Neurocrit Care

https://doi.org/10.1007/s12028-018-0568-y

REVIEW ARTICLE

Infection Prevention in the

Neurointensive Care Unit: A Systematic Review
Aaron Sylvan Lord1*, Joseph Nicholson2 and Ariane Lewis1

© 2018 Springer Science+Business Media, LLC, part of Springer Nature and Neurocritical Care Society

Abstract 
Hospital-acquired infections are common in neurointensive care units. We sought to review interventions which may
reduce infection rates in neurocritically ill populations. We conducted a systematic review of studies targeting adult
patients in neuro-intensive care units (neuro-ICUs) with an intervention designed to prevent ICU-acquired infections.
Our outcome of interest was change in the prevalence or rates of infection between active and control arms of these
studies. We excluded studies based on the following criteria: no English full-text version available; pediatric popula-
tion; non-neurosciences ICU population; pre- or intraoperative methods to prevent infection; lack of discrete data
for infection rates/prevalence; studies that were purely observational in nature and did not test an intervention; and
studies performed in resource limited settings. We initially retrieved 3716 results by searching the following data-
bases: PubMed/MEDLINE, EMBASE via Ovid, and Cochrane CENTRAL via Ovid. No date or language limits were used
in the search. Computerized deduplication was conducted using EndNote followed by a confirmatory manual review
resulting in 3414 citations. An additional 19 manuscripts were identified through review of references. The screening
process followed a standard protocol, using two screeners at the title/abstract level to determine relevance and at
the full-text level to determine eligibility for inclusion. The 3427 titles/abstracts were independently screened by two
board-certified neurointensivists to determine relevance for full-text review, and 3248 were rejected. The remaining
179 abstracts were reviewed in full text using predetermined inclusion/exclusion criteria. Ultimately, 75 articles met
our inclusion criteria and were utilized in the final analysis. The reviewed literature highlights the need for collabora-
tive, multi-disciplinary, and multi-pronged approaches to reduce infections. Rates of VRI, SSI, VAP, CAUTI, and CLABSI
can approach zero with persistence and a team-based approach.
Keywords:  Hospital-acquired infections, Neurointensive care, Infection prevention, Care bundles, Surgical site
infections, Ventriculitis, Ventilator-associated pneumonia

Introduction pneumonias  (VAPs) in mixed medical and surgical ICU

Hospital-acquired infections are common in intensive patients [1–4]. However, patients in neurointensive care
care units (ICUs) including those dedicated to the care units may have distinct characteristics that predispose
of neurological patients. There is an extensive literature them to infections when compared to a general ICU
on preventing general ICU infections such as catheter- patient population including high utilization of external
associated urinary tract infections (CAUTIs), central ventricular drains (EVDs), high rates of dysphagia and
line-associated blood stream infections (CLABSIs), sur- urinary retention, and the presence of stroke and brain
gical site infections (SSIs), and ventilator-associated injury-induced immunosuppression. Given the increas-
ing focus on reducing rates of hospital-acquired infec-
tions and the growing use of these metrics as proxies for
*Correspondence: [email protected] quality of care, we sought to review current interventions
1
Departments of Neurology and Neurosurgery, New York University which may reduce rates of these infections in neurocriti-
School of Medicine, New York, NY, USA
Full list of author information is available at the end of the article cally ill populations.
Methods dramatically decrease SSI rate. Le et  al. [7] reported a
This systematic review was conducted and reported in pre/post study on the effect of a perioperative care bun-
accordance with PRISMA statement [5]. This systematic dle to reduce SSI and other complications after cranio-
review protocol was registered in PROSPERO (Regis- plasty. The bundle included four doses of perioperative
tration Number: CRD42017057281) in February 2017. vancomycin, a barrier dressing through postoperative day
Search strategies were developed in collaboration with 3, and decolonization of the surgical incision using topi-
a medical librarian. The research question was framed cal chlorhexidine from postoperative day 4–7.
using the PICO format, and searches were developed In development of their bundle, the authors noted
using a combination of keywords and subject headings the high rate of methicillin-resistant Staphylococcus
specific to population and intervention terms (see Sup- aureus colonization compared to other patients in the
plemental Material for search terms). The population neurosciences ICU (19 vs. 6%). The care bundle led to
of interest was adult patients in neurosciences ICUs in decreased rates of SSI (24–3%, p = 0.02) and eliminated
developed countries. The studied interventions were the need for redo cranioplasty (19 vs 0%, p = 0.02). Hale
either (a) randomized-controlled trials (RCTs) or (b) et  al. [8] report on a pre- and postoperative bundle to
cohort studies with measurements before and after a reduce SSI. They introduced a “Craniotomy Checklist”
protocol change. We only selected studies with primary which included preoperative chlorhexidine gluconate
or secondary outcomes of prevention of ICU-acquired (CHG) shampoo, preoperative CHG-alcohol skin prep,
infections. Exclusion criteria for studies included: no postoperative incision care orders, and a new glycemic
English full-text version available; pediatric population; control protocol. Craniotomy SSI decreased from 4.4 to
non-neurosciences ICU population; pre- or intraopera- 1.2% (p = 0.03). As with other care bundles, it is difficult
tive methods to prevent infection; lack of discrete data to assess the efficacy of individual components, but the
for infection rates/prevalence; studies that were purely effectiveness of the bundles demonstrated here denotes
observational in nature (i.e., did not test a protocol that very low rates of SSI are possible.
change); and resource limited settings. In a slightly difference approach, Adogwa et  al. [9]
We initially retrieved 3716 results by searching the reported on use of negative pressure wound therapy on
following databases: PubMed/MEDLINE, EMBASE via patients undergoing long-segment thoracolumbar fusion.
Ovid, and Cochrane CENTRAL via Ovid. No date or In 46 patients, the device was left in place until postoper-
language limits were used in the search. Computerized ative day 3 compared with 114 controls. All patients had
deduplication was conducted using EndNote followed subfascial drains. The authors demonstrated a relative
by a confirmatory manual review resulting in 3414 cita- decrease of 30% in rates of SSI (11 vs. 15%, p = 0.04) as
tions. An additional 19 manuscripts were identified well as decreased rates of wound dehiscence (6 vs. 12%,
through review of references. The screening process fol- p = 0.02).
lowed a standard protocol, using two screeners at the
title/abstract level to determine relevance and at the Infection Prevention for Patients with External Ventricular
full-text level to determine eligibility for inclusion. The Drains (EVDs)
3427 titles/abstracts were independently screened by EVD Bundles
two board-certified neurointensivists to determine rel- Table  1 describes the protocol details of 18 different
evance for full-text review, and 3248 were rejected. The EVD bundles extracted from 17 manuscripts describ-
remaining 179 abstracts were reviewed in full text using ing 15 unique cohorts [10–26]. All of the manuscripts
predetermined inclusion/exclusion criteria. Ultimately, are quasi-experimental studies assessing the association
75 articles met our inclusion criteria and were utilized between implementation of an EVD protocol and the
in the final analysis. An additional six articles on glyce- rate of ventriculostomy-related infections (VRIs). The
mic control were added during the review process of this protocols focus on a number of key categories related
paper bringing total number of articles to 75. See Fig. 1 to EVD placement and maintenance including: (a) use
for PRISMA flowchart [5]. GRADE criteria were used of a checklist; (b) insertion setting (operating room vs.
to determine quality of evidence at the study level. After ICU); (c) strictness of sterility during the procedure; (d)
full-text review, studies were grouped based on type of hair removal; (e) catheter-type; (f ) tunneling; (g) use of
intervention [6]. occlusive dressings; (h) post-procedure prophylactic
antibiotics; (i) reduction in frequency of cerebrospinal
Results fluid (CSF) sampling; (j) aseptic technique for sampling
Methods to Reduce Surgical Site Infections or manipulation of drain; and k) routine EVD exchanges
The following studies demonstrated that attempts to for extended dwell times. All protocols were developed,
decrease SSI via vigilant adoptions of care bundles can implemented, and published by the multi-disciplinary
 Idenficaon Records idenfied through Addional records idenfied
database searching through other sources
(n = 3716) (n = 13)

