J Neurosurg Pediatrics 14:167172, 2014

AANS, 2014

Predictors of the need for cerebrospinal fluid diversion in

patients with myelomeningocele
Clinical article
Blake C. Phillips, M.D.,1 Michael Gelsomino, M.D.,1
Ambre L. Pownall, M.S.N., A.P.R.N., P.P.C.N.P.-B.C., 3 Eylem Ocal, M.D.,1,3
Horace J. Spencer, M.S., 2 Mark S. OBrien, M.D.,1,3
and Gregory W. Albert, M.D., M.P.H.1,3
Departments of 1Neurosurgery and 2Biostatistics, University of Arkansas for Medical Sciences; and
3
Division of Neurosurgery, Arkansas Childrens Hospital, Little Rock, Arkansas
Object. Many patients with myelomeningocele (MMC) develop hydrocephalus, and most will undergo CSF
diversion. The goal of this retrospective study was to determine whether there was a change in the shunt rate over
the 7 consecutive years of the study. The authors will also identify the criteria used to determine the need for shunt
placement.
Methods. During a 7-year period, 73 patients underwent MMC closure at Arkansas Childrens Hospital. The
shunt rate for each year was calculated. Clinical characteristics were evaluated, including apneic and bradycardic
spells, CSF leak, level of the MMC, head circumference, and rate of head growth. In addition, radiological images
were reviewed, and the frontooccipital horn ratio (FOHR), ventricular index (VI), and thalamooccipital distance
(TOD) were calculated. Comparisons were made between those patients who underwent shunt placement and those
who did not.
Results. One patient was excluded due to death in the perinatal period. Of the 72 remaining patients, 54 (75%)
underwent placement of a ventriculoperitoneal shunt. This rate did not change significantly over time. Between the
cohorts with and without a shunt there was no significant difference in age, sex, or race. There was no significant
difference in apneic episodes or bradycardic episodes. There was a statistically significant difference in fontanelle
characteristics, head circumference at birth, and rate of head growth. Patients who required CSF diversion had a mean
head growth of 0.32 cm/day compared with those who did not receive a shunt (0.13 cm/day; p < 0.05). All radiological parameters were found to be statistically significant.
Conclusions. In this study, several classic indicators of hydrocephalus in the neonate were not found to be significantly associated with the need for CSF diversion. Fontanelle characteristics, head circumference at birth, and
head growth velocity were associated with the need for shunt placement. Imaging information including the VI, TOD,
and FOHR are statistically significant measures to evaluate prior to placement of a ventriculoperitoneal shunt. The
optimal patient with MMC for CSF diversion will have full to tense fontanelle, increasing head circumference of
more than 3 mm/day, and radiological evidence of an elevated VI, TOD, and/or FOHR.
(http://thejns.org/doi/abs/10.3171/2014.4.PEDS13470)

Key Words myelomeningocele head circumference anterior fontanelle

spina bifida hydrocephalus ventriculoperitoneal shunt congenital

yelomeningocele (MMC) is one of the most

common developmental anomalies of the CNS.
Many of these patients also develop hydrocephalus, most of whom eventually undergo CSF diversion.
The rate of CSF shunt placement in these patients varies
significantly in published studies, from 52% to 92%, with

Abbreviations used in this paper: FOHR = frontooccipital horn

ratio; MMC = myelomeningocele; MOMS = Management of
Myelomeningocele Study; OFC = occipitofrontal circumference;
TOD = thalamooccipital distance; VI = ventricular index.

J Neurosurg: Pediatrics / Volume 14 / August 2014

at least 1 group reporting a decrease in shunt rate over

time at their institution.1,6
Cerebrospinal fluid shunting has numerous complications, most notably shunt failure and shunt infection.
Studies have suggested that patients with greater numbers
of shunt revisions have poorer performance on neuropsychological testing.7,8 Another study showed that shunt infection has an adverse effect on the patients intelligence
quotient.17 These data provide incentive to reduce the rate
of shunt placement in patients with MMC and to ensure
that only patients who will benefit from CSF shunting will
have this procedure performed.
167

