cranial US of the 10 infants with significant
abnormalities done at either 40 weeks postmenstrual
age or at discharge were reviewed. In the IVH group,
the cranial US were consistent with resolving IVH and
the presence of a ventriculoperitoneal shunt placement
in the two infants who required that intervention. Of the
seven infants with an initial diagnosis of PVL, only
infants 3, 4, and 7 had late findings of PVL (0.6% of

Table 1 Hospital Course and Demographic Information of Four Infants with IVH
Characteristic Infant 1 Infant 2 Infant 3 Infant 4

Gestational age (wk) 30 31 31 31

Birthweight (g) 1365 1570 1695 1980
Mode of delivery Vaginal Cesarean section Vaginal Cesarean section
Apgar scores 1 min, 6; 5 min, 7 1 min, 8; 5 min, 9 1 min, 7; 5 min, 9 1 min, 2; 5 min,
6; 10 min, 8
IVH grade II and III I and III III and IV III and IV
Ventilator support Y Y N N
Seizure activity N Y N Y
Hydrocephalus Y N Y N
VPS Y N Y N
IVH, intraventricular hemorrhage grade; VPS, ventriculoperitoneal shunt; Y, yes; N, no.
20 AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 24, NUMBER 1 2007

Table 2 Hospital Course and Demographic Information of Seven Infants with PVL
Characteristic Infant 1 Infant 2 Infant 3 Infant 4 Infant 5 Infant 6 Infant 7*

Gestational 32 31 30 33 31 31 31
age (wk)
Birthweight (g) 2155 1392 1645 1630 1865 1430 1980
Mode of Cesarean Cesarean Cesarean Cesarean Cesarean Cesarean Cesarean
delivery section section section section section section section
Apgar scores 1 min, 4; 1 min, 6; 1 min, 3; 1 min, 9; 1 min, 8; 1 min, 7; 1 min, 2; 5 min,
5 min, 4; 5 min, 8 5 min, 7 5 min, 9 5 min, 9 5 min, 8 6; 10 min, 8
10 min, 7
Placental Y N N Y N N N
abruption
Infections N N CONS meningitis; N Serratia N N
Pseudomonas conjunctivitis
pneumonia
Ventilator Y N Y N N N N
support
Seizure Y N Y N N N Y
activity
Other MVA PTD FAS hypotension Maternal Monochorionic, Maternal No fetal
PIH diamniotic twin tobacco use movement
1 d PTD,
poor BPP
*The same as infant 4 in Table 1.
PVL, periventricular leukomalacia, Y, yes; N, no; CONS, coagulase-negative Staphylococcus epidermis; MVA, motor vehicular accident; PTD,
prior to delivery; FAS, fetal alcohol syndrome; PIH, pregnancy induced hypertension; BPP, biophysical profile.

the study population). Other findings in the PVL group routine screening US were limited to infants less than 30
were mild VM and choroid plexus cyst. weeks gestation. All 10 of these infants had clinically
There were a total of 388 infants of 24 and 29 significant abnormal US findings associated with an
weeks gestation who were born during the study period. increased risk of adverse neurodevelopmental outcome.
A total of 70 had abnormal US, and 38 (9.8%) US were As noted in the results, each infant had either historical
of clinical significance. Significant IVH, PVL, VM, and or clinical indications warranting US examination. Had
other anomalies occurred 4.6%, 3.6%, 0.3%, and 1.3%, a screening US not been obtained, infants 1 and 3 in the
respectively. The other abnormalities found on US IVH group might not have been diagnosed and under-
included porencephalic cyst; thalamic, posterior fossa, gone surgical intervention until much later. Arguably,
and frontal lobe hemorrhages; and extensive subarach- however, an US would likely have been clinically
noid hemorrhage with intraparenchymal extension. mandated based on the findings of accelerated head
growth in the two infants (infants 1 and 3) who
ultimately required surgical intervention. Infants 2
DISCUSSION and 4 did not have accelerated head growth, but both
In our study of 486 infants of 30 to 33 weeks gestation, subjects exhibited seizure activity that would have
10 infants with abnormalities would have been missed if warranted an US.
Four of the seven infants with PVL on screening
Table 3 Distribution and Percentage of Clinically US had no evidence of PVL on the US done at 40 weeks
Significant Abnormal Ultrasound in Each Gestational
postmenstrual age. It will be important to determine the
Age Group
developmental outcome of this group. Volpe reports that
Gestational Age (wk)

