Clinical Neurology and Neurosurgery 116 (2014) 13–19
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Review
Third-generation cephalosporins as antibiotic prophylaxis in neurosurgery: What’s the evidence? Weiming Liu a,c,∗,1 , Marian Christoph Neidert b,1 , Rob J.M. Groen c , Christoph Michael Woernle b , Hajo Grundmann d a
Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Tiantan Xili 6, 100050 Beijing, China Department of Neurosurgery, University Hospital Zurich, Frauenklinikstrasse 10, CH-8091 Zurich, Switzerland c Department of Neurosurgery, University Medical Center Groningen, University of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands d Department of Medical Microbiology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands b
a r t i c l e
i n f o
Article history: Received 6 April 2013 Received in revised form 17 October 2013 Accepted 27 October 2013 Available online 1 November 2013 Keywords: Antibiotic prophylaxis Neurosurgery Meta-analysis Surgical site infections Third-generation cephalosporins
a b s t r a c t To analyze the role of third-generation cephalosporins as prophylactic antibiotics in neurosurgery. We reviewed the literature for data from randomized controlled trials (RCTs) on third-generation cephalosporins compared to other antibiotic regimen in neurosurgery. End point of the RCTs was the occurrence of surgical site infections (SSIs) – data were pooled in a fixed-effects meta-analysis. Five randomized controlled trials enrolling a total of 2209 patients were identified. The pooled odds ratio for SSIs (overall) with third-generation cephalosporins prophylaxis in the five RCTs was 0.94 (95% CI, 0.59–1.52; P = 0.81). No significant difference between third-generation cephalosporins and alternative regimen was identified. When analyzing organ SSIs (osteomyelitis, meningitis, and others intracranial infections) in data derived from four RCTs (1596 patients), third-generation cephalosporins failed to show superiority (pooled odds ratio 0.88; 95% CI 0.45–1.74; P = 0.72). Third-generation cephalosporin antibiotic prophylaxis fails to show superiority over conventional regimens regarding both incisional and organ related SSIs in neurosurgery. © 2013 Elsevier B.V. All rights reserved.
Contents 1. 2.
3.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials and methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Types of interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Types of outcome measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Data collection and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6. Subgroup analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Included trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Risk of bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Third-generation cephalosporins compared to conventional antibiotic prophylaxis regarding the overall rate of SSIs . . . . . . . . . . . . . . . . . . . 3.4. Third-generation cephalosporins compared to conventional antibiotics regarding the rate of organ SSIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Insufficient data for additional analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. Adverse events related to antibiotic prophylaxis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7. Publication bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14 14 14 14 14 14 14 15 15 15 15 16 16 16 16 16
∗ Corresponding author at: Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Tiantan Xili 6, 100050 Beijing, China. Tel.: +86 1067096512; fax: +86 1067096523. E-mail address: dr [email protected] (W. Liu). 1 These authors contributed equally to this paper. 0303-8467/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.clineuro.2013.10.015
14
4. 5.
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Although surgical site infections (SSIs) are not common in neurosurgery, 1–5% of craniotomies [1] and 3% of spinal procedures [2] are complicated either by infections at the incision site (incisional infections) or even involve the central nervous system and its surrounding structures (organ infections). For cerebrospinal fluid (CSF) shunt insertions infection rates range between 1.5% and 38% [3]. While all SSIs increase morbidity, mortality, length of hospital stay, and cost [4], postoperative infections in neurosurgery are associated with a particular high burden of disease [5]. Therefore, antibiotic prophylaxis is widely considered as an option to reduce SSIs in neurosurgical procedures. Since clinical trials determined the effectiveness of antibiotic prophylaxis given shortly before the start of the operation, most guidelines recommend the perioperative administration of antibiotics [6–8]. Cefazolin, a first-generation cephalosporin, has been originally advocated for its adequate coverage for clean and cleancontaminated operations [9]. It has a good safety profile, acceptable pharmacokinetics, and reasonable cost per dose [10]. Since a change in the spectrum toward more Gram-negative bacteria with broader antibiotic resistance has been observed, experts suggested to use extended spectrum beta-lactam antibiotics [11]. Thirdgeneration cephalosporins have been put forward for their better Gram-negative coverage and favorable pharmacokinetic and pharmacodynamic properties. Thus, third-generation cephalosporins achieve CSF concentrations that are expected to be well above the minimum inhibitory concentrations of the most frequent pathogens. Studies on the use of third-generation cephalosporins for antibiotic prophylaxis in neurosurgery showed good results, suggesting effective SSIs prevention with acceptable side effects [12–14]. This article will focus on randomized controlled trials comparing third-generation cephalosporins to conventional antibiotic prophylaxis regimens which had already shown efficiency in previous RCTs.
17 18 18
second-generation cephalosporins (i.e. cefazolin and cefuroxim), aminoglycosides (gentamicin), and glycopeptides (vancomycin). 2.3. Types of outcome measures We considered surgical site infections (SSIs) including superficial incisional surgical site infections, deep incisional surgical site infection and organ infection as the primary endpoint. We defined SSIs based on clinical evidence of infection and according to the CDC/NHSN surveillance definition of health care-associated infection criteria [15] as follows: Superficial incisional primary (SIP) SSI or superficial incisional secondary (SIS) SSI: involves only skin and subcutaneous tissue of the incision within 30 days after surgery. Deep incisional primary (DIP) or deep incisional secondary (DIS) SSI: involves deep soft tissues (e.g. fascial and muscle layers) of the incision within 30 days after surgery (1 year if implant is in place). Organ/space surgical site infections involve organs and organ cavities beyond the fascia and muscular layers that were opened or manipulated during the operative procedure. Infections are regarded as SSIs if occurring within 30 days after surgery or 1 year if implant is in place. In neurosurgical procedures the specific sites of organ/space SSIs include: BONE: osteomyelitis MEN: meningitis IC: intracranial infection, including brain abscess, subdural or epidural infection and encephalitis DISC: disk space We also evaluated adverse drug effects reported, including nausea, vomiting, diarrhea, allergic reactions, and others which might have led to premature discontinuation of prophylaxis and to other adverse effects in relation to the antibiotics used.
2. Materials and methods
2.4. Search strategy
2.1. Inclusion criteria We identified RCTs comparing third-generation cephalosporins to conventional antibiotic regimens in neurosurgery. Neurosurgical procedures included craniotomies, spinal procedures, ventriculoperitoneal (VP) shunts, external ventricular drains (EVDs), biopsies and stereotactic surgery. For the purpose of analyzing which antibiotic regimens are likely to be more effective and safe, we excluded placebo controlled studies as well as trials comparing just dosing or timing of a single antibiotic regimen.
We searched the CENTRAL (Cochrane Central Register of Controlled Trials, 2012, Issue 6), MEDLINE, EMBASE, LILACS (Latin American and Caribbean Center on Health Sciences Information) and CMB (Chinese Biomedical Database). The cut-off time of the retrieved documents was the end of November 2012. We used the search strategy described in detail in Table 1 to search MEDLINE and CENTRAL. We combined the MEDLINE search strategy with the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity-maximizing version [16]. We adapted the search strategy for EMBASE, LILACS and CMB.