Records aer duplicates removed

(n = 3427)
Screening

Records screened Records excluded

(n = 3427) (n =3248)

Full-text arcles assessed

for eligibility Full-text arcles excluded
Eligibility

(n = 179) (n = 110)
Not Adult Neuro ICU specific: 38
Pre- or Intra-op: 21
Purely observaonal: 14
Meeng abstract later published
as full manuscript: 10
Insufficient Data: 9
No English or Full Text: 8
Guidelines/Reviews: 5
Not related to infecon
prevenon: 4
Included

Resource Limited Se—ng: 1

Studies included in
qualitave synthesis 6 addional arcles added
(n = 75) in review process

Fig. 1  PRISMA 2009 flow diagram. Reproduced from: Moher et al. [77]

stakeholders in care for patients with EVDs. All protocols different phases provides evidence that a multi-pronged
involved staff education and many included implementa- approach is likely necessary to achieve very low rates of
tion of periodic infection control rounds. VRI.
The large number of interventions that occur with
implementation of each bundle makes it difficult to assess Silver and Antibiotic‑Impregnated EVDs
the effectiveness and necessity of each component of the A number of studies evaluated the benefits of antibiotic-
bundle. Additionally, the heterogeneous patient popula- impregnated EVDs as compared with standard catheters.
tions, variations in surveillance techniques, and unique Wong et  al. [27] conducted a prospective RCT compar-
definitions of VRI complicate comparisons of the preva- ing the infection rate in patients with standard EVDs
lence of VRI across institutions. However, these bundles (who received dual prophylactic systemic antibiotics) and
demonstrate that implementation of a comprehensive patients with EVDs impregnated with clindamycin and
EVD bundle that emphasizes aseptic technique through- rifampicin (who received only periprocedural systemic
out the entire lifespan of the EVD has the ability to signif- antibiotics). They found no ventriculostomy-related
icantly decrease rates of VRI and that rates approaching 0 infections in either group and no significant difference in
are possible. The incremental improvements seen in the the rates of extracranial infections (51 vs. 46%, p = 0.617).
studies which introduced aspects of their bundles over In a 2003 prospective RCT of 288 patients at six medical
Table 1  Characteristics of external ventricular brain bundles
Study Study Inser- OR Full Second Hair Iodine/ Anti- Tunneling Occlu- Routine Post- No Routine Other Pre- Post-
quality tion place- barrier pair removal chlo- biotic/ sive dress- proce- Daily EVD protocol protocol
check- ment pre- of gloves rhex- silver- dress- ing dure CSF exchange infection infection
list pre- cau- idine EVDs ing change antibi- sam- rate rate
ferred tions otics pling

Bader Very N Y Y N Shave Y N N Y N Y Y N No irriga- 5/9 (55.5%) 0/10 (0%)

et al. low tion or
[10] tubing
changes;
In lieu
of CSF
draws,
send
entire
drainage
bag using
aseptic
tech-
nique;
aseptic
tech-
nique for
zeroing/
manipula-
tion
Korinek Low N Y N N Use Y N Y (3 cm) N Y (q72H) N Y N Send 16/161 10/216
et al. clippers drainage (9.9%) (4.6%)
[11] on full bag for
scalp infectious
work-up
and if pos-
itive do
an aseptic
draw from
tubing
Dasic Low N Y Y Y Shave Y N Y (mini- Y N N Y N 10 mg IT 14/51 (27%) 7/59 (12%)
et al. mum vanco
[12] 10 cm) after
place-
ment
Harrop Low N N Y N Use clip- Y N N Y N N Y N 22/327 23/281
et al. pers (6.7%) (8.2%)
Part 1
[13]
Table 1  continued
Study Study Inser- OR Full Second Hair Iodine/ Anti- Tunneling Occlu- Routine Post- No Routine Other Pre- Post-
quality tion place- barrier pair removal chlo- biotic/ sive dress- proce- Daily EVD protocol protocol
check- ment pre- of gloves rhex- silver- dress- ing dure CSF exchange infection infection
list pre- cau- idine EVDs ing change antibi- sam- rate rate
ferred tions otics pling
Harrop Low N N Y N Use clip- Y Y N Y N N Y N 22/327 2/195 (1.0%)
et al. pers (6.7%)
Part 2
[13]
Honda Low N N Y N Use clip- Y N N Y Y (q48H) N N N Cap and 3.56/1000 2/1000 EVD
et al. pers mask for EVD Days days
[14] everyone
Part 1 in room;
sterile
gauze/
Tegaderm
at inser-
tion site
Honda Low N N Y N Use clip- Y Y N Y Y (q48H) N N N Cap and 3.56/1000 0.87/1000
et al. pers mask for EVD Days EVD Days
[14] everyone
Part 2 in room;
sterile
gauze/
Tegaderm
at inser-
tion site
Lever- Very N Y N N Use N N Y (> 5 cm) N N N Y N Use aseptic 20–37% 9%
stein- low clippers tech-
Van on full nique for
Hall scalp sampling;
et al. Used
[15] algorithm
for diag-
nosis of
suspected
VRI
Table 1  continued
Study Study Inser- OR Full Second Hair Iodine/ Anti- Tunneling Occlu- Routine Post- No Routine Other Pre- Post-
quality tion place- barrier pair removal chlo- biotic/ sive dress- proce- Daily EVD protocol protocol
check- ment pre- of gloves rhex- silver- dress- ing dure CSF exchange infection infection
list pre- cau- idine EVDs ing change antibi- sam- rate rate
ferred tions otics pling
Amini Low Y N Y Y Use clip- Y Y N N Y (q72H) N N N Checklist 9.20% 0.46%
et al. pers devel-
[19] oped; RN
Rah- can stop
man procedure
et al. if not
[17] adher-
Kubilay ing to
et al. checklist;
[18] devel-
oped
Neuro-
science
Infection
Preven-
tion Team
Hill Low Y N Y Y Use clip- Y N Y Y N N N N Weekly 16/1000 1st year:
et al. pers infection EVD Days 4.5/1000
[19] control EVD days;
rounds; 2nd Year:
aseptic 1.3/1000
tech- EVD days;
nique for 3rd year
zeroing/ 0/1000
manipu- EVD days
lation;
Improved
documen-
tation;
increased
surveil-
lance
Lwin Low N Y N N Not indi- N N N N N N Y Y (Day 10) Strict 5/82 (6.1%) 3/79 (3.8%)
et al. cated aseptic
[20] technique
Part 1 for all
manipula-
tions
Table 1  continued
Study Study Inser- OR Full Second Hair Iodine/ Anti- Tunneling Occlu- Routine Post- No Routine Other Pre- Post-
quality tion place- barrier pair removal chlo- biotic/ sive dress- proce- Daily EVD protocol protocol
check- ment pre- of gloves rhex- silver- dress- ing dure CSF exchange infection infection
list pre- cau- idine EVDs ing change antibi- sam- rate rate
ferred tions otics pling
Lwin Low N Y N N Not indi- N Y N N N N Y Y (Day 10) Strict 5/82 (6.1%) 0/73% (0%)
et al. cated aseptic
[20] technique
Part 2 for all
manipula-
tions
Flint Low N N Y N Use clip- Y Y Y (3–5 cm) Y N N Y N Strict 9/143 (6.3%) 1/119 (0.8%)
et al. pers on aseptic 7.35/1000 0.79/1000
[21] large technique EVD Days EVD Days
area for all
manipu-
lations;
chlo-
rhexidine
biopatch
Cama- Low N Y Y N Use Y N Y (5 cm) Y Y (q24) Y Y N CSF draws 9.5% 4.8% 7/1000
cho clippers only from 14/1000 EVD days
et al. on full distal EVD Days
[22] scalp point
Zakaria Low N Y N N Not indi- Y Mix Y N N N Y N Weekly 54/234 18/132
et al. cated infection (23.1%) (13.6%)
[23] control 21.5/1000 13.7/1000
rounds; EVD days EVD days
aseptic
tech-
nique for
manipula-
tion
Chatzi Low Y Y Y N Not indi- N N N N N N Y Y (Day 7) Unit-based 28% 10.6%
et al. cated reeduca- 18/1000 7.1/1000
[24] tion twice EVD days EVD days
monthly;
Full barrier
precau-
tions/
asepsis
when
accessing
the drain
Angulo Very N N N N Not indi- N Y N N N N Y N 2.6/1000 0.6/1000
et al. low cated EVD days EVD days
[25]
 infection centers, Zabramski et  al. [28] compared standard EVDs
with minocycline/rifampin-impregnated EVDs and
protocol