70% of infants with periventricular white matter in-
Status 30 31 32 33 jury, especially those with small focal areas of necrosis,
diffuse gliosis, and myelin loss, are not detected consis-
Abnormal US 3 5 1 1
tently by ultrasonography.4 It is possible that magnetic
Normal US 81 109 179 107
resonance imaging would have detected abnormalities
% 3.6 4.4 0.6 0.9
not seen on the US done at 40 weeks.
p ¼ 0.04 by x2 for linear trend toward fewer abnormal US from 30 to
33 weeks gestation.
Mild VM (0.5 to 1.0 cm) was seen in 11 of the
US, ultrasound. subjects in the study. The degree of VM is a prediction
 ROUTINE CRANIAL ULTRASOUNDS IN PREMATURE INFANTS/HARRIS ET AL
of adverse neurodevelopmental outcome for preterm
infants. Ventricular dilation greater than 1 cm taken at
the midbody of the lateral ventricle on sagittal scan is
associated with cerebral palsy, mental retardation, and
neuropsychiatric disorders at 2 to 9 years of age.1
Although ventriculomegaly was seen in infants with
grade III/IV IVH, no infant with a sole diagnosis of
VM had ventricular dilation greater than 1 cm.
There were several infants who had intracranial
abnormalities other than IVH, PVL, and VM. Three
were greater than 30 weeks gestation. Those infants with
significant findings had clinical and historical findings,
such as dysmorphic facies and trauma, which would have
warranted an US.
Among the infants less than 30 weeks gestation,
the proportion with clinically significant US abnormal-
ities was greater than those of 30 to 33 weeks gestation.
It is expected that the more immature the infant, the
greater the likelihood of intracranial abnormalities and a
more complicated clinical course. Upon data review,
most but not all infants in the group greater than 30
weeks gestation had a clinical indicator or history that
warranted screening US.
Based on the review of our population, we have
amended our screening US policy. Now, infants 30
weeks gestation or less receive a routine screening US
on DOL 7 to 10 and a follow-up US at 40 weeks
postmenstrual age or at discharge. Our decision to use
30 weeks as the lower limit is a conservative one, and is
based on our finding of one infant who had significant
IVH.
In our study, infants greater than 30 weeks ges-
tation with clinically significant abnormalities on US had
historical and clinical findings including placental abrupt-
ion, abnormal biophysical profile, dysmorphic facies,
seizures, severe hypotension, accelerated head growth,
and hydrocephalus that would have indicated the need for
prompt US evaluation. Our study supports the recom-
mendation by the Academy of Neurology and the Child
Neurology Society that screening US can be limited but
suggests that the gestational age cut off should be 30
weeks or less. With our revised screening US policy, we
21
have achieved an 80% reduction in the number of routine
screening cranial ultrasounds.
REFERENCES
1. Ment LR, Bada HS, Barnes P, et al. Practice parameter:
neuroimaging of the neonate. Report of the Quality Stand-
ards Subcommittee of the American Academy of Neurology
and the Practice Committee of the Child Neurology Society.
Neurology 2002;58:1726–1738
2. Fanaroff AA, Martin RJ. Intracranial hemorrhage. In:
Papile, L-A, eds. Neonatal-Perinatal Medicine. Diseases
of the Fetus and Infant. St. Louis, MO: Mosby; 2002:879–
881
3. Rezaie P, Dean A. Periventricular leukomalacia, inflamma-
tion and white matter lesions within the developing nervous
system. Neuropathology 2002;22:106–132
4. Volpe JJ. Brain injury in the premature infant-from patho-
genesis to prevention. Brain Dev 1997;19:519–534
5. Ment LR, Vohr B, Allan W, et al. The etiology and outcome
of the cerebral ventriculomegaly at term in very low birth
weight preterm infants. Pediatrics 1999;104:243–248
6. Papile LS, Burstein J, Burstein R, Koffler H. Incidence and
evolution of the subependymal intraventricular hemorrhage: a
study of infants with weights less than 1500 grams. J Pediatr
1978;92:529–534
7. Boal DK, Watterberg KL, Miles S, Gifford KL. Optimal
cost-effective timing of cranial ultrasound screening in
low-birth weight infants. Pediatr Radiol 1995;25:425–
428
8. Perlman J, Rollins N. Surveillance protocol for the detection of
intracranial abnormalities in premature neonates. Arch Pediatr
Adolesc Med 2000;154:822–826
9. Batton D, Holtrop P, DeWitte D, Pryce C, Roberts C.
Current gestational age-related incidence of major intra-
ventricular hemorrhage. J Pediatr 1994;125:623–625
10. Harding D, Kuschel C, Evans N. Should preterm infants
born after 29 weeks’ gestation be screened for intraventricular
hemorrhage? J Paediatr Child Health 1998;34:57–59
11. Behrman R, Kliegman R, Jenson H. The nervous system. In:
Haslam R, ed. Nelson Textbook of Pediatrics. Philadelphia,
PA: WB Saunders; 2000:1795
12. Holzman C, Paneth N, Little R, Pinto-Martin J, Neonatal
Brain Hemorrhage Study Team. Perinatal brain injury in
premature infants born to mothers using alcohol in
pregnancy. Pediatrics 1995;95:66–72