Childs Nerv Syst
DOI 10.1007/s00381-017-3544-5
ORIGINAL PAPER
Conversion of external ventricular
drainage to ventriculo-peritoneal shunt: to change or not
to change the proximal catheter?
Jehuda Soleman 1 & Haggai Benvenisti 1 & Shlomi Constantini 1 & Jonathan Roth 1
Received: 19 June 2017 / Accepted: 12 July 2017
# Springer-Verlag GmbH Germany 2017
Abstract
Purpose In this study, we investigate the occurrence rate of
early shunt infection and malfunction in pediatric patients af-
ter converting an external ventricular drainage (EVD) to a
ventriculo-peritoneal shunt (VPS) without replacing the ven-
tricular catheter.
Methods Data was retrospectively reviewed for 17 pediatric
patients (11 male (64.7%), mean age 7.5 years, range 0.25–
15 years) who underwent 18 consecutive direct conversions of
tunneled EVD to VPS without replacing the ventricular cath-
eter between 2008 and 2017. In each case, the EVD was
inserted in sterile fashion within the operating room and
tunneled subcutaneously 5–7 cm away from the insertion site.
Primary outcome measure was the occurrence of early (within
30 days) VPS infection or malfunction. The mean follow-up
time was 56.8 months (±35.7 months).
Results The mean period of EVD before VPS placement was
9.0 days (±3.6 days, range 2–18 days). Five patients had shunt
infections/malfunctions. One patient (5.6%) had an early
shunt infection after 30 days. One patient had a late shunt
infection after 9 months. One patient had an early shunt mal-
function after 9 days. Two patients (11.1%) had late shunt
malfunctions after 6.5 months and 9 years. There were no
other incidents of shunt-related complications or shunt-
related mortality.
Conclusion In the pediatric population, the conversion of a
tunneled EVD to a VPS without replacing the ventricular
catheter can be safely done. Cranial entry is spared, while
* Jonathan Roth
jonaroth@gmail.com
1
Department of Pediatric Neurosurgery, Dana Children’s Hospital,
Tel-Aviv Medical Center, 6 Weizmann Street, Tel Aviv 64239, Israel
the rates of shunt infection and malfunction do not increase
significantly.
Keywords Pediatric neurosurgery . Hydrocephalus . Shunt
infection . Shunt malfunction
Introduction
External ventricular drainages (EVD) are used for a vari-
ety of pathologies causing acute hydrocephalus, or fol-
lowing ventricular surgery, and are therefore one of the
most common neurosurgical procedures [1, 2]. Some of
these patients will develop chronic hydrocephalus and
their EVD will need to be converted to a permanent
ventriculo-peritoneal shunt (VPS).
There are no clear guidelines regarding the conversion
of an EVD to a VPS, although this too is a frequent
neurosurgical procedure. Some neurosurgical centers ad-
vocate use of a new site for the VPS placement, since
contamination of the EVD site may increase the risk of
VPS infection [3, 4]. Others prefer replacing the existing
EVD catheter with a new proximal catheter, while using
the same EVD site, since it provides an atraumatic en-
trance of the proximal catheter through a preexisting track
and therefore minimizes the risk of hemorrhage [3].
It has been our practice to convert tunneled EVDs to
VPS by using the same EVD site without replacing the
existing EVD catheter. This option avoids brain parenchy-
mal passes, which might cause iatrogenic bleeds or cath-
eter malpositioning. Our objective in this study is to ana-
lyze the correlation of this practice to rates of shunt in-
fection and malfunction.

Methods
The study protocol was approved by the local ethics commit-
tee. Data obtained from 17 consecutive pediatric patients who
underwent 18 direct conversions of an EVD to a VPS for
hydrocephalus due to various reasons between 2008 and
2017 were retrospectively reviewed (Table 1). In each case,
the EVD was placed in the operating room in sterile fashion
and tunneled subcutaneously 5–7 cm away from the insertion
site. Patients were excluded from the study if the EVD was
tunneled to the abdomen and externalized there, if during the
VPS insertion the EVD catheter was replaced, or if patients
were older than 18 years of age. Follow-up data was available
for all patients, with a mean follow-up of 56.8 months
(±35.7 months, range 2–102.6 months). Primary outcome
measure was the occurrence of early (within 30 days) VPS
infection or malfunction. Secondary outcome measures were
the occurrence of late VPS infection or malfunction, other
surgical complications, and mortality.