11.5%
found that the standard EVDs were twice as likely to
Post-

rate

become colonized (37 vs. 18%, p < 0.0012). Many stud-

ies by other authors confirmed the findings by Zabram-
infection
protocol

ski et  al., [29–31], but two retrospective observational

20.9%

cohort studies found no significant difference in the rate

Pre-

rate

of infections for patients with standard EVDs and those

with antibiotic-impregnated EVDs [32, 33]. Although
prior rou-
2–3 days;

tine EVD
changes
CSF draws

stopped

Mikhaylov et al. found that antibiotic-impregnated EVDs

Other

ever

were associated with lower 3-year mortality rate than

standard EVDs (15.8 vs. 24.6%, p = 0.21), Shekhar et  al.
exchange

found that 20% of patients with antibiotic-impregnated

Routine

EVDs died and 12% of patients with standard EVDs died.

EVD

The types of antibiotics in antibiotic-impregnated EVDs

N

vary, but in a study that compared the rates of infection

pling
Daily

sam-

in patients with minocycline/rifampin-impregnated cath-

CSF
No

eters and clindamycin/rifampin-impregnated catheters,

antibi-
proce-

Abla et al. [34] found no positive cerebrospinal fluid cul-

Post-

otics
dure

tures in either cohort.

N

Silver-bearing catheters have also been used to pre-

Routine

change
dress-

vent meningitis. One study of 164 patients that compared

ing

standard EVDs with silver-bearing EVDs showed that

N

the occurrence of a positive cerebrospinal fluid culture,

Tunneling Occlu-

dress-

colonization of the catheter tip, or cerebrospinal fluid

sive

ing

white blood cell count > 4cells/μl was significantly less

in patients with silver-bearing catheters (18.9 vs. 33.7%,
CSF cerebrospinal fluid, EVD external ventricular drain, IT intrathecal, OR operating room, RN registered nurse

p = 0.04) [35].
While there are insufficient data to state whether silver-
N

bearing or antibiotic-impregnated catheters are superior

biotic/
silver-
EVDs
Anti-

to one another, the above evidence suggests that both are

N

superior to standard EVDs with respect to infection pre-

Iodine/

vention. Silver- or antibiotic-impregnated catheters are

rhex-
idine
chlo-

included in many EVD bundles in neurosciences ICUs

N

where the VRI rate approached zero. Additionally, use of

removal

Not indi-
cated

silver- or antibiotic-impregnated catheters may reduce

Hair

need for prolonged prophylactic systemic antibiotics.

of gloves
Second

Prolonged Prophylactic Antibiotics for EVDs

The data as to whether prolonged prophylactic antibiotics
barrier pair

improve VRI rates are mixed. In one RCT of 255 patients

comparing single (cefepime) and dual (ampicillin/sulbac-
tions
cau-
pre-
Full

tam and aztreonam) antibiotics for the duration of EVD

N

presence, the authors found 12% of patients in the single

place-

ferred
ment

group and 6% of patients in the dual group developed

pre-
OR

VRI (p = 0.7) [36]. In a separate RCT, Poon et  al. [37]