2.2. Types of interventions
2.5. Data collection and analysis
Studies were included that compared the effectiveness of third-generation cephalosporins with conventional antibiotics. We considered third-generation cephalosporins most frequently used in clinical routine such as cefdinir, cefixime, ceftibuten, ceftriaxone, cefotaxime, ceftizoxime, or cefoperazone. The conventional regimen was regarded as perioperative administration of a single or a combination of antibiotic compounds that had already shown efficiency in previous RCTs, including broad-spectrum penicillins (ampicillin), first- or
Three authors (WL, HZ, MN) independently selected trials for inclusion. Outcomes were cross-checked – ambiguities or misinterpretations were resolved through discussion and consensus finding. Two authors (WL, HZ) independently assessed potential biases of the selected trials according to the Cochrane Handbook for Systematic Reviews of Interventions [16] based on six domains, (i) random sequence generation, (ii) allocation concealment, (iii) blinding, (iv) incomplete outcome data, (v) selective reporting, and
Rationale
(vi) others, and rated the trials according to their findings as (i) low risk of bias, (ii) high risk of bias and (iii) uncertain risk of bias. We also considered the prognostic equivalence between patients in the two treatment arms and the completeness and length of follow-up. We assessed heterogeneity in all analyses with an I2 statistic value of ≥50% taken to indicate statistical heterogeneity. We analyzed the data using ReviewManager 5.1 [17] and calculated the summary estimates using the fixed-effect model. We conducted a visual inspection of the funnel plot of the studies for any obvious asymmetry that could indicate publication bias.
15
Less toxicity, better CSF penetration, and Gram-negative coverage Higher CSF and tissue concentration to MIC ratio Higher CSF and tissue concentration to MIC ratio, longer half-life, minimal adverse effects Improved coverage of beta-lactamase producers Higher CSF and tissue concentration to MIC ratio
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
Ceftriaxone Trimethoprim-sulfamathoxazole At least 1 year 107 VP shunt
2000
2001
2007
2008
Whitby
Zhu
Erman
Nejat
[25]
Cefoperazone–sulbactam Cefazoline At least 2 weeks 483 Cranial, spinal, burr-hole, shunt
1993 Pons
[11]
Ceftriaxone 6 weeks Craniotomy, burr-hole, shunt, transsphenoidal, spinal
180
Ampicillin-sulbactam
Cefotaxime Trimethoprim-sulfamethoxazole 6 weeks 613
Vancomycin and gentamicin 3 months
Year Author
17 brain abscess* tw 18 cerebral abscess* tw 19 exp encephalitis/ 20 brain inflammation* 21 encephaliti* 22 spin* infection 23 or/8–22 24 exp neurosurgery/ 25 neurosurg* tw 26 exp neurosurgical procedures/ 27 neurosurgical procedur* tw 28 exp craniotomy/ 29 craniotom* tw 30 brain operation* tw 31 or/24–30 32 7 and 23 and 31
Table 2 Details of the 5 included trials.
1 exp anti-bacterial agents/ 2 antibiotic*.tw. 3 antimicrob*.tw. 4 antibacter*.tw 5 bacteriocid*.tw 6 antimycobacter*.tw 7 or/1–6 8 exp central nervous system infections/ 9 exp infection/ 10 infect*.tw. 11 exp meningitis/ 12 meningit*.tw. 13 pachymeningit* 14 epidural infection* 15 subdural infection* 16 exp brain abscess/
Ref
Table 1 Search strategies in MEDLINE and CENTRAL.
[23]
Procedures
In all trials, proper randomization was achieved by means of computer-generated lists. Concealment of patient allocation to different treatment arms of patients or clinicians (blinding) was not mentioned for most trials. Blinding is important for outcome assessment, but it is generally difficult to perform [20] and to evaluate [21] in surgical procedures. For example, the assessment
Craniotomy, burr-hole, shunt
3.2. Risk of bias
[22]
Cases
Follow up time
Our search strategy identified 45 clinical trials. Of these, 15 non-randomized trials were excluded leaving 30 randomized control trials (RCTs). Of these 30 RCTs, 17 compared the intervention with placebo, 3 compared dosing or the route of administration of a single antibiotic, and 3 assessed antimicrobial impregnated ventricular catheters compared to conventional catheters. According to the inclusion criteria mentioned above, we excluded these trials. Seven RCTs compared different antibiotic regimens. Two more trials were excluded because they compared cefepime (fourth-generation cephalosporin) [18] and cefuroxime (secondgeneration cephalosporin) [19] respectively. Thus, a total of 5 RCTs comparing conventional antibiotic regimen to third-generation cephalosporins could be included in this analysis, details are shown in Table 2.
826
3.1. Included trials
Cranial, spinal, transsphenoidal
Conventional antibiotic regimen
3. Results
[26]
Third-generation cephalosporins
We examined the outcome of different types of SSIs (overall, superficial incisional SSIs, deep incisional SSIs, organ SSIs) according to the previous definitions. We conducted subgroup analyses based on the types of surgical procedures, for example craniotomy, shunt, and spinal procedure. The type of bacteria such as Grampositive bacteria, Gram-negative bacteria, and methicillin-resistant Staphylococcus aureus (MRSA) was also considered.