Baseline characteristics of the patients included the follow-
ing: age, gender, underlying diagnosis, duration of EVD prior
to conversion to VPS, whether the EVD was changed during
this time period, ventriculitis or meningitis caused by EVD,
and cerebrospinal fluid (CSF) values (leukocytes,
polymorphonucleocytes (PMN), red blood cells (RBC), glu-
cose, and protein) prior to EVD conversion to the VPS.
Analysis was completed of such factors as age, sex, duration
of EVD placement, EVD location, underlying diagnosis, and
CSF values which might influence VPS infection and
malfunction.
Surgical technique
In most of the cases, EVD was either inserted frontally or
occipitally through a small burr hole or through the cranioto-
my opening. In one patient, the EVD was inserted into the 4th
ventricle through an open craniotomy of the posterior fossa.
CSF was obtained intraoperatively for laboratory analysis.
The catheter was tunneled for a distance of 5–7 cm using the
tunneling device available in the EVD kit, connected to a
closed CSF collecting system, and dressed in sterile fashion.
Frontally placed EVDs were tunneled to the occipital region,
while occipitally placed EVDs were tunneled to the frontal, or
shoulder or scapula regions. All ventricular catheters used
were antibiotic impregnated (Bactiseal® EVD catheter,
DePuy Synthes®; rifampin and clindamycin). All patients re-
ceived prophylactic antibiotics as long as the EVD was in
place (perioperatively and for the first 24 h vancomycin and
rocephin, thereafter cefuroxime).
Following the acute phase, if patients remained dependent
on CSF drainage, conversion of the EVD to a VPS was con-
sidered. VPS placement was only performed if preoperative
CSF cultures obtained from the EVD were negative.
Childs Nerv Syst
All peritoneal catheters inserted during the conversion of
EVD to VPS were antibiotic-impregnated devices
(Bactiseal®, DePuy Synthes®; rifampin and clindamycin).
All patients received prophylactic antibiotics perioperatively
and for the first 24 h (vancomycin and rocephin). The previ-
ously used ventriculostomy site was re-opened in 13 patients
(72.2%) and a distal catheter was subcutaneously tunneled to
the abdomen. In five cases (27.8%), a new, more distal site
along the tunneled catheter was chosen. The EVD was sharply
divided, either 1–2 cm from its emergence from the skull or at
the new incision site, and the distal cut end of the EVD was
pulled out after closure of all wounds. Approximately, 5 ml of
CSF was obtained for analysis. A PS Medical valve
(Medtronic®) was connected to the proximal and distal cath-
eter. Once flow of CSF through the distal catheter was assured,
the catheter was implanted into the peritoneum. Early postop-
erative imaging was not performed routinely; rather, an MRI
was completed at a later stage, or as clinically indicated.
Statistical analysis
All statistical analyses were done using SPSS (Statistics
Version 24.0, IBM Corp, 2012). Contingency tests were done
using Fisher’s exact test. For non-parametric tests, the Mann-
Whitney U test was used. A p value of <0.05 was considered
significant.
Results
A total of 17 patients (11 male, 64.7%) with a mean age of
7.6 years (±4.9 years, range 0.25–15 years) were included in
the study. Patient demographics are presented in Tables 1 and
2. The mean period of EVD before VPS placement was
9.0 days (±3.6 days, range 2–18 days). In four patients
(22.2%), the EVD was changed during this period of time,
before being replaced with a VPS. There were no EVD-
related infections; however, four patients had CSF infection
prior to EVD placement (Table 1).
Overall shunt infection rate was 11.1% (n = 2).