B. C. Phillips et al.
In 10% of infants with MMC, hydrocephalus is apparent at birth, with patients having macrocrania and
massive ventriculomegaly.2 While these patients will
likely benefit from CSF shunting, there are no widely
applied criteria for CSF shunting in other patients with
MMC with less profound hydrocephalus. Previously reported indicators of the need for a shunt include level of
the lesion, clinical signs of elevated intracranial pressure
such as a tense or bulging fontanelle, bradycardia, sunsetting eyes, increasing head circumference, and increasing
ventricular size.6,14,16
The Management of Myelomeningocele Study
(MOMS) explored the shunt rate of patients with prenatal
versus postnatal MMC repair as one of its primary end
points. This study demonstrated a statistically significant
difference in CSF diversion, with the rate of shunting in
the fetal closure group (65%) lower than in the postnatal
closure group (92%).1 Another recent publication indicated a shunt rate of 52% from 1 institution using stringent
shunt criteria.6 This institution was not involved in the
MOMS trial and performed exclusively postnatal MMC
repair. Also, while the results of the MOMS trial are
compelling, the rate of shunt placement for the postnatal repair group was high compared with other published
results.6
The goal of this retrospective study was to determine
the shunt rate over 7 consecutive years for patients at our
institution and to determine whether there was a change
in the shunt rate over time. We also sought to identify the
criteria used to determine the requirement for shunting at
our institution and identify which factors were the most
predictive of the need for CSF shunting.

Methods
Study Population

Retrospective data were collected on all infants with

MMCs repaired at Arkansas Childrens Hospital from
January 1, 2005, to December 31, 2011. This cohort consisted of 73 total patients, with 1 exclusion due to death in
the immediate perinatal period from multiple congenital
anomalies and withdrawal of care. All MMC repairs were
performed postnatally. This study was approved by the
Institutional Review Board of the University of Arkansas
for Medical Sciences.

Clinical and Anatomical Measurements

The anatomical level of the lesion was determined

from the neurosurgical operative report. The anatomical
level for this study was defined as the most rostral level
of the bone defect and compared using a 2-sample t-test.
Data on apneic spells, bradycardic spells, and fontanelle
characteristics were gathered from the neonatal intensive
care daily notes and confirmed in nursing documentation. These data were compared using Fishers exact test.
Gestational age was determined from the obstetric report
on the admissions records. Occipitofrontal circumference
(OFC) measurements were included only if performed by
a member of the division of neurosurgery. Primary OFC
measurements at birth were compared between those

168

patients with and without shunts as well as the rate of

change in OFC measurement when applicable. Rate of
change was calculated using Time 0 as birth and Time
1 as OFC just prior to shunt placement (for those with
shunts) or OFC at 3-month follow-up (for those without
shunts). These results were compared using a 2-sample
t-test.
The frontooccipital horn ratio (FOHR) was obtained
using CT scans prior to CSF diversion in those receiving
shunts (48 of 54 patients) and the most recent follow-up
for those who did not require a shunt (15 of 18 patients).
The FOHR was determined on axial images by averaging the frontal and occipital horn widths and then dividing by the interparietal distance.12 Measurements made
on 3 consecutive slices were evaluated and the largest
measurement was used for each value. The FOHR was
measured in triplicate on 3 separate days by a single observer (B.C.P.). The average of the 3 measurements for
each patient was used in the final analysis. The FOHR
for patients with and without shunts was compared using
a 2-sample t-test, with comparisons made independent of
age.
The ventricular index (VI) and the thalamooccipital
distance (TOD) were obtained from cranial ultrasonography examinations when performed prior to placement of
a ventriculoperitoneal shunt in those receiving shunts (49
of 54 patients) and within the first month of life for those
who did not require a shunt (17 of 18 patients). The VI
was defined as the distance between the falx cerebri and
the lateral wall of the anterior horn at its widest point in
the coronal plane.4 The TOD was defined as the distance
between the outermost point of the thalamus at its junction with the choroid plexus and the outermost part of the
occipital horn in a parasagittal plane.3,4 The VI and TOD
were measured in triplicate on 3 consecutive days by a
single observer (B.C.P.). The average of the 3 measurements for each value for each patient was used in the final
analysis. These measurements for patients with and without shunts were compared using a 2-sample t-test with
comparisons made independent of age.
Complications

Complications were recorded after reviewing daily

progress and operative notes. Special attention was paid
to complications in regard to hydrocephalus, including
the development of a pseudomeningocele or CSF leak
from the operative site. Any event of death was recorded.
All shunt revisions and infections were recorded.