Ceftizoxime
2.6. Subgroup analyses
16
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
(3.2%) receiving third-generation cephalosporin prophylaxis and 37 cases of postoperative infection in 1081 subjects (3.4%) receiving conventional prophylaxis. A fixed-effects meta-analysis showed a pooled OR of 0.94 (95% CI, 0.59–1.52; P = 0.81) indicating no significant benefit for third-generation cephalosporins (Fig. 1). There was no indication of heterogeneity between the results of the different trials (P = 0.98; I2 = 0%). 3.4. Third-generation cephalosporins compared to conventional antibiotics regarding the rate of organ SSIs We accept the CDC/NHSN definition of SSIs as mentioned above [15]. Organ/space surgical site infections occur within 30 days postoperatively (1 year if implant is in place). In neurosurgical procedures the specific sites of organ/space SSIs include BONE (osteomyelitis), MEN (meningitis) and IC (intracranial infection, including brain abscess, subdural or epidural infection and encephalitis). The study Whitby et al. did not differentiate between the different types of SSIs, so we did not consider this study in the evaluation of organ SSIs. The rest four trials yielded 1596 patients. There were 17 cases of postoperative organ infections among 813 patients (2.1%) receiving third-generation cephalosporin prophylaxis and 19 cases of postoperative organ infection among 783 patients (2.4%) on conventional antibiotics. A fixed-effects meta-analysis showed a pooled OR of 0.88 (95% CI, 0.45–1.74; P = 0.72), indicating no significant difference between two antibiotic regimen in protecting organ SSIs. There was also no indication of heterogeneity between the results of the different trials (P = 0.69; I2 = 0%). 3.5. Insufficient data for additional analyses
Fig. 1. Strategies to avoid the risks of bias. Green plus sign: low risk; red minus sign: high risk; blank: unclear risk in parenthesis bias addressed by respective strategy. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
of superficial skin infections is to some degree subjective to the observer. Although we accept the results of the RCTs, we must bear in mind the bias-generating consequences that result from the loss of blinding. According to the definition of SSIs, a follow-up period of at least 1 month (1 year with implant) postoperatively was regarded as crucial. There are three trials that have incomplete outcome data due to insufficient follow-up periods [11,22,23] (shown in Table 2). All three trials include shunt cases, thus a follow-up period of at least 1 year is needed because an implant was left in place. Pons’ trial [24] does not include shunt cases and has a followup period of 3 months, however it is not mentioned whether other implants were left in place. Consequently, we consider their outcome data as unclear. In one trial [25], there was no information available about presence or absence of side-effects of the drugs used. The anticipated risk of bias of the studies included was scored by the 3 reviewers of the present meta-analysis, and is summarized in Fig. 1. 3.3. Third-generation cephalosporins compared to conventional antibiotic prophylaxis regarding the overall rate of SSIs The five included RCTs enrolled a total of 2209 patients. There were 36 cases of postoperative infection among 1128 patients
The included RCTs do not have enough information for additional analyses regarding the two antibiotic regimens such as a comparison of different surgical procedures: craniotomy, shunt or spinal procedures. We also made an attempt to determine possible changes in the bacterial spectrum after the use of different antibiotics prophylaxis regimen, but even when stratifying for broad categories such as Gram-positive, Gram-negative bacteria and MRSA, the trials cannot provide sufficient precision. Hence, the small number of data available did not allow for further statistical analysis on this item (Figs. 2 and 3). 3.6. Adverse events related to antibiotic prophylaxis Information on adverse events was missing in the trial of Nejat et al. [25]. There were no allergic or adverse drug reactions in both regimen groups in Zhu [23] and Erman [11] trials. In the trial of Whitby et al. [22], adverse clinical events happened in 7.6% of the patients receiving the third-generation cephalosporin (cefotaxime) vs. 6.4% in those receiving conventional prophylaxis (trimethoprim/sulfamethoxazole). Most cases experienced minor reactions such as skin rashes. The difference was not significant. In the trial of Pons et al. [26], no adverse drug reactions to thirdgeneration cephalosporins (ceftizoxime) were ascertained, but six patients in the group receiving conventional prophylaxis (vancomycin/gentamicin) suffered from hypotension and flushing. 3.7. Publication bias Publication bias is the phenomenon that statistically significant results are more likely to be published and cited. A funnel plot was used to test for this form of bias in our meta-analysis. When an unbiased sample of trials performed is studied with a funnel plot, the observed effect sizes should range symmetrically around the true effect size, which will be most accurately estimated by the
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
17
Fig. 2. Summary of the meta-analysis of third-generation cephalosporins vs conventional antibiotics regarding the overall rate of SSIs. Each horizontal line represents the results of a single RCT. The central square marks the OR point estimate for each study and the size of the square is proportionate to the weight given to the trial. The left and right end points of the horizontal line mark the ends of the CI for the individual trial’s OR estimate. The diamond represents the pooled estimate from the meta-analysis; it’s center marks the summary OR point estimate and the tips the 95% confidence intervals.
Fig. 3. Summary of the meta-analysis of randomized trials on third-generation cephalosporins vs conventional antibiotics regarding the rate of organ SSIs. Each horizontal line represents the results of a single study. Each horizontal line represents the results of a single RCT. The central square marks the OR point estimate for each study and the size of the square is proportionate to the weight given to the trial. The left and right end points of the horizontal line mark the ends of the CI for the individual trial’s OR estimate. The diamond represents the pooled estimate from the meta-analysis; it’s center marks the summary OR point estimate and the tips the 95% confidence intervals.
largest trials, resulting in a symmetrical plot shaped like an inverted funnel. Since there are only five trials included in this review, the issue of publication bias needs serious attention. Fig. 4 shows that the trials included in this meta-analysis seem to be approximately symmetrical in both analyses. 4. Discussion In general, the third-generation cephalosporins used in the five included studies (cefotaxime, cefoperazone, ceftriaxone, ceftazidime or ceftizoxime) are considered to be of clinical efficacy, favorable pharmacokinetics, and low frequency of adverse effects as well as good coverage of Gram-negative organisms, especially those with narrow and broad spectrum -lactamases [27]. These
Fig. 4. A funnel plot evaluating publication bias in third-generation cephalosporins vs. conventional prophylaxis regarding the overall rate of SSIs. The plot shows that the five included RCTs in this meta-analysis seem approximately symmetrically ranged around the overall effect size estimate, shown by the vertical dashed line in the center.