& Early shunt infection occurred in one patient (5.6%) after
30 days. This patient presented at an early age with an
infected shunt (Escherichia coli and candida) 4 years after
shunt placement (the shunt was originally inserted due to a
trapped 4th ventricle in association with a Dandy Walker
malformation). The shunt was externalized, antibiotic and
antifungal medication was administered, and once cul-
tures were negatives, it was replaced by a new shunt sys-
tem. Two years later, the patient presented with suspected
abdominal infection. The shunt was externalized, and
once proven infected (methicillin sensitive S.
aureus (MSSA)), the shunt was removed and an
Childs Nerv Syst

Table 1 Baseline and outcome characteristics of 18 consecutive cases included

No. Age Sex Underlying diagnosis Indication EVD Duration of EVD changed Shunt Timing of Shunt Timing of
(years) for shunt location EVD (days) before shunt infection infection malfunction malfunction

1 9 f Papillary pineal tumor SD f 18 yb n n

2 11 m SEGA SD f 11 n n n
3 13 m Metastatic PAa IVH, TP o 7 yc n n
4 13 m Metastatic PAa Trapped pf 11 n n yf 9 days
4th
5 15 m Intraventricular PA Late HCP f 10 yd n n
and
IVH
6 2 m Infra ependymoma SD o 11 n n n
7 8 m Infra medulloblastoma SD o 12 n n n
8 5 m Metastatic GBM IVH, TP, f 2 n n n
aHCP
9 3 m IVH due to AVM SD f 12 n n n
10 13 f Crouzon, encepahlocele, SD f 6 n yi 9 months n
congenital HCP
11 3 m Brain abscess causing HCP Infec, SD o 10 n n n
12 4 f C-H PA, shunt infection Infec, SD f 6 n n n
13 14 m Metastatic aHCP, TP o 7 n n yg 6.5 months
oligodendroglioma
14 11 f SEGA SD f 6 ye n n
15 2 f Hypothalamic OPG SD f 6 n n yh 9 years
16 6 f IVH due to AVM SD f 12 n n n
17 0.25 f IVH of prematurity, shunt Infec, SD o 7 n n n
infection
18 4 m Dandy walker, trapped 4th Infec, SD f 8 n yj 30 days n
ventricle, shunt infection

m male, f female, SEGA subependymal giant cell astrocytoma, PA pilocytic astrocytoma, GBM glioblastoma multiforme, IVH intraventricular hemor-
rhage, AVM arterio-venous anomaly, HCP hydrocephalus, OPG optic pathway glioma, Infra infratentoriell, C-H chiasmatic hypothalamic, SD shunt
dependency after EVD insertion at primary operation (e.g., tumor resection, bleeding evacuation, treatment of infection, etc.), TP tumor progression
(causing IVH or shunt malfunction), Infec (shunt-) infection leading to EVD insertion, aHCP acute hydrocephalus, y yes, n no, f frontal, o occipital, pf
posterior fossa
a
Case 3 and 4 are the same patient
b
Clotting of first EVD 3 days after tumor resection therefore EVD replaced
c
After tumor resection, IVH occurred which was removed surgically and at surgery, EVD was changed
d
2 years after primary tumor resection surgery developed HCP. Endoscopical septum pellucidotomy performed and EVD placed. IVH the next day
clotted EVD therefore changed
e
5 days after tumor resection surgery dislocation of EVD therefore changed. 2 days later, edoscopical septum pellucidotomy done, EVD changed during
the procedure
f
Dislocation of ventricular catheter in the posterior fossa leading to distal revision
g
Shunt clotted due to tumor progression (metastatic, intraventricular) leading to distal revision
h
Trapped lateral ventricle 9 years after tumor resection, underwent endoscopic septum pellucidotomy and replacement and repositioning of ventricular
catheter through the pellucidotomy
i
Streptococcus pneumonia
j
Methicillin sensitive Staphylococcus aureus

EVD was placed. After obtaining sterile CSF cul-
tures, the EVD was converted after 8 days to a
ventriculo-artrial shunt (VAS). Thirty days later, the
patient presented with fever and neurological decline
caused by an MSSA shunt infection. The shunt was
externalized and the patient treated with antibiotics
until the CSF was sterile. A new VPS was then
implanted. There were no further infections in this
patient during subsequent follow-up.