Statistical Analyses

The statistical significance of all clinical and radiological variables was determined using the univariate
Cox proportional hazards model. Multivariate Cox proportional hazards analysis was also performed. In this
analysis, 3 models were used. Model 1 included only
clinical characteristics (gestational age, sex, apneic episodes, bradycardic episodes, fontanelle description, head
circumference at birth, and head growth rate). Model 2
included the clinical characteristics from Model 1 in addition to the radiological parameters obtained from head
J Neurosurg: Pediatrics / Volume 14 / August 2014

Cerebrospinal fluid diversion in a myelomeningocele population

ultrasonography (VI and TOD). Model 3 included all
clinical and radiological variables.

Results

Of the 72 patients reviewed, 54 (75%) underwent

CSF diversion, with 81% of these patients requiring the
insertion of a ventriculoperitoneal shunt within the first
month. Annual shunt rates varied from 45% to 100%
(Fig. 1). There was not a statistically significant change in
the rate of patients with MMC undergoing CSF diversion
through the study period.
The anatomical level of the MMC was thoracic in
7 patients, lumbar in 64 patients, and sacral in 1 patient.
Seventy-five percent of the patients with thoracic-level
MMC had shunts, and 75.4% of the patients with lumbar
MMC had shunts. The single patient with a sacral MMC
did not have a shunt. There was no statistically significant
difference in anatomical level between patients with and
without shunts in this study.
Apneic and bradycardic spells were measured to determine their predictive value for CSF diversion. Of the
72 patients, apneic spells were recorded in 17 patients
(Table 1). Apneas were recorded in 11 (20.4%) of the 54
patients with shunts and 6 (33.3%) of the 18 patients without shunts. This difference was not statistically significant. Bradycardic episodes were recorded in 19 (35.2%)
of the 54 patients with shunts and 5 (27.8%) of the 18
patients without shunts (Table 1). This difference was also
not statistically significant.
A description of the fontanelle characteristics was
available for all 72 patients. Fontanelle description was
divided into sunken/flat, full/soft, and tense, and recorded
in the perinatal period or immediately prior to CSF di-

version. For the 18 patients without shunts, the fontanelle

was described as sunken/flat in 14 patients, full/soft in 3
patients, and tense in 1. For the 54 patients with shunts,
the fontanelle was described as sunken/flat in 13 patients,
full/soft in 33 patients, and tense in 8 patients (Table 1).
These values were found to be statistically significant,
with a fontanelle described as sunken/flat less likely to
require a shunt than those with a fontanelle described as
full/soft or tense (p < 0.05). Using sunken/flat as the referent, the calculated odds ratios for the full/soft and tense
groups were 11.9 (95% CI 2.9148.15) and 8.6 (95% CI
0.9478.67), respectively.
The OFC was available for all 72 patients. For patients who underwent shunt placement, the OFC percentile was determined using the last measurement recorded
before shunt placement. Among patients without shunts,
the OFC percentile was determined using the measurement taken at the last recorded follow-up with the patient, up to 3 months of age. The rate of CSF diversion
was analyzed to determine if there was a statistically
higher or lower shunt rate among OFC percentile ranges
(Table 1). The mean OFC was found to be 32.8 cm for
patients without shunts and 34.8 cm for patients with
shunts. These values were not statistically significant.
Forty-five of the 72 patients with recorded OFC had multiple recordings available for analysis of rate of change.
The mean rate of head growth was 0.13 cm/day and 0.32
cm/day for patients without and with shunts, respectively
(Table 1). This difference in the rates between patients
with and without shunts was statistically significant (p
< 0.05).
Radiological imaging was available for all 72 patients, with CT performed in 62 of 72 patients and head
ultrasonography performed in 68 of 72 patients. Com-

Fig. 1. Line graph displaying the percentage of patients with MMC who received shunts per year. There was not a statistically
significant change in shunt rate over the period evaluated.

