Journal of Clinical Neuroscience 18 (2011) 1201–1205

Contents lists available at ScienceDirect

Journal of Clinical Neuroscience

journal homepage: www.elsevier.com/locate/jocn

Clinical Study

Marked reduction in wound complication rates following decompressive

hemicraniectomy with an improved operative closure technique
Michael E. Sughrue, Orin G. Bloch, Geoffrey T. Manley, Shirley I. Stiver ⇑
Department of Neurological Surgery, Brain and Spinal Injury Center, University of California at San Francisco, 1001 Potrero Avenue, Building 1, Room 101,
Box 0899, San Francisco, CA 94143, USA

a r t i c l e i n f o a b s t r a c t

Article history: Although decompressive hemicraniectomy with dural expansion and bone flap removal is a potentially
Received 19 December 2010 life-saving procedure, concerns remain regarding the morbidity associated with this approach. We and
Accepted 7 January 2011 others have noted the high rate of wound complications resulting from this technique, often associated
with cerebrospinal fluid (CSF) absorption problems. Here, we present our experience with an improved
technique for wound closure after unilateral decompressive hemicraniectomy with a wide cruciate dur-
Keywords: otomy. Data for all patients who underwent a decompressive hemicraniectomy at our institution from
Complication
October 2005 to October 2009 were gathered prospectively. Starting in mid 2008, we adopted an alter-
Hemicraniectomy
Infection
nate approach to operative wound closure, which involved skin closure with a running Monocryl absorb-
Wound able stitch, and prolonged subgaleal drainage. We compared the rates of wound complication using this
approach with those obtained with earlier conventional closure techniques. Over a 1 year period, we
dramatically reduced the rate of wound complications in patients undergoing hemicraniectomy at our
hospital using this new (Monocryl technique, 0% (n = 29) compared to other techniques, 35% (n = 98),
chi-squared [v2] p < 0.001). Patients closed using our new technique experienced markedly reduced rates
of wound infection (p < 0.01), and CSF leak (p < 0.05), compared to other, more standard, techniques.
Thus, attention to closure of hemicraniectomy wounds can markedly reduce the rate of wound compli-
cations, thus improving the risk-to-benefit ratio of this procedure.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction the ability of retrospective analyses to capture minor complications,

Decompressive hemicraniectomy with dural expansion and
bone flap removal is a potentially life-saving procedure for patients
with a variety of acute brain injuries, such as traumatic brain injury,
spontaneous intracerebral hemorrhage, and malignant middle cere-
bral artery infarction.1–5 One common objection to the upfront, pro-
phylactic approach to hemicraniectomy is the morbidity associated
with leaving the dura open and the bone flap off.6,7 Craniectomy
procedures have been noted to have high rates of neurosurgical
complications such as wound dehiscence, cerebrospinal fluid (CSF)
leak, and wound infection exceeding those of routine, elective
supratentorial craniotomies in which the dura is typically closed
and the bone flap replaced. In addition craniectomy mandates a sec-
ond surgery for subsequent replacement of the bone flap. In the lit-
erature, the rates of wound complication for craniectomy vary
markedly, from 3% to 40%.4,5,7,8 Potential explanations for this wide
variability include differences in technique (that is, the use of dura-
plasty compared to durotomy techniques), the aggressive use of CSF
diversion in up to 40% of patients in some series, and limitations in
such as a CSF leak requiring bedside repair or a wound infection
treated with antibiotics.
Recently, we have developed a protocol for operative closure of
craniectomy wounds based on the premise that the scalp is gener-
ally a well-vascularized structure, and if well approximated, it will
heal. Extra-axial CSF hygromas commonly develop beneath the
scalp defect following decompressive craniectomy with an open
durotomy. Given a ‘‘head-start’’, rigorous wound approximation
and early scalp healing prophylaxes against CSF leak that may arise
from tension put on the healing incision by these hygromas. We
present here our experience with a Monocryl (poliglecarone 25;
Ethicon J & J, Somerville, NJ, USA) technique for closure of hemicra-
niectomy incisions. Using this new approach for operative closure,
we have observed marked reduction in our complication rates for
these large cranial wounds.