& Late shunt infection occurred in one patient (5.6%) after
9 months. This patient suffered from Crouzon syndrome
and underwent a Lefort distraction at the age of 11 years
 Childs Nerv Syst

Table 2 Cumulative data of 18 consecutive cases with conversion of progression, a trapped 4th ventricle became evident. This
external ventricular drainage (EVD) to ventriculo-peritoneal (VP) shunt
was addressed by an infratentorial craniotomy, resection
Factor Value of the tonsils, opening of arachnoidal webs toward the
spinal subarachnoid space, and insertion of a lumbo-
Mean age in years (±SD, range) 7.6 (±4.9, 0.25–15) peritoneal tube from the 4th ventricle toward the spinal
Sex (male, %) 11 (64.7%) subarachnoid space, acting as a local stent. Since no clin-
Mean days of EVD (±SD, range) 9.0 (±3.6, 2–18) ical and radiological improvement was observed post-op-
EVD location (n, %) eratively, the patient was taken to surgery again 3 days
Frontal 11 (61.1) later, where an open redo surgery was done and an EVD
Occipital 6 (33.3) left in the 4th ventricle. The patient improved, becoming
Posterior fossa 1 (5.6) shunt dependent, and therefore the EVD was converted
Change of EVD (n, %) 4 (22.2) into a VPS after 11 days. Nine days later, a neurological
CSF values at EVD insertion (mean ± SD) decline was seen, while on imaging a recurrence of the
Leukocytes (per mm3) 26.8 (±35.8) trapped 4th ventricle was apparent. The patient was taken
PMN (%) 30.6 (±32.0) to surgery, where a dislocation of the ventricular catheter
RBC (per mm3) 2081.8 (±3778.9) from within the 4th ventricle was seen.
Glucose (mg/dl) 67.6 (±20.2) & Late shunt malfunction occurred in two patients (11.1%).
Protein (mg/dl) 178.2 (±178.1) In one patient, a proximal shunt malfunction occurred due
CSF values before VP shunt insertion (mean ± SD) to intraventricular tumor progression after 6.5 months.
Leukocytes (per mm3) 29.5 (±35.2) The second patient developed a trapped lateral ventricle
PMN (%) 35.9 (±38.3) 9 years after partially resecting a hypothalamic optic path-
RBC (per mm3) 3027.3 (±3785.5) way gliomas and VPS insertion. An endoscopic septum
Glucose (mg/dl) 61.2 (±26.0) pellucidotomy was done, and at the same time the ventric-
Protein (mg/dl) 131.4 (±135.3) ular catheter was exchanged and repositioned through the
CSF values at or after VP shunt insertion (mean ± SD) septum pellucidotomy.
Leukocytes(per mm3) 10.1 (±23.8)
PMN (%) 19.7 (±30.3) No other VPS-related surgical complications or mortality
RBC (per mm3) 1353.8 (±1947.6) occurred during the follow-up period.
Glucose (mg/dl) 56.2 (±16.3) Results of the CSF samples collected during EVD inser-
Protein (mg/dl) 92.9 (±114.6) tion, before VPS insertion, and during or after VPS insertion
(mean 4.6 ± 4 days, range 0–17 days after VPS insertion) are
PMN polymorphonucleocytes, RBC red blood cells, CSF cerebrospinal presented in Table 2. All analyzed factors (age, sex, duration
fluid of EVD placement, location of EVD, replacement of EVD,
underlying diagnosis, and CSF values) showed no significant
association with shunt infection or malfunction.
and an endoscopic trans-nasal repair of a basal
encephalocele and EVD placement at the age of 13 years.