Y

found rates of VRI and extracranial infections were lower

check-
Study Inser-
quality tion

in patients who received ampicillin/sulbactam and aztre-

Table 1  continued

list

onam until EVD removal as compared with patients who

received periprocedural ampicillin/sulbactam (3 vs. 11%,
low
Very

p = 0.01 for VRI and 20 vs. 42%, p = 0.002 for extracranial

infections).
et al.
Study

[26]
Phan
 While these studies showed superiority of aggres- EVD Exchanges
sive antibiotic prophylaxis of EVDs, other studies found Only one study provides experimental evidence on
results questioning their efficacy [26, 38, 39]. One ret- the effect of routine EVD exchange on rates of VRI.
rospective study comparing 209 patients who received This small RCT of 103 patients compared regular EVD
cefuroxime for the duration of EVD placement and 99 exchange every 5  days to exchange only when clinically
patients who received it periprocedurally (for three or indicated [42]. Patients requiring a new catheter had it
less doses) found that the infection rates were nearly placed contralateral to existing catheter. The authors
identical (3.8 vs. 4.0%). The authors estimated that dis- found VRI in 8% of patients active arm versus 4% in the
continuation of prolonged prophylactic antibiotics would control arm (p = 0.44). The mean number of EVDs in
lead to a savings of $80,000 per year in direct drug costs the active group was 2.4, while only one patient needed
[38]. Additionally, in a prospective observational study of a catheter exchange in the control group (for occlusion).
866 patients with antibiotic-impregnated catheters, Mur- All infections occurred after day 10 in both groups.
phy et  al. found that there was no significant difference A small study of 32 patients investigated the effect of
in the rate of infection between patients who were given decreasing the frequency of changing the EVD drainage
cefazolin or vancomycin for the duration of EVD place- set [43]. They demonstrated a reduction in “evidence of
ment versus those who were given antibiotics peripro- ventriculitis” from 69 to 37% by changing the drainage set
cedurally (1.1 vs. 0.4%, p = 0.22). The authors noted that less frequently, from 3 to 7 days. However, the high rate
use of periprocedural antibiotics led to cost savings of of infection in this study makes generalization difficult.
$162,516 in drug costs and reduction in nosocomial
infections [39]. Dellit et  al. [40] also found that limiting Methods to Prevent VRI at the EVD Insertion Site
prophylactic antibiotic use to periprocedural adminis- In an effort to reduce infections with typical skin com-
tration decreased the incidence of Clostridium difficile mensals, Schodel et al. [44] examined the effect of using
infections. Of note, they used antibiotic-impregnated a bolt-kit to isolate the EVD as it traverses the skin upon
EVDs and there was no change in number of positive skull entry. In this pre/post study, the authors found a
CSF cultures after discontinuation of prolonged prophy- small but significant decrease in VRI using the bolt-kit
lactic antibiotics. system compared to traditional twist-drill hole only (4.9
These studies suggest that the utility of prolonged pro- vs. 6.8%, p = 0.034). The results remained significant after
phylactic antibiotics is dependent on a number of fac- controlling for patient age, drainage time, and number of
tors including background rate of VRI, concurrent use punctures.
of other infection prevention methods or “bundles,” and Drawing on parallel efforts to reduce CLABSIs, one
use of silver or antibiotic-impregnated EVDs. Notably, study examined the isolated effect of chlorhexidine-con-
decreased antibiotic use may be associated with lower taining dressings on rates of VRI [45]. The authors saw
rates of Clostridium difficile and growth of resistant a reduction from 6.98 to 1.70 VRI per 1000 EVD days,
bacteria. almost exclusively due to decreased infection with staph-
ylococcus species.
Frequency of CSF Sampling Similarly, another quasi-experimental study examined
One study directly addressed the effect of decreasing the the effect of a cyanoacrylate adhesive (Dermabond) on
frequency of CSF sampling on rates on frequency of VRIs rates of VRI in patients mostly suffering from non-trau-
[41]. In this quasi-experimental study of mostly suba- matic intracranial hemorrhage [46]. Dermabond, accord-
rachnoid hemorrhage (SAH) and traumatic brain injury ing to the authors, is a “water-catalyzed adhesive that is
(TBI) patients with EVDs, the authors looked at the effect especially formulated for use on skin and indicated for
of daily CSF sampling versus every 3-day sampling on the primary closure of small, clean surgical wounds….
frequency of “suspected” VRI (defined as those who were and as a barrier against common bacterial microbes.” The
treated with antibiotics) and “proven” VRI (defined as authors showed a reduction in VRI frequency from 15 to
those who were treated and had positive CSF cultures). 4% (p = 0.002), mostly due to reduction in rates of staphy-
The incidence of “suspected” VRI decreased from 17 to lococcus infections. This reduction supports the authors’
11% and that of “proven” VRI from 10 to 3%. After con- conclusions that Dermabond reduces VRI by providing a
trolling for other factors associated with VRI, the authors barrier to the entry of gram-positive skin flora along the
found the odds ratio for VRI was 0.44 in the every 3-day EVD tract.
sampling group compared to daily.
Infection Prevention for Patients with Intracranial Pressure ICU-acquired infections, while two showed overall lower
Monitors rates of infections (in SAH and TBI populations). How-
Only one study examined the effect of prolonged pro- ever, the authors of the study that showed decreased rates
phylactic antibiotics in patients with parenchymal of infection with intensive insulin management in SAH
intracranial pressure (ICP) monitors [47]. This study of patients were unable to replicate their findings in a fol-
279 patients with TBI evaluated the effect of a protocol low-up study which on patients with TBI [53]. A recent
change from use of a broad-spectrum antibiotic (ceftri- meta-analysis pooled the rates of ICU-acquired pneumo-
axone) to a narrow-spectrum antibiotic (cefazolin) with nia and found no difference in pooled rates of pneumonia
the hypothesis that tailoring coverage would not change between different arms of glucose control across studies
overall central nervous system (CNS) infection rates (RR 1.04, 95% CI 0.82–1.32, p = 0.73) [49].
but would result in less resistant infections when they
occurred in both the CNS and systemically. The authors Infection Prevention after CNS Trauma
found an insignificant difference in CNS infections rate Five studies reported the results of RCTs examining the
between the narrow-spectrum group (1.7%) and the use of prophylactic antibiotics in patients with basilar
broad-spectrum group (4.4%). Of note, 19 out of 119 skull fractures and evidence of CSF leak [50–54]. Of note,
patients in the narrow-spectrum group received no anti- these five studies were the subject of a 2015 Cochrane
biotics for reasons not described. There was a higher pro- systematic review and meta-analysis which found no
portion of systemic infections arising from resistant gram benefit to prophylactic antibiotics for reducing the rates
negative species in the broad-spectrum group, especially of meningitis or meningitis-related mortality in this
in patients that developed infections later in the hospital population [55]. The studies are summarized in Table 2.
course. There were no experimental or quasi-experimen- None of the five studies found a significant reduction
tal studies investigating infection rates with and without in risk of meningitis with prophylactic antibiotic usage,
antibiotic prophylaxis for patients with ICP monitors. although one study did show a decreased rate of com-
bined meningitis and SSI. With one exception, the rates
Infection Prevention for Patients with Subdural or of meningitis in this clinical setting were low, and most
Subgaleal Drains authors recommended against use of prophylaxis based
One study from our group evaluated the effect of a pro- on their findings.
tocol change which eliminated the use of prolonged
prophylactic antibiotics (cefazolin or vancomycin) in Prophylactic Antibiotics to Prevent Hospital‑Acquired
postoperative neurosurgical patients with subdural and Infections in the Neurosciences ICU
subgaleal drains [48]. The 105 patients in the antibi- Prophylactic Antibiotics for Comatose Patients
otic group received 513 doses of cefazolin and 77 doses We identified three studies (2 RCTs, 1 quasi-experimen-
of vancomycin postoperatively until the time of drain tal) that examined whether prophylactic antibiotics were
removal, compared to six doses of cefazolin and one dose effective in lowering rates of ventilator-associated pneu-
of vancomycin postoperatively for the 80 patients in the monia (VAP) in patients presenting with coma [56–58].
no antibiotic group. The rates of SSIs in both groups were Sirvent et  al. [56] published an open-label RCT of 100
similar (1 superficial and 1 deep in the antibiotics group) patients in coma (mostly due to TBI (86%) with some
versus one infection in the no-antibiotics group. There stroke and post-neurosurgery patients) with Glasgow
were two cases of Clostridium difficile in the antibiotics Coma Score (GCS) ≤ 12 and expected mechanical venti-
group versus none in the no-antibiotics group, but this lation > 72 h. The experimental arm (n = 50) received two
was insignificant (p = 0.5). We also noted a cost savings doses of high-dose cefuroxime (1500 mg) 12 h apart com-
of $887.50 per patient in the no-antibiotics group, mostly pared to usual care for the control arm. The incidence of
due to increased costs of treating three patients with pneumonia in the experimental arm was 24% compared
resistant infections in the prolonged antibiotics group. to 50% in the control group (p = 0.007) with most pneu-
There were no resistant organisms grown in the no-anti- monia occurring within the first 4  days of mechanical
biotics group. ventilation. Despite the reduction in incidence of pneu-
monia, there was no difference in morbidity or mortal-
Glycemic Control to Reduce ICU‑Acquired Infections ity between groups. Acquarolo et  al. [57] performed
Six studies reported the results of RCTs on intensive a small randomized, open-label study using 3  days of
insulin therapy versus more liberal glycemic control ampicillin-sulbactam (3  g every 6  h) in 38 comatose
strategies and its effect on hospital-acquired infections patients (GCS ≤ 8) on mechanical ventilation in a mixed
in mixed groups of neurocritically ill patients [49–54]. TBI, stroke, and cardiac arrest population. A reduction
Four of the six studies showed no difference in rates of in early onset (within first 4  days) pneumonia was seen
 (58% control vs. 21% prophylaxis, p = 0.02) without an

cebo group and 0.9% in prophylaxis

Overall high rate of meningitis; inclu-

Three patients in placebo group and

were treated for suspected menin-

one patient in the penicillin group
effect on functional outcome, but the study was under-

2.1% in placebo / 0.0% in treatment If including scalp wounds, rates of

meningitis/SSI were 8.7% in pla-
toward more invasive, resistant

sion criteria required pneumo-

gitis but had negative cultures
powered for this outcome. A quasi-experimental study
Nasopharyngeal flora changed

of 129 ventilated patients (71 in prophylaxis group, 58

gram negative species

in control) with coma (GCS ≤ 8) compared a single dose

of ceftriaxone (2 g) to control in reduction in early onset
(within first 4  days) pneumonia [58]. The groups had
similar age (56 vs. 59), APACHE (17 vs. 18), and GCS
Comments

cephalus
(5 for both groups) scores. The incidence of early onset

group
None

pneumonia was lower in the prophylaxis arm compared

to control (4.4 vs. 23.1/1000 days of mechanical ventila-
tion, p = 0.02). No mortality benefit was seen, though the
1/26 in placebo; 0/26 in treatment