features make third-generation cephalosporins the antibiotic of choice in many clinical settings. However, third-generation cephalosporins are generally not recommended for surgical prophylaxis [7,28]. Despite these recommendations, they have been accepted by the medical community and are today in use in many countries as the most common drugs in surgical prophylaxis. In neurosurgery, some studies [12–14] showed favorable results for third-generation cephalosporins as prophylactic antibiotics. The RCTs included in this analysis also aimed to test for other advantages of third-generation cephalosporins such a superior side effect profile compared to conventional antibiotics. Third-generation cephalosporins are increasingly considered for prophylaxis in neurosurgery mainly because of the observed changes in the spectrum of bacteria causing SSIs in neurosurgical patients (toward an increasing number of Gram-negative SSIs) [27,28]. This is also of particular interest since postoperative meningitis is mostly caused by Gram-negative bacteria [29,30]. Because organ SSIs in neurosurgery are associated with more serious consequences, third-generation cephalosporin prophylaxis was tested. Despite the theoretical advantages of third-generation cephalosporines for antibiotic prophylaxis such as broad bacterial coverage, efficacy in other clinical settings, favorable pharmacokinetics and dynamics, and low frequency of adverse effects, our meta-analysis indicates that on the basis of available pooled data there is no convincing superiority of third-generation cephalosporins over conventional regimes regarding the overall rate of SSIs or organ SSIs after neurosurgical procedures. The pathophysiologic mechanism of SSIs, especially regarding the organ infections, is not very clear. Most studies [30–33] focused on SSIs risk factors; they found that CSF leaks, the extent of procedure, underlying patient status (American Society of Anesthesiologists score), early re-operation, duration of surgery, as well as the number of days of external ventricular drainage and intracranial pressure monitoring were independent risk factors. Among them, CSF leaks carried the highest risk and ensuing meningitis
18
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
was predominantly caused by Gram-negative bacteria. Thus current expert opinion regards postoperative invasion of the CNS as the most likely route of infection [30,34]. In our opinion, clinical research is not sufficient to clarify this matter and experimental research is needed. Third-generation cephalosporins are erroneously regarded as favorable for their blood–brain barrier (BBB) permeability. This feature was also mentioned as one of the main reasons that this group of antibiotics was studied in the RCTs which were subsequently included in our analysis (see Table 2). However, in terms of BBB permeability third-generation cephalosporins are identical to other cephalosporins, they poorly pass the blood–brain barrier if the meninges are not inflamed a feature largely caused by their low lipophilicity [35]. Nevertheless, the minimum bactericidal concentrations (MBCs) of the extended spectrum cephalosporins for common pathogens are generally low; thus, therapeutic drug concentrations can easily be achieved in the CSF [35]. That is why third-generation cephalosporins demonstrate encouraging clinical efficacy in the treatment (not prophylaxis) of a broad range of intracranial bacterial infections. There is an “equivalence” principle in selecting antibiotic for prophylaxis. It means that there were no differences among the antibiotic regimens used [1]. So antibiotics should be selected consistent with current antibiotic knowledge principles. Therefore, antibiotic selection is based on cost, adverse-effect profile, ease of administration, pharmacokinetic profile, and antibacterial activity. In neurosurgery, Staphylococcus aureus remains a major pathogen [36] in despite of changes toward a more Gram-negative bacterial spectrum. In this setting, cefazolin (first-generation cephalosporine) is still recommended as the antibiotic of choice in recently updated guidelines [37]. Meanwhile, we must be aware that the widespread use of thirdgeneration cephalosporins has been associated with increases in extended spectrum -lactamase (ESBL)-mediated resistance amongst Gram-negative pathogens, antibiotic-associated diarrhea due to C. difficile, and MRSA and enterococci [38,39]. Therefore, there is a call to restrict the use of these agents [40]. In order to reduce the selection pressure guidelines are careful not to recommend these compounds for perioperative prophylaxis in neurosurgery. 5. Conclusions In theory, third-generation cephalosporins have unique pharmacodynamics and pharmacokinetics, which make them very appealing for the use in perioperative neurosurgical prophylaxis. However, this meta-analysis of RCTs does not support the superiority of third-generation cephalosporins over conventional antibiotics in this setting. References [1] Barker FG. Efficacy of prophylactic antibiotics for craniotomy. Neurosurgery 1994;35:484–92. [2] Abdul-Jabbar A, Takemoto S, Weber MH, Hu SS, Mummaneni PV, Deviren V, et al. Surgical site infection in spinal surgery: description of surgical and patient-based risk factors for postoperative infection using administrative claims data. Spine 2012;37:1340–5. [3] BC C. Shunt Infection. In: Winn HR, editor. Youmans neurological surgery. Philadelphia: Saunders; 2004. p. 3419–25. [4] Horan TC, Culver DH, Gaynes RP, Jarvis WR, Edwards JR, Reid CR. Nosocomial infections in surgical patients in the United States, January 1986–June 1992. National Nosocomial Infections Surveillance (NNIS) System. Infection Control and Hospital Epidemiology 1993;14:73–80. [5] O’Keeffe AB, Lawrence T, Bojanic S. Oxford craniotomy infections database: a cost analysis of craniotomy infection. British Journal of Neurosurgery 2012;26:265–9. [6] Scottish Intercollegiate Guidelines Network. Antibiotic prophylaxis in surgery. A national clinical guideline; 2008.
[7] American Society of Health-System Pharmacists. ASHP Therapeutic Guidelines on Antimicrobial Prophylaxis in Surgery. Am J Health Syst Pharm 1999;56:466–513. [8] No authors Listed. Antimicrobial prophylaxis for surgery. Treatment Guidelines from the Medical Letter 2009;7:47–52. [9] Scher KS. Studies on the duration of antibiotic administration for surgical prophylaxis. The American Surgeon 1997;63:59–62. [10] Page CP, Bohnen JM, Fletcher JR, McManus AT, Solomkin JS, Wittmann DH. Antimicrobial prophylaxis for surgical wounds. Guidelines for clinical care. Archives of Surgery (Chicago, IL: 1960) 1993;128:79–88. [11] Erman T, Yilmaz M, Demirhindi H, Arslan A. Postsurgical infection. comparative efficacy of intravenous cefoperazone/sulbactam and cefazoline in preventing surgical site infection after neurosurgery. Neurosurg Q 2007;17: 166–9. ´ [12] Bidzinski J, Barczewska M, Marchel A. Ceftriaxone (Rocephin Roche) in shortterm program of perioperational prophylaxis in neurosurgery. Neurologia i Neurochirurgia Polska 1993;27:55–61. [13] Zhao JZ, Wang S, Li JS, Yang J, Zang JT, He Q, et al. The perioperative use of ceftriaxone as infection prophylaxis in neurosurgery. Clinical Neurology and Neurosurgery 1995;97:285–9. [14] Arnaboldi L. Antimicrobial prophylaxis with ceftriaxone in neurosurgical procedures. A prospective study of 100 patients undergoing shunt operations. Chemotherapy 1996;42:384–90. [15] Horan TC, Andrus M, Dudeck M a. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. American Journal of Infection Control 2008;36:309–32. [16] Higgins JPT, Sally G. Cochrane handbook for systematic reviews of interventions, Version 5.1.0. The Cochrane Collaboration; 2011. [17] The Nordic Cochrane Centre. The Cochrane Collaboration. RevMan; 2011. [18] Wong GKC, Poon WS, Lyon D, Wai S. Cefepime vs. ampicillin/sulbactam and aztreonam as antibiotic prophylaxis in neurosurgical patients with external ventricular drain: result of a prospective randomized controlled clinical trial. Journal of Clinical Pharmacy and Therapeutics 2006;31:231–5. [19] Tai CC, Want S, Quraishi NA, Batten J, Kalra M, Hughes SPF. Antibiotic prophylaxis in surgery of the intervertebral disc. A comparison between gentamicin and cefuroxime. Journal of Bone and Joint Surgery British Volume 2002;84:1036–9. [20] Gurusamy K, Gluud C, Nikolova D, Davidson B. Assessment of risk of bias in randomized clinical trials in surgery. Br J Surg 2009;96:342–9. [21] Sackett DL. Commentary: Measuring the success of blinding in RCTs: don’t, must, can’t or needn’t? International Journal of Epidemiology 2007;36: 664–5. [22] Whitby M, Johnson BC, Atkinson RL, Stuart G. The comparative efficacy of intravenous cefotaxime and trimethoprim/sulfamethoxazole in preventing infection after neurosurgery: a prospective, randomized study. British Journal of Neurosurgery 2000;14:13–8. [23] Zhu XL, Wong WK, Yeung WM, Mo P, Tsang CS, Pang KH, et al. A randomized, double-blind comparison of ampicillin/sulbactam and ceftriaxone in the prevention of surgical-site infections after neurosurgery. Clinical Therapeutics 2001;23:1281–91. [24] Pons VG, Denlinger SL, Guglielmo BJ, Octavio J, Flaherty J, Derish PA, et al. Ceftizoxime versus vancomycin and gentamicin in neurosurgical prophylaxis. Neurosurgery 1993;33:537. [25] Nejat F, Tajik P, Khashab MEl, Kazmi SS, Khotaei GT. A randomized trial of ceftriaxone versus trimethoprim-sulfamethoxazole to prevent ventriculoperitoneal shunt infection. J Microbiol Immunol Infect 2008;41:112–7. [26] Pons VG, Denlinger SL, Guglielmo BJ, Octavio J, Flaherty J, Derish PA, et al. Ceftizoxime versus vancomycin and gentamicin in neurosurgical prophylaxis: a randomized, prospective, blinded clinical study. Neurosurgery 1993;33:416–22 [discussion 422–3]. [27] Page MGP. Cephalosporins in clinical development. Expert Opinion on Investigational Drugs 2004;13:973–85. [28] Institute for Clinical Systems Improvement. Antibiotic Prophylaxis for Surgical Site Infection Prevention in Adults Fourth Edition; 2010. ˜ [29] Núnez-Pereira S, Pellisé F, Rodríguez-Pardo D, Pigrau C, Sánchez JM, Bagó J, et al. Individualized antibiotic prophylaxis reduces surgical site infections by Gram-negative bacteria in instrumented spinal surgery. European Spine Journal 2011;20(Suppl. 3):397–402. [30] Reichert MCF, Medeiros EAS, Ferraz FAP. Hospital-acquired meningitis in patients undergoing craniotomy: incidence, evolution, and risk factors. American Journal of Infection Control 2002;30:158–64. [31] Korinek A-M, Baugnon T, Golmard J-L, van Effenterre R, Coriat PPL. Risk factors for adult nosocomial meningitis after craniotomy: role of antibiotic prophylaxis. Neurosurgery 2006;62(Suppl. 2):532–9. [32] Kourbeti IS, Jacobs AV, Koslow M, Karabetsos D, Holzman RS. Risk factors associated with postcraniotomy meningitis. Neurosurgery 2007;60:317–25 [discussion 325–6]. [33] Korinek A-M, Golmard J-L, Elcheick A, Bismuth R, Van Effenterre R, Coriat P, et al. Risk factors for neurosurgical site infections after craniotomy: a critical reappraisal of antibiotic prophylaxis on 4,578 patients. British Journal of Neurosurgery 2005;19:155–62. [34] McGirt MJ, Zaas A, Fuchs HE, George TM, Kaye K, Sexton DJ. Risk factors for pediatric ventriculoperitoneal shunt infection and predictors of infectious pathogens. Clinical Infectious Diseases 2003;36:858–62. [35] Christ W. Pharmacological properties of cephalosporins. Infection 1991;19(Suppl. 5):S244–52.
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19 [36] Fry DE, Barie PS. The changing face of Staphylococcus aureus: a continuing surgical challenge. Surgical Infections 2011;12:191–203. [37] Bratzler DW, Dellinger EP, Olsen KM, Perl TM, Auwaerter PG, Bolon MK, et al. Clinical practice guidelines for antimicrobial prophylaxis in surgery. American Journal of Health-system Pharmacy 2013;70:195–283. [38] Wilcox MH. The tide of antimicrobial resistance and selection. International Journal of Antimicrobial Agents 2009;34(Suppl. 3):S6–10.
19
[39] Kernodle DS, Classen DC, Burke JP, Kaiser AB. Failure of cephalosporins to prevent Staphylococcus aureus surgical wound infections. JAMA 1990;263: 961–6. [40] Lee S-O, Lee ES, Park SY, Kim S-Y, Seo Y-H, Cho YK. Reduced use of thirdgeneration cephalosporins decreases the acquisition of extended-spectrum beta-lactamse-producing Klebsiella pneumoniae. Infection Control and Hospital Epidemiology 2004;25:832–7.
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Review
Third-generation cephalosporins as antibiotic prophylaxis in neurosurgery: What’s the evidence? Weiming Liu a,c,∗,1 , Marian Christoph Neidert b,1 , Rob J.M. Groen c , Christoph Michael Woernle b , Hajo Grundmann d a
Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Tiantan Xili 6, 100050 Beijing, China Department of Neurosurgery, University Hospital Zurich, Frauenklinikstrasse 10, CH-8091 Zurich, Switzerland c Department of Neurosurgery, University Medical Center Groningen, University of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands d Department of Medical Microbiology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands b
a r t i c l e
i n f o
Article history: Received 6 April 2013 Received in revised form 17 October 2013 Accepted 27 October 2013 Available online 1 November 2013 Keywords: Antibiotic prophylaxis Neurosurgery Meta-analysis Surgical site infections Third-generation cephalosporins
a b s t r a c t To analyze the role of third-generation cephalosporins as prophylactic antibiotics in neurosurgery. We reviewed the literature for data from randomized controlled trials (RCTs) on third-generation cephalosporins compared to other antibiotic regimen in neurosurgery. End point of the RCTs was the occurrence of surgical site infections (SSIs) – data were pooled in a fixed-effects meta-analysis. Five randomized controlled trials enrolling a total of 2209 patients were identified. The pooled odds ratio for SSIs (overall) with third-generation cephalosporins prophylaxis in the five RCTs was 0.94 (95% CI, 0.59–1.52; P = 0.81). No significant difference between third-generation cephalosporins and alternative regimen was identified. When analyzing organ SSIs (osteomyelitis, meningitis, and others intracranial infections) in data derived from four RCTs (1596 patients), third-generation cephalosporins failed to show superiority (pooled odds ratio 0.88; 95% CI 0.45–1.74; P = 0.72). Third-generation cephalosporin antibiotic prophylaxis fails to show superiority over conventional regimens regarding both incisional and organ related SSIs in neurosurgery. © 2013 Elsevier B.V. All rights reserved.
Contents 1. 2.
3.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials and methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Types of interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Types of outcome measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Data collection and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6. Subgroup analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Included trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Risk of bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Third-generation cephalosporins compared to conventional antibiotic prophylaxis regarding the overall rate of SSIs . . . . . . . . . . . . . . . . . . . 3.4. Third-generation cephalosporins compared to conventional antibiotics regarding the rate of organ SSIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Insufficient data for additional analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. Adverse events related to antibiotic prophylaxis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7. Publication bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14 14 14 14 14 14 14 15 15 15 15 16 16 16 16 16
∗ Corresponding author at: Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Tiantan Xili 6, 100050 Beijing, China. Tel.: +86 1067096512; fax: +86 1067096523. E-mail address: dr [email protected] (W. Liu). 1 These authors contributed equally to this paper. 0303-8467/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.clineuro.2013.10.015
14
4. 5.