The EVD was converted into a shunt after 6 days, since Discussion
the patient became EVD dependent. Nine months later, the
patient presented with a trans-nasal CSF rhinorrhea and In this study, we demonstrated that direct conversion of an
streptococcus pneumonia meningitis. The leak was treated EVD to a VPS in children is a valid and safe option. The
surgically, while the VP shunt was externalized and the conversion of an EVD to a VPS is a common neurosurgical
patient treated with antibiotics. Once the CSF cultures procedure; however, no data exists about the postoperative
showed no bacterial growth, the VPS was reinserted into infection rate if the EVD catheter is not replaced. To our
the abdomen. knowledge, this is the first study describing and analyzing this
practice in children, and even in the general population. The
Overall shunt malfunction rate was 16.7% (n = 3). VPS infection rate in our study, within a rather long follow-up
time (mean of approx. 4.5 years), is comparable to the pub-
& Early shunt malfunction occurred after 9 days in one pa- lished literature on the rates of pediatric VPS infection [5–7].
tient (5.6%). This patient suffered from a metastatic In our series, there were two infections, one early and one late.
pilocytic astrocytoma (PA) within the third ventricle and In one patient, an underlying infection before EVD insertion
underwent repeated supratentorial tumor surgeries and in- was apparent. In the other patient, the infection resulted from a
sertion of an occipital VPS. Later, due to tumor CSF rhinorrhea after surgical repair of a basal encephalocele.
Childs Nerv Syst
In four patients, the EVD was placed due to an underlying
CSF or VPS infection (patients 11, 12, 17, and 18 in
Table 1), and converted directly to a VPS once the infection
was treated and CSF cultures showed no bacterial growth.
Although in one of these patients, an early shunt infection
occurred, no significant correlation was seen between under-
lying CSF or VPS infection, at the time of EVD insertion, and
VPS infection after direct conversion of EVD to VPS.
However, great care should be taken when converting an
EVD directly to a VPS in situations where an underlying
CSF or VPS infection was apparent, since the risk of infection
might be increased due to a biofilm on the EVD. Further
studies with a larger population group are warranted, before
we can draw conclusions for this specific patient population.
Early proximal shunt malfunction occurred in one patient
after conversion of EVD to a VPS, after the EVD was initially
placed in the fourth ventricle due to trapped 4th ventricle. The
obstruction of the catheter recurred after 9 days, and therefore
a redo surgery was done, in which the ventricular catheter was
repositioned. Whether the early catheter dislocation was due
to the conversion surgery itself, or due to the fact that VPS
placed in the posterior fossa are more prone to dislocate, re-
mains unclear [8]. However, when disconnecting the ventric-
ular catheter from the EVD system and connecting it to the
VPS system, one should be attentive not to pull out the ven-
tricular catheter and to minimize the amount of manipulation
around the ventricular catheter. Late shunt malfunction oc-
curred in two patients. In one case, a proximal shunt malfunc-
tion occurred due to tumor progression of a metastatic intra-
ventricular tumor. In the second case, no shunt malfunction
was seen; however, since a septum pellucidotomy was done,
due to a trapped lateral ventricle, the ventricular catheter was
exchanged and repositioned, going from one lateral ventricle
to the other. Thus, it seems obvious that these events were not
related to the fact that the EVD catheter was not replaced
during the conversion to a VPS. Preoperative CSF values
drawn from our patients did not seem to influence the rate of
infection or malfunction. High protein and/or high RBC levels
in the CSF before VPS placement were shown to influence
shunt malfunction negatively in laboratory studies [9]. On the
other hand, a clinical study in patients undergoing VPS place-
ment after suffering an aneurysmal subarachnoid hemorrhage
suggested otherwise [3]. In our study, no correlation between
high protein and RBC levels in the CSF and higher shunt
malfunction was seen. A relative paucity in the literature exists
on the prognostic value of preoperative and postoperative CSF
values for VPS malfunction or infection. Based on the pre- or
postoperative CSF values of our small cohort, no correlation
between the CSF values and higher risk for VPS infection or
malfunction was seen.