2. Methods
2.1. Patient population

⇑ Corresponding author. Tel.: +1 415 206 5198. Data for all patients admitted to, or evaluated by, the Neurosur-
E-mail address: sstiver@neurosurg.ucsf.edu (S.I. Stiver). gery service at the San Francisco General Hospital (SFGH) from

0967-5868/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jocn.2011.01.016
1202 M.E. Sughrue et al. / Journal of Clinical Neuroscience 18 (2011) 1201–1205
October 2005 to October 2009 were gathered prospectively and
stored in a database. All patients receiving a unilateral, frontotem-
poral–parietal–occipital decompressive hemicraniectomy with
wide cruciate durotomy for traumatic brain injury, or intracerebral
hemorrhage, were evaluated for inclusion into this study. We sel-
dom perform bifrontal decompressive craniectomy, and these pa-
tients are not included in this study. This study was conducted
with the approval of the University of California San Francisco
Committee for Human Research.
Because our main goal was to study the rates of wound compli-
cations, we excluded all patients who died within 5 days of sur-
gery. There were no excluded patients who experienced wound
complications. We also excluded any patient with a degloving
scalp injury that required a tissue coverage procedure, such as a
rotational flap, for closure. Patients undergoing hemicraniectomy
with open or depressed skull fractures and scalp lacerations that
could be debrided and primarily closed were included in this ser-
ies. Hemicraniectomies for patients with penetrating injuries were
included if the entry and exit wounds could be debrided and closed
primarily.
2.2. Decompressive Hemicraniectomy
A large, reverse question mark incision was made from the mid-
line posterior to the hairline curving over the parietal–occipital re-
gion, reflecting forward above the pinna, and extending down
anterior to the tragus. The ipsilateral cerebral hemisphere was sur-
gically decompressed through a large hemicraniectomy, typically
measuring at least 12 cm by 15 cm. Additional bone was removed
from the sphenoid wing and low in the temporal region above the
zygoma to decompress the basal cisterns and upper brainstem. The
dura was opened slowly in a cruciate fashion to avoid hypotensive
arrest. Extra-axial hematomas were evacuated and hemostasis was
obtained using standard techniques. In all patients, Gelfilm (Pfizer,
New York, NY, USA) was placed in the subdural space on the corti-
cal surface to maintain adequate tissue planes for future cranio-
plasty. The dura was not closed and the dural leaflets were laid
over the top of the Gelfilm. One or two Jackson Pratt (JP) drains
were left in the subgaleal space in all patients, regardless of the
method of skin closure. In all patients the galea layer was closed
with an inverted, interrupted 2-0 or 3-0 vicryl suture. Prior to July
2008, skin closure was at the discretion of the surgeon and meth-
ods utilized included staples 0.5 cm to 1 cm apart, running 3-0 ny-
lon with sutures typically separated by 1 cm, and interrupted 3-0
nylon vertical mattress sutures separated by 1 cm. Beginning in
July 2008, we initiated a protocol (see section 2.3) using Monocryl
suture for scalp wound closure that was followed by all surgeons at
SFGH for decompressive hemicraniectomy procedures. Following
closure of the craniectomy wound, intracranial pressure monitors
(ventriculostomy when this was possible) and brain tissue oxygen
monitors were placed routinely either in the operating room (OR),
or in the Intensive Care Unit (ICU). Post-operatively, we seldom
used continuous CSF drainage in patients with decompressive cra-
niectomy. Post-operative care also included formal nutritional
evaluation for all patients while in the ICU, and early initiation of
enteral feeding in all patients otherwise unable to take an oral diet.