J Neurosurg: Pediatrics / Volume 14 / August 2014

169

B. C. Phillips et al.
TABLE 1: Characteristics of hydrocephalus in a cohort with MMC
Variable

Without Shunts (n = 18)

With Shunts (n = 54)

mean gestational age SD (wks)

males (%)
race (%)
Caucasian
African American
Hispanic
other
apneic episodes (%)
bradycardic episodes (%)
CSF leak (%)
fontanelle description (%)
sunken/flat
full/soft
tense
mean OFC SD (cm)
mean head growth SD (cm/day)
mean FOHR SD
mean VI SD
mean TOD SD

36.3 3.3
11 (61.1)

37.5 2.3
29 (53.7)

14 (77.8)
1 (5.6)
2 (11.1)
1 (5.6)
6 (33.3)
5 (27.8)
0

42 (77.8)
5 (9.3)
6 (11.1)
1 (1.9)
11 (20.4)
19 (35.2)
4 (7.4)

14 (77.8)
3 (16.7)
1 (5.6)
32.8 3.8
0.13 0.077
0.446 0.084
1.77 0.36
2.35 0.97

13 (24.1)
33 (61.1)
8 (14.8)
34.8 3.5
0.316 0.264
0.549 0.089
2.33 0.83
4.04 1.61

p Value
0.1850*
0.7848
0.8270

0.3379
0.7737
0.5660
0.0003

0.0611*
0.0126*
0.0004*
0.0004*
<0.0001*

* Two-sample t-test with Satterthwaite adjustment for unequal variances.

Fishers exact test.
Forty-six patients with shunts, 15 without shunts.
Forty-nine patients with shunts, 17 without shunts.

puted tomography was used to determine the FOHR. The

mean FOHR for patients without and with shunts was
0.45 and 0.55, respectively (Table 1). Head ultrasonography was used to determine the VI and TOD. The mean VI
for patients without and with shunts was 1.77 cm and 2.33
cm, respectively (Table 1). The mean TOD for patients
without and with shunts was 2.35 cm and 4.04 cm, respectively (Table 1). All radiological measurements were
compared using a 2-sample t-test and were found to be
statistically significant (p < 0.05).
Univariate Cox proportional hazards analysis demonstrated statistical significance for fontanelle description, head circumference at birth, rate of head growth,
and all radiological parameters (Table 2). Multivariable
Cox proportional hazards analysis was performed using
3 models (Table 3). Considering only clinical characteristics and excluding the radiological parameters, fontanelle
description and head circumference at birth were significant predictors of the need for shunt placement. Considering these clinical characteristics and radiological parameters from head ultrasonography, fontanelle description
and TOD were significant. Considering clinical characteristics and all radiological parameters, head growth rate
and FOHR were significant.
Of the 72 MMCs closed, 4 had CSF leaks documented at the MMC repair site, which is consistent with
previous studies.11 All 4 of these patients eventually required CSF diversion (Table 1). During the study period,
26 (48%) of the 54 patients who underwent CSF diversion
170

required a revision. Seven (27%) of these revisions were

due to shunt infection. During the reporting period, the
mortality rate was 6.8% (5/73) prior to exclusions, similar
to previously reported rates.14 Two of the deceased were
in the cohort without shunts, but after thorough chart review, neither of these deaths was due to hydrocephalus:
one was from alleged child neglect, and the other was due
to necrotizing enterocolitis. The death of the only excluded patient was due to multiple congenital anomalies requiring withdrawal of care. Of the 2 patients who died in
the shunt group, one death was from cardiac arrest after
septic shock, and the other from an indeterminate cause.