Generally, when converting an EVD to a VPS, in many
neurosurgical centers, a clean, new site is used for the VPS
insertion. The rationale is to avoid higher infection rates,
which might be caused due to contamination of the EVD site.
Rammos et al. introduced a technique where the VPS was
placed through the same EVD site without risking higher rates
of shunt infection [3]. However, in their series, the ventricular
catheter was replaced, still exposing the patients to all of the
risks associated with ventricular catheter replacement (espe-
cially in relatively small or distorted ventricles) [3].
We focused specifically on various EVD-related factors
which we believe may increase the safety of direct conversion
of EVD to VPS:
& The EVD is placed in the operating room under strict
sterile conditions, and is tunneled at least 5 cm away from
the incision, since this was shown to reduce infection rates
[10, 11].
& The EVD is tunneled away from a potential future VPS
tract.
& We use only impregnated EVD and VPS tubing, since it
has been shown that the introduction of antibiotic-
impregnated ventricular catheters, and better care of the
EVD site, significantly reduces EVD-related infections [3,
4, 12–14]. In addition, a recent meta-analysis showed that
the rate of VPS infection is reduced when antibiotic-
impregnated tubing is used [15].
& Wounds are kept as small as possible. In addition, we
emphasize meticulous care of the local wound and the
EVD exit site.
& If the EVD was tunneled sufficiently distant from the
EVD incision, we choose a more distal site along the
tunneled catheter for the incision. In our cohort, infection
rate was comparable between the group where the same
EVD site was utilized, and the group where a new, more
distal incision was done (one infection each).
& Our patients are kept on prophylactic antibiotics as long as
they have an EVD in place and prophylactic antibiotic
treatment is administered during and for 24 h after the
conversion of EVD to VPS. Although controversial, a
prolonged course of prophylactic antibiotics throughout
the EVD’s lifespan was shown to reduce infection rates.
On the other hand, in the prolonged treatment group, the
organisms were more virulent and included MRSA and
Candida species [16, 17]. A survey among pediatric neu-
rosurgeons showed heterogeneous practice of antibiotic
prophylaxis for VPS placement. However, when EVD
was inserted, antibiotics were administered by all centers
for as long as the EVD was in place [18]. Antibiotic pro-
phylaxis during VPS surgery was significantly associated
with lower shunt infection rates [7].
Our pediatric series suggests that not only is it safe
to use the same EVD site for VPS conversion, it seems
safe to leave the old ventricular catheter in place and
not replace it. These results may reflect our EVD policy

as stated above, and may not be extrapolated to other
patient populations.
This retrospective study is subject to all the limitations of
data collection inherent in such work. Our study includes a
rather small sample size and might therefore be underpow-
ered. A control group for comparison was not apparent, since
this procedure has been our practice for the conversion of all
EVDs to VPS for the last 10 years. Further studies with larger
cohorts and control groups are warranted in order to further
evaluate and confirm our results.
Conclusion
This review of our experience supports the position that in
pediatric patients, direct conversion of an EVD to a VPS can
be safely done at the same EVD site and without exchanging
the ventricular catheter. Manipulation of the brain, with all the
associated risks, is avoided, and the rates of shunt infection
and malfunction do not seem to increase significantly.
Acknowledgements We would like to thank Mrs. Adina Sherer for the
medical proof reading of this manuscript.
Compliance with ethical standards
Conflict of interest None.
Disclosure of funding None.