2.3. SFGH wound closure protocol for decompressive hemicraniectomy
Following the placement of Gelfilm on the cortical surface with
overlying of the dural leaflets, two size 7 JP drains are introduced
into the subgaleal space and tunneled approximately 2 cm from
the wound edge. These drains are sutured into place with a 3-0 ny-
lon stitch, with the goal of tightening the skin around the drain so
that no leakage occurs. Inverted, interrupted 2-0 or 3-0 vicryl ga-
leal stitches are placed about 1 cm apart from each other. The skin
is then closed with a running 4-0 Monocryl stitch. The goal with
the Monocryl skin stitch is to take numerous small bites that are
closely spaced together (within 3–4 mm), thereby enabling metic-
ulous approximation of the two sides of the incision. Note that be-
cause this is an absorbable stitch, the bites should be very small
because the stitch does not need to be removed, and generally will
fall out on its own in 3 to 4 weeks. If open skull fractures or pene-
trating head injuries are near or in-line with the incision, theses are
debrided and closed primarily with the same galeal closure and 4-0
Monocryl running stitch. The wound is covered in a thin layer of
bacitracin, and an overlying sterile dressing is applied.
The craniectomy dressing is removed on post-operative day 2,
and the wound is left open to air. The wound is cleaned with saline
and dressed with bacitracin for 5 days, starting on post-operative
day 3. Typically, JP drains are removed on days 3 to 5 but may re-
main for as long as 8 days post-operatively. Upon removal, the
drain site is closed with a figure of eight, 3-0 nylon stitch. The mon-
ocryl stitches are left in and resorb on their own.
2.4. Data analysis
We defined an overall rate of wound complications using two
approaches. In the standard approach, we defined wound compli-
cations as any wound problem requiring a return to the operating
room or placement of a ventriculostomy. After determining that
the Monocryl cohort did not have any wound complications using
these criteria, we defined a second rate of wound complications
using more strict criteria. In these strict criteria, wound complica-
tions were defined as any evidence of CSF leak, cellulitis, or wound
dehiscence requiring any additional intervention. This included re-
opening of an existing ventriculostomy for CSF drainage, oversew-
ing of the wound, or placement of a patient on empiric antibiotics
for suspected wound cellulitis. Importantly, these strict criteria
likely overestimate the rate of wound complications as would be
viewed by most neurosurgeons; however, they provide a rigorous
assessment of the intensity of wound management required to
care for patients undergoing decompressive hemicraniectomy.
‘‘Wound infections’’ were defined as purulent drainage from the
cranial incision or extra-axial empyema found on re-operation.
‘‘CSF leak’’ was defined as any event of CSF leaking from the cranial
incision requiring oversewing, operative wound repair, placement
or opening of a ventriculostomy, or a CSF diversion procedure.
‘‘Wound dehiscence’’ was defined as a separation of the wound
edges requiring over-sewing of skin edges at the bedside or return
to the OR for revision or closure. ‘‘Complex wounds’’ were defined
as surgical wounds with features that rendered them high risk for
wound complications and infections. This included decompressive
hemicraniectomies that involved open skull fractures, significant
scalp lacerations penetrating the galea in or around the planned re-
verse question mark incision, and penetrating head injuries.
The specifics of the bone and soft tissue injury and the tech-
niques of wound closure were recorded from the admission emer-
gency record and the operative reports. Patients were carefully
followed and wound complications tracked prospectively. Results
of wound complication from skin closure with staples, running ny-
lon and vertical mattress nylon stitches were compared to those
obtained with the new Monocryl wound closure protocol.
2.5. Statistical analysis
Binary outcomes, such as wound complications, were compared
using the chi-squared (v2) test. Continuous variables are presented
as mean ± standard error (SE). Comparisons of continuous vari-
ables, such as age and initial Glasgow Coma Scale score were made
using an analysis of variance (ANOVA) using aggregated standard
deviations after determining that the data were distributed nor-
 M.E. Sughrue et al. / Journal of Clinical Neuroscience 18 (2011) 1201–1205
mally. Post-hoc comparisons were performed to further analyze
statistically significant data. Bonferroni corrections were per-
formed for subgroup analyses when relevant. A p < 0.05 was con-
sidered statistically significant.