21.5% in placebo / 18.9% in treat-

OR 1.0 (no infections in the three

prophylaxis group had lower mechanical ventilation and

OR 1.0 (no infections in either

ICU days (10 and 15 day control vs. 6 and 10 prophylaxis,

p = 0.01 and 0.02).
Rates of meningitis

Prophylactic Antibiotics for Patients with Severe Ischemic

Stroke
groups)
group)

While the majority of patients included in the seven

ment

published RCTs on use of prophylactic antibiotics after

stroke had mild strokes and thus would not be admitted
to a neurosciences ICU, two studies evaluating the role
of prophylactic antibiotics in patients with severe stroke
Not specified; median 7.0 days

met our inclusion criteria [59, 60]. The Mannheim Infec-

Antibiotic duration (days)

tion in Stroke Study (MISS) was a RCT evaluating the

efficacy of prophylactic mezlocillin/sulbactam for 4 days
(range 4–21 days)

compared to usual care in patients with stroke who had

Table 2  Studies of prophylactic antibiotics in basilar skill fractures with CSF leak

modified Rankin scores > 3 [59]. The study enrolled 30

At least 5

patients in each group within 4  h of symptom onset.

Infections developed in 50% of patients with prophylaxis
Ampicillin or cephalothin 1 g IV q6H 10

and in 90% of patients with usual care in the first 10 days

3

(p = 0.002). The reduction was mostly due to decreased

157 Treatment group 1:  Ceftriaxone 1 g

 Ampicillin/Sulfadiazine 1 g/0.5 g

rates of urinary tract infections (UTIs). The authors

160 Treatment group 1:  Penicillin G

Treatment group 2:  Penicillin G

also note that while the infection rate in the prophylac-

Penicillin G 5 M units IV Q6H

tic group was high, most of these infections occurred

109 Ceftriaxone 1 g IV daily

after prophylactic antibiotics were stopped, with a mean

1.2 M units IV daily

20 M IV units daily

IV intravenous, OR odds ratio, Q6H every 6 hours, SSI surgical site infection
Treatment group 2:

day of infection of 5.1 compared to day 3.3 in the con-

trol arm (p = 0.003). Functional outcomes at 90  days
Antibiotic

were improved in patients in the prophylaxis arm, with

IV daily

Q6H

18 patients with mRS 3–4 versus 9 in the control group

p < 0.05).
The PANTHERIS trial randomized 80 patients with
10

52
Study quality N

severe middle cerebral artery infarcts to either 5 days of

prophylactic intravenous moxifloxacin (ppx) or placebo
Moderate

Moderate

Moderate

Moderate

within 36  h of symptom onset. Both intention-to-treat

(ITT) and per-protocol (PP) analyses were performed
High

due to early deaths, withdrawal of consent, and protocol

violations. The primary outcome was total infection rate
Demetriades et al. [53]

at day 11. In the ITT analysis (n = 79), the rate of infec-

Klastersky et al. [52]

Eftekhar et al. [54]

Ignelzi et al. [50]

tion was 15% in the treated group and 33% in the placebo
Hoff et al. [51]