References
1. Cinibulak Z, Aschoff A, Apedjinou A, Kaminsky J, Trost HA,
Krauss JK (2016) Current practice of external ventricular drainage:
a survey among neurosurgical departments in Germany. Acta
Neurochir 158:847–853
2. van de Beek D, Drake JM, Tunkel AR (2010) Nosocomial bacterial
meningitis. N Engl J Med 362:146–154
3. Rammos S, Klopfenstein J, Augspurger L, Wang H, Wagenbach A,
Poston J et al (2008) Conversion of external ventricular drains to
ventriculoperitoneal shunts after aneurysmal subarachnoid hemor-
rhage: effects of site and protein/red blood cell counts on shunt
infection and malfunction. J Neurosurg 109:1001–1004
4. Sloffer CA, Augspurger L, Wagenbach A, Lanzino G.
Antimicrobial-impregnated external ventricular catheters: does
Childs Nerv Syst
the very low infection rate observed in clinical trials apply to daily
clinical practice? Neurosurgery 2005;56:1041–1044; discussion −4
5. Raffa G, Marseglia L, Gitto E, Germano A (2015) Antibiotic-
impregnated catheters reduce ventriculoperitoneal shunt infection
rate in high-risk newborns and infants. Child’s Nerv Syst: ChNS:
Off J Int Soc Pediatr Neurosurg 31:1129–1138
6. Soleman J, Schneider CA, Pfeifle VA, Zimmermann P, Guzman R
(2017) Laparoscopic-assisted ventriculoperitoneal shunt placement
in children younger than the age of 1 year. World Neurosurg 99:
656–661
7. Xu H, Hu F, Hu H, Sun W, Jiao W, Li R et al (2016) Antibiotic
prophylaxis for shunt surgery of children: a systematic review.
Child’s Nerv Syst: ChNS: Off J Int Soc Pediatr Neurosurg 32:
253–258
8. Bokhari I, Rehman L, Hassan S, Hashim MS (2015) Dandy-Walker
malformation: a clinical and surgical outcome analysis. J Coll
Physicians Surg–Pak: JCPSP 25:431–433
9. Brydon HL, Bayston R, Hayward R, Harkness W (1996) The effect
of protein and blood cells on the flow-pressure characteristics of
shunts. Neurosurgery 38:498–504 discussion 5
10. Berger A, Weninger M, Reinprecht A, Haschke N, Kohlhauser C,
Pollak A (2000) Long-term experience with subcutaneously
tunneled external ventricular drainage in preterm infants. Child’s
Nerv Syst: ChNS: Off J Int Soc Pediatr Neurosurg 16:103–109
discussion 10
11. Friedman WA, Vries JK (1980) Percutaneous tunnel
ventriculostomy. Summary of 100 procedures. J Neurosurg 53:
662–665
12. Cui Z, Wang B, Zhong Z, Sun Y, Sun Q, Yang G et al (2015) Impact
of antibiotic- and silver-impregnated external ventricular drains on
the risk of infections: a systematic review and meta-analysis. Am J
Infect Control 43:e23–e32
13. Fried HI, Nathan BR, Rowe AS, Zabramski JM, Andaluz N,
Bhimraj A et al (2016) The insertion and management of external
ventricular drains: an evidence-based consensus statement : a state-
ment for healthcare professionals from the neurocritical care socie-
ty. Neurocrit Care 24:61–81
14. Root BK, Barrena BG, Mackenzie TA, Bauer DF (2016) Antibiotic
impregnated external ventricular drains: meta and cost analysis.
World Neurosurg 86:306–315
15. Konstantelias AA, Vardakas KZ, Polyzos KA, Tansarli GS, Falagas
ME (2015) Antimicrobial-impregnated and -coated shunt catheters
for prevention of infections in patients with hydrocephalus: a sys-
tematic review and meta-analysis. J Neurosurg 122:1096–1112
16. Dasic D, Hanna SJ, Bojanic S, Kerr RS (2006) External ventricular
drain infection: the effect of a strict protocol on infection rates and a
review of the literature. Br J Neurosurg 20:296–300
17. Poon WS, Ng S, Wai S (1998) CSF antibiotic prophylaxis for neu-
rosurgical patients with ventriculostomy: a randomised study. Acta
Neurochir Suppl 71:146–148
18. Biyani N, Grisaru-Soen G, Steinbok P, Sgouros S, Constantini S
(2006) Prophylactic antibiotics in pediatric shunt surgery. Child’s
Nerv Syst : ChNS : Off J Int Soc Pediatr Neurosurg 22:1465–1471