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Although surgical site infections (SSIs) are not common in neurosurgery, 1–5% of craniotomies [1] and 3% of spinal procedures [2] are complicated either by infections at the incision site (incisional infections) or even involve the central nervous system and its surrounding structures (organ infections). For cerebrospinal fluid (CSF) shunt insertions infection rates range between 1.5% and 38% [3]. While all SSIs increase morbidity, mortality, length of hospital stay, and cost [4], postoperative infections in neurosurgery are associated with a particular high burden of disease [5]. Therefore, antibiotic prophylaxis is widely considered as an option to reduce SSIs in neurosurgical procedures. Since clinical trials determined the effectiveness of antibiotic prophylaxis given shortly before the start of the operation, most guidelines recommend the perioperative administration of antibiotics [6–8]. Cefazolin, a first-generation cephalosporin, has been originally advocated for its adequate coverage for clean and cleancontaminated operations [9]. It has a good safety profile, acceptable pharmacokinetics, and reasonable cost per dose [10]. Since a change in the spectrum toward more Gram-negative bacteria with broader antibiotic resistance has been observed, experts suggested to use extended spectrum beta-lactam antibiotics [11]. Thirdgeneration cephalosporins have been put forward for their better Gram-negative coverage and favorable pharmacokinetic and pharmacodynamic properties. Thus, third-generation cephalosporins achieve CSF concentrations that are expected to be well above the minimum inhibitory concentrations of the most frequent pathogens. Studies on the use of third-generation cephalosporins for antibiotic prophylaxis in neurosurgery showed good results, suggesting effective SSIs prevention with acceptable side effects [12–14]. This article will focus on randomized controlled trials comparing third-generation cephalosporins to conventional antibiotic prophylaxis regimens which had already shown efficiency in previous RCTs.
17 18 18
second-generation cephalosporins (i.e. cefazolin and cefuroxim), aminoglycosides (gentamicin), and glycopeptides (vancomycin). 2.3. Types of outcome measures We considered surgical site infections (SSIs) including superficial incisional surgical site infections, deep incisional surgical site infection and organ infection as the primary endpoint. We defined SSIs based on clinical evidence of infection and according to the CDC/NHSN surveillance definition of health care-associated infection criteria [15] as follows: Superficial incisional primary (SIP) SSI or superficial incisional secondary (SIS) SSI: involves only skin and subcutaneous tissue of the incision within 30 days after surgery. Deep incisional primary (DIP) or deep incisional secondary (DIS) SSI: involves deep soft tissues (e.g. fascial and muscle layers) of the incision within 30 days after surgery (1 year if implant is in place). Organ/space surgical site infections involve organs and organ cavities beyond the fascia and muscular layers that were opened or manipulated during the operative procedure. Infections are regarded as SSIs if occurring within 30 days after surgery or 1 year if implant is in place. In neurosurgical procedures the specific sites of organ/space SSIs include: BONE: osteomyelitis MEN: meningitis IC: intracranial infection, including brain abscess, subdural or epidural infection and encephalitis DISC: disk space We also evaluated adverse drug effects reported, including nausea, vomiting, diarrhea, allergic reactions, and others which might have led to premature discontinuation of prophylaxis and to other adverse effects in relation to the antibiotics used.
2. Materials and methods
2.4. Search strategy
2.1. Inclusion criteria We identified RCTs comparing third-generation cephalosporins to conventional antibiotic regimens in neurosurgery. Neurosurgical procedures included craniotomies, spinal procedures, ventriculoperitoneal (VP) shunts, external ventricular drains (EVDs), biopsies and stereotactic surgery. For the purpose of analyzing which antibiotic regimens are likely to be more effective and safe, we excluded placebo controlled studies as well as trials comparing just dosing or timing of a single antibiotic regimen.
We searched the CENTRAL (Cochrane Central Register of Controlled Trials, 2012, Issue 6), MEDLINE, EMBASE, LILACS (Latin American and Caribbean Center on Health Sciences Information) and CMB (Chinese Biomedical Database). The cut-off time of the retrieved documents was the end of November 2012. We used the search strategy described in detail in Table 1 to search MEDLINE and CENTRAL. We combined the MEDLINE search strategy with the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity-maximizing version [16]. We adapted the search strategy for EMBASE, LILACS and CMB.
2.2. Types of interventions
2.5. Data collection and analysis
Studies were included that compared the effectiveness of third-generation cephalosporins with conventional antibiotics. We considered third-generation cephalosporins most frequently used in clinical routine such as cefdinir, cefixime, ceftibuten, ceftriaxone, cefotaxime, ceftizoxime, or cefoperazone. The conventional regimen was regarded as perioperative administration of a single or a combination of antibiotic compounds that had already shown efficiency in previous RCTs, including broad-spectrum penicillins (ampicillin), first- or
Three authors (WL, HZ, MN) independently selected trials for inclusion. Outcomes were cross-checked – ambiguities or misinterpretations were resolved through discussion and consensus finding. Two authors (WL, HZ) independently assessed potential biases of the selected trials according to the Cochrane Handbook for Systematic Reviews of Interventions [16] based on six domains, (i) random sequence generation, (ii) allocation concealment, (iii) blinding, (iv) incomplete outcome data, (v) selective reporting, and
Rationale
(vi) others, and rated the trials according to their findings as (i) low risk of bias, (ii) high risk of bias and (iii) uncertain risk of bias. We also considered the prognostic equivalence between patients in the two treatment arms and the completeness and length of follow-up. We assessed heterogeneity in all analyses with an I2 statistic value of ≥50% taken to indicate statistical heterogeneity. We analyzed the data using ReviewManager 5.1 [17] and calculated the summary estimates using the fixed-effect model. We conducted a visual inspection of the funnel plot of the studies for any obvious asymmetry that could indicate publication bias.
15
Less toxicity, better CSF penetration, and Gram-negative coverage Higher CSF and tissue concentration to MIC ratio Higher CSF and tissue concentration to MIC ratio, longer half-life, minimal adverse effects Improved coverage of beta-lactamase producers Higher CSF and tissue concentration to MIC ratio
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
Ceftriaxone Trimethoprim-sulfamathoxazole At least 1 year 107 VP shunt
2000
2001
2007
2008
Whitby
Zhu
Erman
Nejat
[25]
Cefoperazone–sulbactam Cefazoline At least 2 weeks 483 Cranial, spinal, burr-hole, shunt
1993 Pons
[11]
Ceftriaxone 6 weeks Craniotomy, burr-hole, shunt, transsphenoidal, spinal
180
Ampicillin-sulbactam
Cefotaxime Trimethoprim-sulfamethoxazole 6 weeks 613
Vancomycin and gentamicin 3 months
Year Author
17 brain abscess* tw 18 cerebral abscess* tw 19 exp encephalitis/ 20 brain inflammation* 21 encephaliti* 22 spin* infection 23 or/8–22 24 exp neurosurgery/ 25 neurosurg* tw 26 exp neurosurgical procedures/ 27 neurosurgical procedur* tw 28 exp craniotomy/ 29 craniotom* tw 30 brain operation* tw 31 or/24–30 32 7 and 23 and 31
Table 2 Details of the 5 included trials.
1 exp anti-bacterial agents/ 2 antibiotic*.tw. 3 antimicrob*.tw. 4 antibacter*.tw 5 bacteriocid*.tw 6 antimycobacter*.tw 7 or/1–6 8 exp central nervous system infections/ 9 exp infection/ 10 infect*.tw. 11 exp meningitis/ 12 meningit*.tw. 13 pachymeningit* 14 epidural infection* 15 subdural infection* 16 exp brain abscess/
Ref
Table 1 Search strategies in MEDLINE and CENTRAL.