group (p = 0.1). In the PP analysis (n = 66), the infection

rate was 17% in the treated group and 42% in the con-
Study

trol group (p = 0.03). In contrast to the MISS study, the

rate of UTIs was similar between groups (ITT: 7.7% ppx to be higher in head trauma versus vascular patients.
vs. 12.5% control, p = 0.7; PP 8.6% ppx vs. 16.1% con- There was no difference in mortality between groups.
trol, p = 0.5). The rates of pneumonia were reduced to a Hammond and Potgieter [65] conducted a smaller RCT
greater extent but did not reach statistical significance of SDD on 33 patients with mostly non-surgical neurolog-
(ITT: 7.7% ppx vs. 20% control, p = 0.2; PP: 8.6% ppx vs. ical issues (Guillain-Barré syndrome, meningoencepha-
25.8% control, p = 0.1). Of note, only 9% of patients in the litis, status epilepticus, and myasthenia gravis) who were
trial received mechanical ventilation (1 ppx, 6 control). expected to remain intubated for at least 48 h and remain
Survival curve analysis over the first 180 days showed no in ICU for at least five  days. The SDD consisted of the
survival difference between groups (p = 0.6). same agents as in the Korinek study; however, both SDD
and placebo arms received three days of intravenous cefo-
Prevention of VAP in the Neurosciences ICU taxime. There was an overall higher infection rate in the
Ventilator‑Associated Pneumonia Bundles SDD group (46 vs. 32%, no p value reported). The authors
Weireter et  al. [61] performed a quasi-experimental did not publish data on mortality or functional outcomes.
study evaluating the efficacy of a VAP bundle in reduc- The study is flawed by the use of intravenous cefotaxime
ing rates of VAP in a neurotrauma ICU. The interven- in both arms, which the authors state was added because
tion instituted a weekly multi-disciplinary team meeting, “the use of an intravenous placebo might have resulted in
standardized weaning protocols (including empower- the withdrawal of an excessive number of cases from the
ment of respiratory therapists and nursing to proceed study because the need for antibiotics would have led to
with extubation if all protocol criteria were met), an oral unblinding.” However, given that prophylactic intravenous
care regimen, protocols for patient positioning, subglot- antibiotics for mechanical ventilation are not standard of
tic suctioning for patients with endotracheal tubes, and care, the authors dilute the effect of SDD by their design.
aggressive antibiotic stewardship. VAP rates dropped A brief conference abstract reported the results of a
from 21/1000 patient days to  < 1/1000 patient days. quasi-experimental study on the effect of oral and enteral
Mechanical ventilation time also decreased from 5.8 to SDD with intravenous cefotaxime on rates of nosocomial
4.75 days over a 10-year period. pneumonia in a neurotrauma ICU [66]. They demon-
Two additional conference abstracts reported on the strated a significant decrease in rates of pneumonia, but
efficacy of multi-intervention performance improvement additional details were not available in the abstract.
projects in reducing rates of VAP in the neurosciences
ICU [62, 63]. Both reports showed improvement in VAP. Oral Hygiene
Interventions included use of multi-disciplinary work Three articles discussed the use oropharyngeal decon-
groups, VAP bundles, hand hygiene, oral care, and reduc- tamination to prevent VAP in patients with neurotrauma
tion in transport of patients. or severe stroke.
Seguin et al. [67] performed a single-center RCT with
Selective Decontamination of the Digestive Tract 110 patients to evaluate the efficacy of regular oro-
In 1993, two trials reported conflicting results on the pharyngeal application of povidone-iodine in closed
effect of selective decontamination of the digestive tract head trauma patients with GCS ≤ 8 after initial resus-
(SDD) to prevent pneumonia in mechanically ventilated citation and expected need for mechanical ventilation
neurosciences ICU patients. for ≥ 2  days. The incidence of VAP in the decontamina-
Korinek et  al. [64] performed a double-blinded RCT tion group was 8 vs. 39% in the active control (saline,
of SDD on 123 comatose patients intubated for at least p = 0.003) and 42% in the negative control (no oral appli-
5 days after emergent admission to a neurosurgical ICU cation, p = 0.001). Pneumonias occurred earlier (median
for mostly trauma and vascular insults. The study period day 3) in the povidone-iodine group compared to the
consisted of up to 15 days of a suspension of polymyxin E control group (median day 8). There was no difference in
100 mg, tobramycin 80 mg, and amphotericin B 500 mg length-of-stay or mortality between groups.
via nasogastric tube as well as oral application of a paste Based on these results, a multicenter, placebo-con-
consisting of a 2% mixture of the above antimicrobials trolled randomized double-blind trial was performed
and 4% vancomycin. The primary outcome was incidence to evaluate the same research question in an expanded
of bronchopneumonia. Patients were excluded for infec- patient population that included patients with hemor-
tion, extubation, or death prior to day five. The authors rhagic stroke in addition to neurotrauma patients [68].
found a decrease in the rates of both bronchopneumo- In the 150 patients in the intention-to-treat analysis, they
nia (42 vs. 24%, p < 0.04) and total infections (82 vs. 46%, found VAP rates of 31% in the povidone-iodine group
p < 0.001). The authors found the reduction in infection versus 28% in the control group (RR 1.11, CI 0.67–1.82).
Of note, there were five cases of acute respiratory distress
syndrome in the povidone-iodine group and none in the CAUTI rate in the neurosciences ICU decreased from 3
control arm. There were no differences in ICU and Hos- to 1% (p = 0.001) and length-of-stay decreased from 3 to
pital length-of-stay or ICU and 90-day mortality. 2 days (p = 0.02) though mortality and post-ICU disposi-
Wagner et al. [69] performed a quasi-experimental study tion did not change.
to evaluate whether a protocol for oral hygiene care was Patel et  al. [75] reported on a physician-directed daily
effective in reducing rates of pneumonia in both intubated checklist to assess whether a urinary catheter was still
and non-intubated patients admitted with acute ischemic needed and a nursing-directed protocol to utilize bladder
stroke or intracerebral hemorrhage. After instituting the scanning and clean intermittent catheterization prior to
protocol which consisted of campus-wide nursing educa- insertion of a urinary catheter. They found a nonsignificant
tion to perform systematic assessment of oral health and decrease in CAUTI rate from 3.8 to 2.9 infections/1000 cath-
use of Sage Oral Care kits (0.05% cetylpyridinium chloride eter days (p = 0.8) and a decrease in the percentage of cath-
and 1.5% hydrogen peroxide and suction toothbrush, rates eter days in the ICU (69–50%, p = 0.02). Lusby et al. reported
of VAP decreased from 14 to 10% (p = 0.022). Although on a similar protocol which also utilized chlorhexidine baths
there was a higher proportion of intracerebral hemorrhage and reflex cultures only after positive urinalysis. CAUTI rates
patients in the pre-protocol cohort (47 vs. 31%), a post hoc decreased from 8.6 to 3.7/1000 patient days (p = 0.001) [76].
analysis showed no effect modification by stroke subtype
on the efficacy of intervention. Conclusion
Preventing hospital-acquired infections is an impor-
Prevention of Central Line‑Associated Blood Stream tant focus in neurosciences ICUs. There are numerous
Infections in the Neurosciences Intensive Care Unit infection prevention strategies available. The reviewed
In two reports, Fukunaga et  al. [70, 71] reported on the literature highlights the need for collaborative, multi-dis-
efficacy of povidone-iodine ointment and gauze dress- ciplinary, and multi-pronged approaches. Rates of VRI,
ings in reducing rates of CLABSI in a mixed neurologi- SSI, VAP, CAUTI, and CLABSI can approach zero with
cal ICU population. They showed that application of a persistence and a team-based approach. Given the impact
10% povidone-iodine ointment covered with gauze and preventing these infections may have on patient-centered
a Tegaderm resulted in less CLABSI and central line outcomes as well as their use as important benchmarks
colonization than use of Tegaderm alone (0/1000  days for regulatory agencies and payers on measuring the
CLABSI, 0.6/1000  days colonization vs. 5.22/1000  days quality of care, developing protocols and auditing com-
CLABSI, 9/1000 days colonization, both p < 0.05). pliance for prevention infection should be standard prac-
Elsayed et  al. [72] reported on the efficacy of a multi- tice in modern neurosciences ICUs.