Discussion

The overall shunt rate of our patient population was

75%, which is lower than the more commonly reported
range of 80% to 85% in the MMC population.1,14 This rate
is also lower than the 92% shunt rate for infants in the
MOMS trial who underwent postnatal closure.1 The shunt
rate by anatomical level was not statistically significant
as has been reported in other studies, likely due to small
sample size.14 The rate of shunt placement for patients
with thoracic defects was 75%, with reported rates in excess of 90%.14 The rate of shunt placement for patients
with a lumbar defect in this study was 75.4%, which is
similar to reported rates in other studies.14 Lower shunt
rates for lumbar defects were reported as well for patients
with both prenatal and postnatal repair.5,9,13 No significant
J Neurosurg: Pediatrics / Volume 14 / August 2014

Cerebrospinal fluid diversion in a myelomeningocele population

TABLE 2: Results of univariate Cox proportional hazards models
examining the association between time to shunt placement and
each variable*
Variable

No.
Missing

gestational age (1-wk )

female
Caucasian
apneic episodes
bradycardic episodes
soft/full/tense fontanelle
head circumference (1-cm )
rate of head growth (0.01cm/day )
FOHR (0.2-unit )
VI (1-unit )
TOD (1-unit )

HR (95% CI)

p Value

0
0
0
0
0
0
0
12

1.13 (0.991.28)
0.0673
1.16 (0.681.99)
0.5813
1.14 (0.602.17)
0.6890
1.33 (0.622.57) 0.4068
1.06 (0.611.86)
0.8342
3.23 (1.716.09)
0.0003
1.11 (1.031.20)
0.0042
1.28 (1.161.40) <0.0001

11
6
6

1.17 (1.081.25)
1.77 (1.312.39)
1.48 (1.251.74)

<0.0001
0.0002
<0.0001

* = change; HR = hazard ratio.

trend was observed in changes in shunt rate over time,

likely due to the small sample size and relatively short
time period of the study.
This study found a statistically significant correlation between the fontanelle description and eventual need
of CSF diversion. Patients with a fontanelle described as
full/soft or tense had a statistically higher shunt rate compared with patients with a fontanelle described as sunken/
flat. While in this study none of the surgeons used the
fontanelle description as the sole indication for placement
of a shunt, it appears that independent of other signs/
symptoms or radiological data, it was the most useful
clinical indicator for need of CSF diversion. This simple
noninvasive test should be objectively quantified and used
in pediatric neurosurgical practice for evaluating hydrocephalus and detecting elevated intracranial pressure in
the neonate.
Apneic and bradycardic spells have long been associated with hydrocephalus in the neonate. Unfortunately,
both apnea and bradycardia are associated with a myriad
of neonatal morbidities, ranging from gastroesophageal
reflux to cardiac anomalies. Patients with MMC are often
afflicted with many of these anomalies. These 2 objective
signs of hydrocephalus were analyzed in this MMC cohort and neither had a statistically significant association
with the eventual shunt status of the patients. Therefore,
although it may be reasonable to consider further testing
for hydrocephalus in a patient with apneic or bradycardic
spells, these signs do not have a significant correlation
with eventual need for CSF diversion.
Increasing OFC is often noted in infants with hydrocephalus. Patients requiring CSF diversion did have a
rate of head growth almost 2.5 times greater than those
patients who did not undergo placement of a shunt. The
OFC is a noninvasive measurement that can be useful
in the assessment of hydrocephalus, but at a single time
point it was not found to be predictive for CSF shuntJ Neurosurg: Pediatrics / Volume 14 / August 2014

TABLE 3: Results of multivariate Cox proportional hazards

models
Variable*
Model 1 (n = 72)
soft/full/tense fontanelle
head circumference (1-cm )
Model 2 (n = 66)
soft/full/tense fontanelle
TOD (1-unit )
Model 3 (n = 48)
rate of head growth (0.01-cm/day )
FOHR (0.2-unit )

HR (95% CI)
3.23 (1.706.12)
1.11 (1.031.20)
2.26 (1.264.56)
1.37 (1.151,63)
1.22 (1.091.37)
1.12 (1.041.22)

p Value
<0.0001
0.0003
0.0042
<0.001
0.0219
0.0005
<0.0001
0.0009
0.0047

* In Model 1, the predictor pool included patient characteristic variables that had no missing values (see Table 2); only fontanelle type and
head circumference remained in the model after a stepwise variable
selection procedure. In Model 2, the predictor pool included the same
set from Model 1 as well as the radiological measures VI and TOD. In
Model 3, all variables were included in the predictor set.