[23]
Procedures
In all trials, proper randomization was achieved by means of computer-generated lists. Concealment of patient allocation to different treatment arms of patients or clinicians (blinding) was not mentioned for most trials. Blinding is important for outcome assessment, but it is generally difficult to perform [20] and to evaluate [21] in surgical procedures. For example, the assessment
Craniotomy, burr-hole, shunt
3.2. Risk of bias
[22]
Cases
Follow up time
Our search strategy identified 45 clinical trials. Of these, 15 non-randomized trials were excluded leaving 30 randomized control trials (RCTs). Of these 30 RCTs, 17 compared the intervention with placebo, 3 compared dosing or the route of administration of a single antibiotic, and 3 assessed antimicrobial impregnated ventricular catheters compared to conventional catheters. According to the inclusion criteria mentioned above, we excluded these trials. Seven RCTs compared different antibiotic regimens. Two more trials were excluded because they compared cefepime (fourth-generation cephalosporin) [18] and cefuroxime (secondgeneration cephalosporin) [19] respectively. Thus, a total of 5 RCTs comparing conventional antibiotic regimen to third-generation cephalosporins could be included in this analysis, details are shown in Table 2.
826
3.1. Included trials
Cranial, spinal, transsphenoidal
Conventional antibiotic regimen
3. Results
[26]
Third-generation cephalosporins
We examined the outcome of different types of SSIs (overall, superficial incisional SSIs, deep incisional SSIs, organ SSIs) according to the previous definitions. We conducted subgroup analyses based on the types of surgical procedures, for example craniotomy, shunt, and spinal procedure. The type of bacteria such as Grampositive bacteria, Gram-negative bacteria, and methicillin-resistant Staphylococcus aureus (MRSA) was also considered.
Ceftizoxime
2.6. Subgroup analyses
16
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
(3.2%) receiving third-generation cephalosporin prophylaxis and 37 cases of postoperative infection in 1081 subjects (3.4%) receiving conventional prophylaxis. A fixed-effects meta-analysis showed a pooled OR of 0.94 (95% CI, 0.59–1.52; P = 0.81) indicating no significant benefit for third-generation cephalosporins (Fig. 1). There was no indication of heterogeneity between the results of the different trials (P = 0.98; I2 = 0%). 3.4. Third-generation cephalosporins compared to conventional antibiotics regarding the rate of organ SSIs We accept the CDC/NHSN definition of SSIs as mentioned above [15]. Organ/space surgical site infections occur within 30 days postoperatively (1 year if implant is in place). In neurosurgical procedures the specific sites of organ/space SSIs include BONE (osteomyelitis), MEN (meningitis) and IC (intracranial infection, including brain abscess, subdural or epidural infection and encephalitis). The study Whitby et al. did not differentiate between the different types of SSIs, so we did not consider this study in the evaluation of organ SSIs. The rest four trials yielded 1596 patients. There were 17 cases of postoperative organ infections among 813 patients (2.1%) receiving third-generation cephalosporin prophylaxis and 19 cases of postoperative organ infection among 783 patients (2.4%) on conventional antibiotics. A fixed-effects meta-analysis showed a pooled OR of 0.88 (95% CI, 0.45–1.74; P = 0.72), indicating no significant difference between two antibiotic regimen in protecting organ SSIs. There was also no indication of heterogeneity between the results of the different trials (P = 0.69; I2 = 0%). 3.5. Insufficient data for additional analyses
Fig. 1. Strategies to avoid the risks of bias. Green plus sign: low risk; red minus sign: high risk; blank: unclear risk in parenthesis bias addressed by respective strategy. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
of superficial skin infections is to some degree subjective to the observer. Although we accept the results of the RCTs, we must bear in mind the bias-generating consequences that result from the loss of blinding. According to the definition of SSIs, a follow-up period of at least 1 month (1 year with implant) postoperatively was regarded as crucial. There are three trials that have incomplete outcome data due to insufficient follow-up periods [11,22,23] (shown in Table 2). All three trials include shunt cases, thus a follow-up period of at least 1 year is needed because an implant was left in place. Pons’ trial [24] does not include shunt cases and has a followup period of 3 months, however it is not mentioned whether other implants were left in place. Consequently, we consider their outcome data as unclear. In one trial [25], there was no information available about presence or absence of side-effects of the drugs used. The anticipated risk of bias of the studies included was scored by the 3 reviewers of the present meta-analysis, and is summarized in Fig. 1. 3.3. Third-generation cephalosporins compared to conventional antibiotic prophylaxis regarding the overall rate of SSIs The five included RCTs enrolled a total of 2209 patients. There were 36 cases of postoperative infection among 1128 patients
The included RCTs do not have enough information for additional analyses regarding the two antibiotic regimens such as a comparison of different surgical procedures: craniotomy, shunt or spinal procedures. We also made an attempt to determine possible changes in the bacterial spectrum after the use of different antibiotics prophylaxis regimen, but even when stratifying for broad categories such as Gram-positive, Gram-negative bacteria and MRSA, the trials cannot provide sufficient precision. Hence, the small number of data available did not allow for further statistical analysis on this item (Figs. 2 and 3). 3.6. Adverse events related to antibiotic prophylaxis Information on adverse events was missing in the trial of Nejat et al. [25]. There were no allergic or adverse drug reactions in both regimen groups in Zhu [23] and Erman [11] trials. In the trial of Whitby et al. [22], adverse clinical events happened in 7.6% of the patients receiving the third-generation cephalosporin (cefotaxime) vs. 6.4% in those receiving conventional prophylaxis (trimethoprim/sulfamethoxazole). Most cases experienced minor reactions such as skin rashes. The difference was not significant. In the trial of Pons et al. [26], no adverse drug reactions to thirdgeneration cephalosporins (ceftizoxime) were ascertained, but six patients in the group receiving conventional prophylaxis (vancomycin/gentamicin) suffered from hypotension and flushing. 3.7. Publication bias Publication bias is the phenomenon that statistically significant results are more likely to be published and cited. A funnel plot was used to test for this form of bias in our meta-analysis. When an unbiased sample of trials performed is studied with a funnel plot, the observed effect sizes should range symmetrically around the true effect size, which will be most accurately estimated by the
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
17
Fig. 2. Summary of the meta-analysis of third-generation cephalosporins vs conventional antibiotics regarding the overall rate of SSIs. Each horizontal line represents the results of a single RCT. The central square marks the OR point estimate for each study and the size of the square is proportionate to the weight given to the trial. The left and right end points of the horizontal line mark the ends of the CI for the individual trial’s OR estimate. The diamond represents the pooled estimate from the meta-analysis; it’s center marks the summary OR point estimate and the tips the 95% confidence intervals.