intervention approach in reducing CLABSI rates in a
Electronic supplementary material
mixed neuro-ICU population through use of prepacked The online version of this article (https​://doi.org/10.1007/s1202​8-018-0568-y)
central line kits including all necessary sterile equipment, contains supplementary material, which is available to authorized users.
campus-wide education, a central line bundle, antisep-
tic coated catheters, and biopatches. The CLABSI rate Author details
decreased from 6.24 to 3.7 per 1000/catheter days in the 1
 Departments of Neurology and Neurosurgery, New York University School
first year (no p-value reported). Continued enforcement of Medicine, New York, NY, USA. 2 NYU Health Sciences Library, New York
University School of Medicine, New York, NY, USA.
of the intervention led to 0 CLABSIs by year 4.
Author Contributions
Prevention of Urinary Tract Infections in the Neurosciences Dr. Lord was involved with all aspects of the study including conception
and design, acquisition of data, analysis, interpretation, and drafting. Joseph
Intensive Care Unit Nicholson was involved in the conception and design. Dr. Lewis was involved
Wisniewski et al. [73] reported on the efficacy of a nurs- with all aspects of the study including conception and design, acquisition of
ing driven protocol in a neurosurgical ICU to perform data, analysis, interpretation, and drafting.
meatal and foley care using chlorhexidine wipes to reduce Source of Support
rates of catheter-associated urinary tract infections Dr. Lord was supported by CTSA award to NYU School of Medicine
(CAUTIs). The CAUTI rate decreased from 5.8/1000 to (1UL1TR001445).
0/1000 device days (p < 0.001). Compliance with Ethical Standards
Samuel et  al. [74] described a protocol that included
daily assessment of catheter need according to preset cri- Conflict of interest
The authors declare that they have no conflict of interest.
teria, urine culture only with a positive urinalysis if other
causes of fever were excluded, nursing foley maintenance
strategies, and compliance audits. After implement-
ing the protocol with a 6-month education phase, the
References 23. Zakaria R, Tripathy S, Srikandarajah N, Rothburn MM, Lawson DDA. Reduc-
1. Gould CV, Umscheid CA, Agarwal RK, et al. Guideline for prevention of tion of drain-associated cerebrospinal fluid infections in neurosurgical
catheter-associated urinary tract infections 2009. Infect Control Hosp inpatients: a prospective study. J Hosp Infect. 2013;84(3):215–21.
Epidemiol. 2010;31(4):319–26. 24. Chatzi M, Karvouniaris M, Makris D, et al. Bundle of measures for external
2. Blot K, Bergs J, Vogelaers D, Blot S, Vandijck D. Prevention of central cerebral ventricular drainage-associated ventriculitis. Crit Care Med.
line-associated bloodstream infections through quality improvement 2014;42(1):66–73.
interventions: a systematic review and meta-analysis. Clin Infect Dis. 25. Angulo MN, Fagaragan L, Tabbilos SJ, et al. Improving the practice on
2014;59(1):96–105. external ventricular drains: risk and cost reduction through a multi-
3. Muscedere J, Dodek P, Keenan S, et al. Comprehensive evidence-based collaborative approach. Neurocrit Care. 2015;1:S138.
clinical practice guidelines for ventilator-associated pneumonia: preven- 26. Phan K, Schultz K, Huang C, et al. External ventricular drain infec-
tion. J Crit Care. 2008;23(1):126–37. tions at the Canberra Hospital: a retrospective study. J Clin Neurosci.
4. Bratzler DW, Dellinger EP, Olsen KM, et al. Clinical practice guidelines 2016;32:95–8.
for antimicrobial prophylaxis in surgery. Am J Health Syst Pharm. 27. Wong GK, Poon WS, Ng SC, Ip M. The impact of ventricular catheter
2013;70(3):195–283. impregnated with antimicrobial agents on infections in patients with
5. Moher D, Liberati A, Tetzlaff J, Altman DG, Group, P. Preferred reporting ventricular catheter: interim report. Acta Neurochir Suppl. 2008;102:53–5.
items for systematic reviews and meta-analyses: the PRISMA statement. J 28. Zabramski JM, Whiting D, Darouiche RO, et al. Efficacy of antimicrobial-
Clin Epidemiol. 2009;62(10):1006–12. impregnated external ventricular drain catheters: a prospective, rand-
6. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus omized, controlled trial. J Neurosurg. 2003;98(4):725–30.
on rating quality of evidence and strength of recommendations. BMJ. 29. Gutierrez-Gonzalez R, Boto GR, Fernandez-Perez C, del Prado N. Protective
2008;336(7650):924–6. effect of rifampicin and clindamycin impregnated devices against Staphy-
7. Le C, Guppy KH, Axelrod YV, et al. Lower complication rates for cranio- lococcus spp. infection after cerebrospinal fluid diversion procedures.
plasty with peri-operative bundle. Clin Neurol Neurosurg. 2014;120:41–4. BMC Neurol. 2010;10:93.
8. Hale M, Coppa N, Dogan A, Townes J. A multi-disciplinary performance 30. Wright K, Young P, Brickman C, et al. Rates and determinants of
improvement project to reduce craniotomy surgical site infections. Am J ventriculostomy-related infections during a hospital transition to use
Infect Control. 2012;40(5):e47. of antibiotic-coated external ventricular drains. Neurosurg Focus.
9. Adogwa O, Fatemi P, Perez E, et al. Negative pressure wound therapy 2013;34(5):E12.
reduces incidence of postoperative wound infection and dehiscence 31. Mikhaylov Y, Wilson TJ, Rajajee V, et al. Efficacy of antibiotic-impregnated
after long-segment thoracolumbar spinal fusion: a single institutional external ventricular drains in reducing ventriculostomy-associated infec-
experience. Spine J. 2014;14(12):2911–7. tions. J Clin Neurosci. 2014;21(5):765–8.
10. Bader MK, Littlejohns L, Palmer S. Ventriculostomy and intracranial pres- 32. Verberk JDM, Berkelbach van der Sprenkel JW, Arts MP, et al. Preventing
sure monitoring: in search of a 0% infection rate. Heart Lung J Crit Care. ventriculostomy-related infections with antibiotic-impregnated drains in
1995;24(2):166–72. hospitals: a two-centre Dutch study. J Hosp Infect. 2016;92(4):401–4.
11. Korinek AM, Reina M, Boch AL, et al. Prevention of external ventricular 33. Shekhar H, Kalsi P, Dambatta S, Strachan R. Do antibiotic-impregnated
drain–related ventriculitis. Acta Neurochir (Wien). 2005;147(1):39–45 external ventriculostomy catheters have a low infection rate in clinical
discussion-6. practice? A retrospective cohort study. Br J Neurosurg. 2016;30(1):64–9.
12. Dasic D, Hanna SJ, Bojanic S, Kerr RSC. External ventricular drain infection: 34. Abla AA, Zabramski JM, Jahnke HK, Fusco D, Nakaji P. Comparison of two
the effect of a strict protocol on infection rates and a review of the litera- antibiotic-impregnated ventricular catheters: a prospective sequential
ture. Br J Neurosurg. 2006;20(5):296–300. series trial. Neurosurgery. 2011;68(2):437–42 discussion 42.
13. Harrop JS, Sharan AD, Ratliff J, et al. Impact of a standardized protocol 35. Fichtner J, Guresir E, Seifert V, Raabe A. Efficacy of silver-bearing external
and antibiotic-impregnated catheters on ventriculostomy infection rates ventricular drainage catheters: a retrospective analysis. J Neurosurg.
in cerebrovascular patients. Neurosurgery. 2010;67(1):187–91 discussion 2010;112(4):840–6.
91. 36. Wong GKC, Poon WS, Lyon D, Wai S. Cefepime vs. Ampicillin/Sulbactam
14. Honda H, Jones JC, Craighead MC, et al. Reducing the incidence of intra- and Aztreonam as antibiotic prophylaxis in neurosurgical patients with
ventricular catheter-related ventriculitis in the neurology-neurosurgical external ventricular drain: result of a prospective randomized controlled
intensive care unit at a tertiary care center in St Louis, Missouri: an 8-year clinical trial. J Clin Pharm Ther. 2006;31(3):231–5.
follow-up study. Infect Control Hosp Epidemiol. 2010;31(10):1078–81. 37. Poon WS, Ng S, Wai S. CSF antibiotic prophylaxis for neurosurgical
15. Leverstein-van Hall MA, Hopmans TE, van der Sprenkel JW, et al. A bundle patients with ventriculostomy: a randomised study. Acta Neurochir
approach to reduce the incidence of external ventricular and lumbar Suppl. 1998;71:146–8.
drain-related infections. J Neurosurg. 2010;112(2):345–53. 38. Alleyne CH Jr, Hassan M, Zabramski JM, et al. The efficacy and cost of
16. Amini S, Fauerbach L, Archibald L, Friedman W, Layon A. Ventriculostomy- prophylactic and periprocedural antibiotics in patients with external
placement bundle and decreased ventricular infections: a single institu- ventricular drains. Neurosurgery. 2000;47(5):1124–9.
tion’s experience. Crit Care Med. 2011;39:139. 39. Murphy RKJ, Liu B, Srinath A, et al. No additional protection against
17. Rahman M, Whiting JH, Fauerbach LL, Archibald L, Friedman WA. Reduc- ventriculitis with prolonged systemic antibiotic prophylaxis for patients