ing. The initial OFC was not predictive of the need for
CSF diversion, consistent with the widely held belief that
hydrocephalus worsens after MMC closure. Rather, we
found that a more valuable use for OFC is to follow the
measurement of individual patients over an extended period of time. In children who cross percentiles for OFC
measurements, this may be indicative of the progression of hydrocephalus and the subsequent need for shunt
placement. Additionally, in this study we found that patients whose head circumference increased by more than
3 mm per day were more likely to require CSF diversion.
The FOHR makes an accurate estimation of the ventricle size using CT. The FOHR does not change with
age in normal patients.10 In this study, we found a statistically significant difference in this parameter between
patients with and without shunts. As an indirect assessment of ventricular volume, this finding may indicate that
the degree of ventriculomegaly in patients with MMC is
a significant factor in the requirement for CSF shunting
regardless of other signs and symptoms that a patient may
have. Although bias may exist from selective imaging of
patients with suspected hydrocephalus, the FOHR can be
a useful factor as an objective estimate of ventricular size
when deciding if a patient requires a shunt. Almost all of
the MMC patients in this study underwent head ultrasonography as part of our admission protocol. It was also
interesting to find that nearly all of the patients also had
at least 1 CT scan. While there was statistical significance
for both imaging modalities, it appears that head ultrasonography for VI and TOD measurements was as good if
not better than CT for predicting the need for CSF diversion. Therefore, because of the radiation exposure associated with CT, it may be advisable to reserve CT for those
patients with complex intracranial anatomy prior to shunt
placement who cannot undergo MRI.
Many of the variables collected in this study are subjective in nature. Thus, there may be bias when collecting
171

B. C. Phillips et al.
this information. Additionally, the measured head circumference was performed by an array of neurosurgical
staff, and while all use the same method of measurement,
it may be subject to some nuances. Experienced neurosurgical caregivers recognize that there can be significant
differences of OFC measurements between examiners.15
In the end, determining which patients will benefit from
a shunt is based on the surgeons gestalt and is somewhat subjective. Although we use placement of a shunt
as a marker for need of a shunt, the two are likely not the
same. Future studies may include a blinded prospective
analysis of CSF diversion to help eliminate the previously
described biases.

Conclusions

Creating a synopsis of variables that facilitate targeting patients with MMC needing CSF diversion will
aid both the consulting physician and treating specialist.
Eliminating unnecessary CSF diversion and treating only
those who require CSF diversion is the ultimate goal.
While there are a number of signs and symptoms that
have long been associated with hydrocephalus in the neonate, this study has shown that several of them were not
significantly associated with the need for CSF diversion,
particularly apneic and bradycardic episodes. Fontanelle
description and the rate of change of the OFC appear to
be the most important clinical signs when determining
the need for shunt. Additional information collected by
head ultrasonography and CT, including the VI, TOD, and
FOHR, are statistically significant measures to evaluate
prior to placement of a ventriculoperitoneal shunt. The
optimal patient with MMC for CSF diversion will have
a soft/full to tense fontanelle, increasing head circumference of more than 3 mm/day, and radiological evidence of
elevated VI, TOD, and/or FOHR as described.
Disclosure
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this
paper.
Author contributions to the study and manuscript preparation include the following. Conception and design: Ocal, OBrien,
Albert. Acquisition of data: Phillips, Gelsomino, Pownall, Albert.
Anal
ysis and interpretation of data: Phillips, Pownall, Spencer,
Albert. Drafting the article: Phillips, Gelsomino. Critically revising
the article: all authors. Reviewed submitted version of manuscript: all
authors. Approved the final version of the manuscript on behalf of all
authors: Phillips. Statistical analysis: Spencer. Administrative/technical/material support: OBrien, Albert. Study supervision: Albert.
References
1. Adzick NS, Thom EA, Spong CY, Brock JW III, Burrows PK,
Johnson MP, et al: A randomized trial of prenatal versus postnatal repair of myelomeningocele. N Engl J Med 364:993
1004, 2011
2. Bowman RM, McLone DG: Neurosurgical management of
spina bifida: research issues. Dev Disabil Res Rev 16:8287,
2010