3. Results
3.1. Patient characteristics
Over the study period, 167 patients underwent unilateral
decompressive hemicraniectomy. Thirty-five patients expired or
were transferred to another hospital prior to post-operative day
5 and were excluded from further analysis. The study group com-
prised the remaining 132 patients, of whom 127 were analyzed
(five wounds closed using the vertical mattress technique were
not included in the analysis). The clinical characteristics and mech-
anisms of injury of this patient group included 124 patients under-
going decompressive hemicraniectomy for trauma (Table 1 – 120
of 127, plus four patients with wound closure using the vertical
mattress technique), while eight operations were performed for
vascular indications. The length of follow-up ranged from 5 days
to 4 years after surgery, largely dependent on the patient’s
outcome.
A total of 29 patients had their wounds closed with the monoc-
ryl wound closure technique described in section 2.3. We closed 40
patients’ incisions with a running nylon stitch, five patients re-
ceived an interrupted nylon vertical mattress closure, and 58 inci-
sions were closed with staples. There was one wound complication
requiring re-operation for wound washout and repair in the verti-
cal mattress cohort. However, due to small number of patients in
the vertical mattress group, we did not analyze the data for this co-
hort separately in subgroup analyses. Patients closed with monoc-
ryl sutures were statistically older than those closed with either
staples or running nylon sutures (Table 1, ANOVA p < 0.05). The ini-
tial Glasgow Coma Scale score did not differ significantly between
groups (Table 1 ANOVA p = NS). Additionally, the fraction of com-
plex wounds (that is, open skull fractures, large subgaleal lacera-
tions in the operative region, and/or penetrating head injuries)
did not differ between groups (Table 1 v2 p = NS).
3.2. The effect of our wound closure technique on wound complication
rates
The use of the Monocryl wound closure technique dramatically
reduced the overall rate of wound complications following decom-
Table 1
(a, upper) Comparison of clinical characteristics of the patients in this study and (b, lower) mechanisms of injury for the patients in this study
Monocryl
No. of patients 29
Age (years, mean ± SE) 48 ± 3.6
Gender (M/F) 23/6
Admit GCS score (mean ± SE) 8.2 ± 0.8
Complex wounds (%) 6/29 (21%)
Open skull fractures 5/29 (17%)
Mechanism
Fall
Pedestrian versus automobile
GSW
Found down
Bike versus automobile
Assault
Other  19/127 (15)
Intracerebral hemorrhage
F = female, GCS = Glasgow coma scale, GSW = gun shot wounds, M = male, SE = standard error.
Includes being hit by a train, motorcycle accidents, and skateboard accidents.
à
Of 132 patients, five had wound closure using the vertical mattress suturing technique and were not analyzed, leaving 127 treated as shown.
1203
pressive hemicraniectomy with cruciate durotomy compared to
the use of any other closure technique. This was true whether
wound complications were defined using standard criteria (stan-
dard criteria: 95% confidence interval [CI] Monocryl 0%, versus
[vs.] running nylon, 7.1–35%, vs. staples 5.2–26%, v2 p < 0.05)
(Fig. 1) or strict criteria (Monocryl 0% vs. others 35%, v2
p < 0.001). Thus, even when strict criteria were applied to ensure
that we were not missing wound complications in the Monocryl
cohort, there were no noted or suspected wound complications
in the patients undergoing hemicraniectomy at our hospital who
were closed using this new technique. More specifically, no pa-
tients required antibiotics for wound erythema, treatments for
meningitis, reoperation for suspected empyema, oversewing of a
suspected CSF leak site, or CSF drainage from an existing ventricu-
lostomy to control wound leakage.