ing ventriculostomy-related infections to near zero: the eliminating treated with antibiotic-coated external ventricular drains. J Neurosurg.
ventriculostomy infection study. Jt Comm J Qual Patient Saf Jt Comm 2015;122(5):1120–6.
Resour. 2012;38(10):459–64. 40. Dellit TH, Chan JD, Fulton C, et al. Reduction in Clostridium difficile infec-
18. Kubilay Z, Amini S, Fauerbach LL, et al. Decreasing ventricular infections tions among neurosurgical patients associated with discontinuation of
through the use of a ventriculostomy placement bundle: experience at a antimicrobial prophylaxis for the duration of external ventricular drain
single institution. J Neurosurg. 2013;118(3):514–20. placement. Infect Control Hosp Epidemiol. 2014;35(5):589–90.
19. Hill M, Baker G, Carter D, et al. A multidisciplinary approach to end exter- 41. Williams TA, Leslie GD, Dobb GJ, Roberts B, van Heerden PV. Decrease
nal ventricular drain infections in the neurocritical care unit. J Neurosci in proven ventriculitis by reducing the frequency of cerebrospinal fluid
Nurs. 2012;44(4):188–93. sampling from extraventricular drains. J Neurosurg. 2011;115(5):1040–6.
20. Lwin S, Low SW, Choy DK, Yeo TT, Chou N. External ventricular drain infec- 42. Wong GKC, Poon WS, Wai S, et al. Failure of regular external ven-
tions: successful implementation of strategies to reduce infection rate. tricular drain exchange to reduce cerebrospinal fluid infection: result
Singapore Med J. 2012;53(4):255–9. of a randomised controlled trial. J Neurol Neurosurg Psychiatry.
21. Flint AC, Rao VA, Renda NC, et al. A simple protocol to prevent external 2002;73(6):759–61.
ventricular drain infections. Neurosurgery. 2013;72(6):993–9. 43. Duncan C, Laurie K, Lynch M. Reducing the frequency of external
22. Camacho EF, Boszczowski I, Freire MP, et al. Impact of an educa- ventricular drainage set changes may reduce the incidence of clinically
tional intervention implanted in a neurological intensive care unit defined ventriculitis. Aust Crit Care. 2011;24(1):69.
on rates of infection related to external ventricular drains. PLoS ONE. 44. Schodel P, Proescholdt M, Ullrich OW, Brawanski A, Schebesch KM. An
2013;8(2):e50708. outcome analysis of two different procedures of burr-hole trephine and
 external ventricular drainage in acute hydrocephalus. J Clin Neurosci. 62. Schmitz M. Preventing ventilator-associated pneumonia in the neuro-
2012;19(2):267–70. science intensive care unit: a multidisciplinary approach. Am J Infect
45. Scheithauer S, Schulze-Steinen H, Hollig A, et al. Significant reduction Control. 2013;1:S89–90.
of external ventricular drainage-associated meningoventriculitis by 63. Frattalone A, Ziai W, Fellerman D. Effect of independent monitoring and
chlorhexidine-containing dressings: a before-after trial. Clin Infect Dis. quality improvement interventions on the rate of ventilator associated
2016;62(3):404–5. pneumonia in a neurocritical care unit. Neurocrit Care. 2011;1:S264.
46. Bookland MJ, Sukul V, Connolly PJ. Use of a cyanoacrylate skin adhe- 64. Korinek AM, Laisne MJ, Nicolas MH, et al. Selective decontamination
sive to reduce external ventricular drain infection rates. J Neurosurg. of the digestive tract in neurosurgical intensive care unit patients: a
2014;121(1):189–94. double-blind, randomized, placebo-controlled study. Crit Care Med.
47. May AK, Fleming SB, Carpenter RO, et al. Influence of broad-spectrum 1993;21(10):1466–73.
antibiotic prophylaxis on intracranial pressure monitor infections and 65. Hammond JMJ, Potgieter PD. Neurologic disease requiring long-term
subsequent infectious complications in head-injured patients. Surg ventilation: the role of selective decontamination of the digestive tract in
Infect. 2006;7(5):409–17. preventing nosocomial infection. Chest. 1993;104(2):547–51.
48. Lewis A, Sen R, Hill TC, et al. Antibiotic prophylaxis for subdural and 66. Cabrera Santana M, Pena Morant V, Sanchez Ramirez C, et al. Selective
subgaleal drains. J Neurosurg. 2017;126(3):908–12. decontamination of the digestive tract (SDD) effects on nosocomial
49. Kramer AH, Roberts DJ, Zygun DA. Optimal glycemic control in neuro- infections in a neurotraumatic intensive care unit (ICU) in a tertiary-care
critical care patients: a systematic review and meta-analysis. Crit Care. hospital. Intensiv Care Med. 2013;39:S271–2.
2012;16(5):R203. 67. Seguin P, Tanguy M, Laviolle B, Tirel O, Malledant Y. Effect of oro-
50. Ignelzi RJ, VanderArk GD. Analysis of the treatment of basilar skull frac- pharyngeal decontamination by povidone-iodine on ventilator-
tures with and without antibiotics. J Neurosurg. 1975;43(6):721–6. associated pneumonia in patients with head trauma. Crit Care Med.
51. Hoff JT, Brewin AU, Letter HS. : antibiotics for basilar skull fracture. J Neuro- 2006;34(5):1514–9.
surg. 1976;44(5):649. 68. Seguin P, Laviolle B, Dahyot-Fizelier C, et al. Effect of oropharyngeal pov-
52. Klastersky J, Sadeghi M, Brihaye J. Antimicrobial prophylaxis in patients idone-iodine preventive oral care on ventilator-associated pneumonia
with rhinorrhea or otorrhea: a double-blind study. Surg Neurol. in severely brain-injured or cerebral hemorrhage patients: a multicenter,
1976;6(2):111–4. randomized controlled trial. Crit Care Med. 2014;42(1):1–8.
53. Demetriades D, Charalambides D, Lakhoo M, Pantanowitz D. Role of 69. Wagner C, Marchina S, Deveau JA, et al. Risk of stroke-associated pneu-
prophylactic antibiotics in open and basilar fractures of the skull: a rand- monia and oral hygiene. Cerebrovasc Dis. 2016;41(1–2):35–9.
omized study. Injury. 1992;23(6):377–80. 70. Fukunaga A, Naritaka H, Fukaya R, Tabuse M, Nakamura T. Our method of
54. Eftekhar B, Ghodsi M, Nejat F, Ketabchi E, Esmaeeli B. Prophylactic povidone-iodine ointment and gauze dressings reduced catheter-related
administration of ceftriaxone for the prevention of meningitis after infection in serious cases. Dermatology. 2006;212(Suppl 1):47–52.
traumatic pneumocephalus: results of a clinical trial. J Neurosurg. 71. Fukunaga A, Naritaka H, Fukaya R, Tabuse M, Nakamura T. Povidone-
2004;101(5):757–61. iodine ointment and gauze dressings associated with reduced catheter-
55. Ratilal BO, Costa J, Pappamikail L, Sampaio C. Antibiotic prophylaxis for related infection in seriously ill neurosurgical patients. Infect Control
preventing meningitis in patients with basilar skull fractures. Cochrane Hosp Epidemiol. 2004;25(8):696–8.
Database Syst Rev. 2015;4:CD004884. 72. Elsayed A, Mahanes D, Nathan B, Gress D. Prevention of catheter-related
56. Sirvent JM, Torres A, El-Ebiary M, et al. Protective effect of intravenously blood stream infection in the neurointensive care unit. Neurocrit Care.
administered cefuroxime against nosocomial pneumonia in patients with 2010;13:S140.
structural coma. Am J Respir Crit Med. 1997;155(5):1729–34. 73. Wisniewski P, Mulatre M, Ibrahim J, Ashworth S, Aguirre L. Decrease in
57. Acquarolo A, Urli T, Perone G, et al. Antibiotic prophylaxis of early onset CAUTI rate following adoption of new protocols in the ICU. Crit Care Med.
pneumonia in critically ill comatose patients. A randomized study. Inten- 2013;1:A277.
siv Care Med. 2005;31(4):510–6. 74. Samuel S, Bertin M, Rasmussen P, Manno E, Frontera J. Implementation of
58. Valles J, Peredo R, Burgueno MJ, et al. Efficacy of single-dose antibiotic cauti prevention protocol in the Neuro ICU lowers CAUTI rates and length
against early-onset pneumonia in comatose patients who are ventilated. of stay. Crit Care Med. 2014;1:A1550.
Chest. 2013;143(5):1219–25. 75. Patel S, Ibrahim J, Smith C, Safcsak K, Ashworth S. Strategies to reduce
59. Schwarz S, Al-Shajlawi F, Sick C, Meairs S, Hennerici MG. Effects of prophy- urinary catheter usage and rate of catheter-associated urinary tract infec-
lactic antibiotic therapy with mezlocillin plus sulbactam on the incidence tions. Crit Care Med. 2014;1:A1552.
and height of fever after severe acute ischemic stroke: the Mannheim 76. Lusby M, Williams MH, Blaber B, et al. A multi-faceted program to reduce
infection in stroke study (MISS). Stroke. 2008;39(4):1220–7. the rates of catheter-associated urinary tract infections. Neurocrit Care.
60. Harms H, Prass K, Meisel C, et al. Preventive antibacterial therapy in acute 2014;1:S41.
ischemic stroke: a randomized controlled trial. PLoS ONE. 2008;3(5):e2158. 77. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. Preferred
61. Weireter LJ Jr, Collins JN, Britt RC, et al. Impact of a monitored pro- reporting items for systematic reviews and meta-analyses: the PRISMA
gram of care on incidence of ventilator-associated pneumonia: results statement. PLoS Med. 2009;6(7):e1000097. https​://doi.org/10.1371/journ​
of a longterm performance-improvement project. J Am Coll Surg. al.pmed1​00009​7.
2009;208(5):700–4.