172

3. Brouwer MJ, de Vries LS, Groenendaal F, Koopman C, Pistorius LR, Mulder EJ, et al: New reference values for the neonatal cerebral ventricles. Radiology 262:224233, 2012
4. Brouwer MJ, de Vries LS, Pistorius L, Rademaker KJ, Groen
endaal F, Benders MJ: Ultrasound measurements of the lateral
ventricles in neonates: why, how and when? A systematic review. Acta Paediatr 99:12981306, 2010
5. Bruner JP, Tulipan N, Reed G, Davis GH, Bennett K, Luker
KS, et al: Intrauterine repair of spina bifida: preoperative predictors of shunt-dependent hydrocephalus. Am J Obstet Gynecol 190:13051312, 2004
6. Chakraborty A, Crimmins D, Hayward R, Thompson D: Toward reducing shunt placement rates in patients with myelomeningocele. J Neurosurg Pediatr 1:361365, 2008
7. Dennis M, Jewell D, Drake J, Misakyan T, Spiegler B, Hetherington R, et al: Prospective, declarative, and nondeclarative
memory in young adults with spina bifida. J Int Neuropsychol Soc 13:312323, 2007
8. Hetherington R, Dennis M, Barnes M, Drake J, Gentili F:
Functional outcome in young adults with spina bifida and hydrocephalus. Childs Nerv Syst 22:117124, 2006
9. Johnson MP, Sutton LN, Rintoul N, Crombleholme TM, Flake
AW, Howell LJ, et al: Fetal myelomeningocele repair: shortterm clinical outcomes. Am J Obstet Gynecol 189:482487,
2003
10. Kan P, Walker ML, Drake JM, Kestle JR: Predicting slitlike
ventricles in children on the basis of baseline characteristics
at the time of shunt insertion. J Neurosurg 106 (5 Suppl):
347349, 2007
11. Mslman AM, Karda S, Sucu D, Akal A, Ylmaz A,
Sirinolu D, et al: Clinical outcomes of myelomeningocele
defect closure over 10 years. J Clin Neurosci 19:984990,
2012
12. OHayon BB, Drake JM, Ossip MG, Tuli S, Clarke M: Frontal
and occipital horn ratio: a linear estimate of ventricular size
for multiple imaging modalities in pediatric hydrocephalus.
Pediatr Neurosurg 29:245249, 1998
13. Radmanesh F, Nejat F, El Khashab M, Ghodsi SM, Ardebili
HE: Shunt complications in children with myelomeningocele:
effect of timing of shunt placement. Clinical article. J Neurosurg Pediatr 3:516520, 2009
14. Rintoul NE, Sutton LN, Hubbard AM, Cohen B, Melchionni
J, Pasquariello PS, et al: A new look at myelomeningoceles:
functional level, vertebral level, shunting, and the implications
for fetal intervention. Pediatrics 109:409413, 2002
15. Sutter K, Engstrom JL, Johnson TS, Kavanaugh K, Ifft DL:
Reliability of head circumference measurements in preterm
infants. Pediatr Nurs 23:485490, 1997
16. Wakhlu A, Ansari NA: The prediction of postoperative hydrocephalus in patients with spina bifida. Childs Nerv Syst 20:
104106, 2004
17. Wills KE, Holmbeck GN, Dillon K, McLone DG: Intelligence
and achievement in children with myelomeningocele. J Pediatr Psychol 15:161176, 1990

Manuscript submitted September 16, 2013.

Accepted April 17, 2014.
Please include this information when citing this paper: published online May 30, 2014; DOI: 10.3171/2014.4.PEDS13470.
Address correspondence to: Blake C. Phillips, M.D., University
of Arkansas for Medical Sciences, Department of Neurosurgery,
4301 West Markham St., Slot 507, Little Rock, AR 72205. email:
phillipsblakec@uams.edu.

J Neurosurg: Pediatrics / Volume 14 / August 2014