Comparison of the rates of individual subtypes of wound com-
plications (defined using strict criteria) including major or minor
wound infection, major or minor CSF leak, and wound dehiscence
or breakdown requiring bedside or operative repair for the differ-
ent closure techniques are shown in Fig. 2. Again, patients closed
using the Monocryl wound protocol experienced markedly re-
duced rates of wound cellulitis and wound infection (Monocryl
0% vs. running nylon 33% vs. staples 28%, v2 p < 0.01), together with
fewer major or minor CSF leaks (Monocryl 0% vs. running nylon
30% vs. staples 22%, v2 p < 0.05). A statistical trend towards re-
duced rates of wound dehiscence requiring repair or reoperation
was also seen (Monocryl 0% vs. running nylon 18% vs. staples
12%, v2 p = 0.13). A total of eight patients (8%) of the 98 patients
closed with nylon suture or staple techniques required permanent
CSF diversion procedures, while there were no shunts required in
the Monocryl cohort.
4. Discussion
In this study, we present our results with a method for closing
decompressive hemicraniectomy wounds which markedly reduced
our wound complication rate to essentially zero almost immedi-
ately after its initiation. In many instances, this closure technique
was used successfully to close open and penetrating head injuries,
and to handle complex scalp incisions. While this cohort is not
large enough to claim that wound complications are completely
eliminated with this closure technique, we have yet to see a
wound-related complication associated with this technique. Based
on our baseline rates prior to instituting this method, we might
have expected to see approximately 10 minor or major wound
Running nylon Staples p-level
40 58
38 ± 2.4 39 ± 2.1 p < 0.05
31/9 46/12 P = NS
7.1 ± 0.7 7.8 ± 0.6 P = NS
7/40 (18%) 3/58 (5%) P = NS
5/40 (13%) 3/58 (5%) P = NS
No. of patients (%)à
46/127 (36)
19/127 (15)
12/127 (9)
10/127 (8)
7/127 (6)
7/127 (6)
7/127 (6)
1204 M.E. Sughrue et al. / Journal of Clinical Neuroscience 18 (2011) 1201–1205
Fig. 1. Graph of wound complication rates defined using conventional criteria
showing that use of the Monocryl wound closure technique dramatically reduced
the overall rate of wound complications following decompressive hemicraniectomy
with cruciate durotomy compared to the use of any other closure technique. The
vertical bars represent 95% confidence intervals (CI) depicting the overall rates of
wound complications for patients closed using running nylon stitches, skin staples,
and the running Monocryl stitch. The combined group depicts the 95% CI for all
patients closed with anything other than a Monocryl stitch.
Fig. 2. A bar graph of wound complication rates defined using strict criteria. Bar
graphs comparing the rates of any wound complication (labeled as ‘‘All’’), wound
infection (labeled as ‘‘Infection’’), wound dehiscence requiring reoperation (labeled
‘‘I&D’’), cerebrospinal fluid (CSF) leak, and meningitis between patients closed using
the Monocryl technique described (labeled ‘‘M’’), running nylon (labeled ‘‘R’’), and
staples (labeled ‘‘S’’).
complications in this cohort. This suggests that the excess wound-
related morbidity reported after decompressive hemicraniectomy
is related to the inadequacy of standard wound closure techniques
rather than to inherent difficulties with sterility in operative trau-
ma wounds. Our results suggest that routine closure techniques
typically used for standard craniotomies fail for decompressive
craniectomy surgeries wherein the scalp flap is long and based
on a small vascularized pedicle, when the dura is not closed, and
when the wounds are subjected to added tension from extra-axial
CSF hygromas.
Our rates of wound complication for the standard closure
groups using running nylon suture and staples exceed those pub-
lished in the literature. In part, this is the result of our decision
to apply strict criteria for wound complications in this analysis, to-
gether with our institutional system of prospective data collection.
Prospective tracking of wound healing is less likely to miss minor
wound complications, such as scant CSF egress that resolves after
re-opening of a ventriculostomy. This rigor of event tracking is
probably not included in retrospective series with less strict defini-
tions for wound complications.
Regardless, we have found that conventional skin closure meth-
ods applied to decompressive craniectomy surgeries are associated
with unacceptably high rates of wound complication for a supra-
tentorial procedure. The potential reasons for these high complica-
tion rates with decompressive craniectomy procedures are
multifactorial. They include the high rate (25–40%) of CSF absorp-
tion problems and subgaleal hygromas caused by this surgery,6–10
the frequent need to operate in a contaminated field due to open
head injuries,6 the very large scalp flap fed by a relatively narrow
pedicle,6 the frequent injury of the superficial temporal artery dur-
ing emergency exposure of the subtemporalis space,6 and the sys-
temic immunosuppression and critical illness seen in polytrauma
patients with head injuries and elderly stroke patients. These fac-
tors, together with the widely patent dura in the durotomy ap-
proach, make meticulous wound closure even more important.
Alternatively, duraplasty with synthetic materials adds a long
and often tedious step,2 incorporates a foreign body into the heal-
ing wound,2,6,7 and is of unproven efficacy.2 While the duraplasty
technique likely reduces the risk of CSF leak and wound complica-
tions, there are downsides to performing this procedure in this set-
ting, which lead to our interest in eliminating this step. In the
setting of acute trauma, patients, especially with those with poly-
trauma, are often unstable, and the time spent performing dura-
plasty can lead to life-threatening increases in operative times.
Furthermore, the attendant brain swelling that often accompanies
decompressive craniectomy frequently demands a rapid closure to
avoid extracranial brain herniation through the bony defect. We
have found that wound complication rates can be brought to very
acceptable levels with the Monocryl closure technique even with
an open durotomy, thereby abrogating the need to subject the pa-
tient to the time and potential risks of performing a duraplasty.
The apparent improvement in wound outcomes and infection
rates that we achieved by closing our hemicraniectomy wounds
with the Monocryl technique challenges a commonly held belief
regarding wound closures. The successful use of a self-absorbing
stitch that remains in situ for several weeks, argues against the
concept that skin closure devices/sutures are foreign bodies that
predispose to infection and need to be removed as soon as possible.
This is also supported by previously published excellent outcomes
with absorbable sutures in elective craniotomies.11 The skin is an
excellent barrier against infection, and it is waterproof. In light of
our experience, it seems prudent to leave the skin tightly approx-
imated with suture for as long as is needed to completely seal off
the brain and cranial vault. Further, the Monocryl suture has a
small needle and thus the bites are small and close together. We
think the Monocryl closure technique provides a tighter approxi-
mation of the skin edges with fewer gaps through which fluid,
blood, and other vehicles of contamination can pass. By compari-
son, the intervals between staples, running nylon and vertical mat-
tress suture are larger and coarser because they must be placed in a
way that they can be removed.11
The use of prolonged JP drainage in our Monocryl closure co-
hort, occasionally up to 8 days without any wound complications,
argues against the idea that prolonged closed drainage of a cranial
wound increases the risk of infection. This widely-held view that
drains increase infection rates is in part an extrapolation from
the general surgery literature,12 and there is little evidence to sup-
port this contention in cranial wounds. Rather, this approach is
similar to methods proposed for the management of post-spinal
surgery CSF leaks.13 The link between CSF leakage and infection,
however, is well-established.14–19 We would suggest that pro-
 M.E. Sughrue et al. / Journal of Clinical Neuroscience 18 (2011) 1201–1205
longed JP drainage, in combination with a tighter skin closure
stitch, facilitates the early phases of wound healing following
decompressive craniectomy procedures. Most patients with
decompressive craniectomy in our series developed large CSF
hygromas, but no patient closed with the running Monocryl tech-
nique ever leaked CSF.
In conclusion, we present prospective data to suggest that
wound complications after hemicraniectomy with an open, stellate
durotomy are reduced and may be preventable using a running,
fine Monocryl suture technique for skin closure together with pro-
longed subgaleal JP drainage. This technique modification has the
potential to greatly increase the ability of decompressive hemicra-
niectomy to lead to lasting recoveries for patients with life-threat-
ening brain injuries.
References
1. Albanese J, Leone M, Alliez JR, et al. Decompressive craniectomy for severe
traumatic brain injury: evaluation of the effects at one year. Crit Care Med
2003;31:2535–8.
2. Burger R, Duncker D, Uzma N, et al. Decompressive craniotomy: durotomy
instead of duroplasty to reduce prolonged ICP elevation. Acta Neurochir Suppl
2008;102:93–7.
3. Chen CC, Cho DY, Tsai SC. Outcome and prognostic factors of decompressive
hemicraniectomy in malignant middle cerebral artery infarction. J Chin Med
Assoc 2007;70:56–60.
4. Faleiro RM, Faleiro LC, Caetano E, et al. Decompressive craniotomy: prognostic
factors and complications in 89 patients. Arq Neuropsiquiatr 2008;66:369–73.
5. Schirmer CM, Hoit DA, Malek AM. Decompressive hemicraniectomy for the
treatment of intractable intracranial hypertension after aneurysmal
subarachnoid hemorrhage. Stroke 2007;38:987–92.
1205
6. Stiver SI. Complications of decompressive craniectomy for traumatic brain
injury. Neurosurg Focus 2009;26:E7.
7. Yang XF, Wen L, Shen F, et al. Surgical complications secondary to
decompressive craniectomy in patients with a head injury: a series of 108
consecutive cases. Acta Neurochir (Wien) 2008;150:1241–7 [discussion 1248].
8. Kan P, Amini A, Hansen K, et al. Outcomes after decompressive craniectomy for
severe traumatic brain injury in children. J Neurosurg 2006;105:337–42.
9. Waziri A, Fusco D, Mayer SA, et al. Postoperative hydrocephalus in patients
undergoing decompressive hemicraniectomy for ischemic or hemorrhagic
stroke. Neurosurgery 2007;61:489–93 [discussion 493-484].
10. Yang XF, Wen L, Li G, et al. Contralateral subdural effusion secondary to
decompressive craniectomy performed in patients with severe traumatic brain
injury: incidence, clinical presentations, treatment and outcome. Med Princ
Pract 2009;18:16–20.
11. Paolini S, Morace R, Lanzino G, et al. Absorbable intradermal closure of elective
craniotomy wounds. Neurosurgery 2008;62:490–2 [discussion ONS492].
12. Cruse PJ, Foord R. The epidemiology of wound infection. A 10-year prospective
study of 62, 939 wounds. Surg Clin North Am 1980;60:27–40.
13. Hughes SA, Ozgur BM, German M, et al. Prolonged Jackson-Pratt drainage in the
management of lumbar cerebrospinal fluid leaks. Surg Neurol 2006;65:410–4
[discussion 414-415].
14. Blomstedt GC. Infections in neurosurgery: a retrospective study of 1143
patients and 1517 operations. Acta Neurochir (Wien) 1985;78:81–90.
15. Lietard C, Thebaud V, Besson G, et al. Risk factors for neurosurgical site
infections: an 18-month prospective survey. J Neurosurg 2008;109:729–34.
16. Dashti SR, Baharvahdat H, Spetzler RF, et al. Operative intracranial infection
following craniotomy. Neurosurg Focus 2008;24:E10.
17. Kourbeti IS, Jacobs AV, Koslow M, et al. Risk factors associated with
postcraniotomy meningitis. Neurosurgery 2007;60:317–25 [discussion 325-
316].
18. Korinek AM, Golmard JL, Elcheick A, et al. Risk factors for neurosurgical site
infections after craniotomy: a critical reappraisal of antibiotic prophylaxis on
4578 patients. Br J Neurosurg 2005;19:155–62.
19. Idali B, Lahyat B, Khaleq K, et al. Postoperative infection following craniotomy
in adults. Med Mal Infect 2004;34:221–4.