Fig. 3. Summary of the meta-analysis of randomized trials on third-generation cephalosporins vs conventional antibiotics regarding the rate of organ SSIs. Each horizontal line represents the results of a single study. Each horizontal line represents the results of a single RCT. The central square marks the OR point estimate for each study and the size of the square is proportionate to the weight given to the trial. The left and right end points of the horizontal line mark the ends of the CI for the individual trial’s OR estimate. The diamond represents the pooled estimate from the meta-analysis; it’s center marks the summary OR point estimate and the tips the 95% confidence intervals.
largest trials, resulting in a symmetrical plot shaped like an inverted funnel. Since there are only five trials included in this review, the issue of publication bias needs serious attention. Fig. 4 shows that the trials included in this meta-analysis seem to be approximately symmetrical in both analyses. 4. Discussion In general, the third-generation cephalosporins used in the five included studies (cefotaxime, cefoperazone, ceftriaxone, ceftazidime or ceftizoxime) are considered to be of clinical efficacy, favorable pharmacokinetics, and low frequency of adverse effects as well as good coverage of Gram-negative organisms, especially those with narrow and broad spectrum -lactamases [27]. These
Fig. 4. A funnel plot evaluating publication bias in third-generation cephalosporins vs. conventional prophylaxis regarding the overall rate of SSIs. The plot shows that the five included RCTs in this meta-analysis seem approximately symmetrically ranged around the overall effect size estimate, shown by the vertical dashed line in the center.
features make third-generation cephalosporins the antibiotic of choice in many clinical settings. However, third-generation cephalosporins are generally not recommended for surgical prophylaxis [7,28]. Despite these recommendations, they have been accepted by the medical community and are today in use in many countries as the most common drugs in surgical prophylaxis. In neurosurgery, some studies [12–14] showed favorable results for third-generation cephalosporins as prophylactic antibiotics. The RCTs included in this analysis also aimed to test for other advantages of third-generation cephalosporins such a superior side effect profile compared to conventional antibiotics. Third-generation cephalosporins are increasingly considered for prophylaxis in neurosurgery mainly because of the observed changes in the spectrum of bacteria causing SSIs in neurosurgical patients (toward an increasing number of Gram-negative SSIs) [27,28]. This is also of particular interest since postoperative meningitis is mostly caused by Gram-negative bacteria [29,30]. Because organ SSIs in neurosurgery are associated with more serious consequences, third-generation cephalosporin prophylaxis was tested. Despite the theoretical advantages of third-generation cephalosporines for antibiotic prophylaxis such as broad bacterial coverage, efficacy in other clinical settings, favorable pharmacokinetics and dynamics, and low frequency of adverse effects, our meta-analysis indicates that on the basis of available pooled data there is no convincing superiority of third-generation cephalosporins over conventional regimes regarding the overall rate of SSIs or organ SSIs after neurosurgical procedures. The pathophysiologic mechanism of SSIs, especially regarding the organ infections, is not very clear. Most studies [30–33] focused on SSIs risk factors; they found that CSF leaks, the extent of procedure, underlying patient status (American Society of Anesthesiologists score), early re-operation, duration of surgery, as well as the number of days of external ventricular drainage and intracranial pressure monitoring were independent risk factors. Among them, CSF leaks carried the highest risk and ensuing meningitis
18
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19
was predominantly caused by Gram-negative bacteria. Thus current expert opinion regards postoperative invasion of the CNS as the most likely route of infection [30,34]. In our opinion, clinical research is not sufficient to clarify this matter and experimental research is needed. Third-generation cephalosporins are erroneously regarded as favorable for their blood–brain barrier (BBB) permeability. This feature was also mentioned as one of the main reasons that this group of antibiotics was studied in the RCTs which were subsequently included in our analysis (see Table 2). However, in terms of BBB permeability third-generation cephalosporins are identical to other cephalosporins, they poorly pass the blood–brain barrier if the meninges are not inflamed a feature largely caused by their low lipophilicity [35]. Nevertheless, the minimum bactericidal concentrations (MBCs) of the extended spectrum cephalosporins for common pathogens are generally low; thus, therapeutic drug concentrations can easily be achieved in the CSF [35]. That is why third-generation cephalosporins demonstrate encouraging clinical efficacy in the treatment (not prophylaxis) of a broad range of intracranial bacterial infections. There is an “equivalence” principle in selecting antibiotic for prophylaxis. It means that there were no differences among the antibiotic regimens used [1]. So antibiotics should be selected consistent with current antibiotic knowledge principles. Therefore, antibiotic selection is based on cost, adverse-effect profile, ease of administration, pharmacokinetic profile, and antibacterial activity. In neurosurgery, Staphylococcus aureus remains a major pathogen [36] in despite of changes toward a more Gram-negative bacterial spectrum. In this setting, cefazolin (first-generation cephalosporine) is still recommended as the antibiotic of choice in recently updated guidelines [37]. Meanwhile, we must be aware that the widespread use of thirdgeneration cephalosporins has been associated with increases in extended spectrum -lactamase (ESBL)-mediated resistance amongst Gram-negative pathogens, antibiotic-associated diarrhea due to C. difficile, and MRSA and enterococci [38,39]. Therefore, there is a call to restrict the use of these agents [40]. In order to reduce the selection pressure guidelines are careful not to recommend these compounds for perioperative prophylaxis in neurosurgery. 5. Conclusions In theory, third-generation cephalosporins have unique pharmacodynamics and pharmacokinetics, which make them very appealing for the use in perioperative neurosurgical prophylaxis. However, this meta-analysis of RCTs does not support the superiority of third-generation cephalosporins over conventional antibiotics in this setting. References [1] Barker FG. Efficacy of prophylactic antibiotics for craniotomy. Neurosurgery 1994;35:484–92. [2] Abdul-Jabbar A, Takemoto S, Weber MH, Hu SS, Mummaneni PV, Deviren V, et al. Surgical site infection in spinal surgery: description of surgical and patient-based risk factors for postoperative infection using administrative claims data. Spine 2012;37:1340–5. [3] BC C. Shunt Infection. In: Winn HR, editor. Youmans neurological surgery. Philadelphia: Saunders; 2004. p. 3419–25. [4] Horan TC, Culver DH, Gaynes RP, Jarvis WR, Edwards JR, Reid CR. Nosocomial infections in surgical patients in the United States, January 1986–June 1992. National Nosocomial Infections Surveillance (NNIS) System. Infection Control and Hospital Epidemiology 1993;14:73–80. [5] O’Keeffe AB, Lawrence T, Bojanic S. Oxford craniotomy infections database: a cost analysis of craniotomy infection. British Journal of Neurosurgery 2012;26:265–9. [6] Scottish Intercollegiate Guidelines Network. Antibiotic prophylaxis in surgery. A national clinical guideline; 2008.
[7] American Society of Health-System Pharmacists. ASHP Therapeutic Guidelines on Antimicrobial Prophylaxis in Surgery. Am J Health Syst Pharm 1999;56:466–513. [8] No authors Listed. Antimicrobial prophylaxis for surgery. Treatment Guidelines from the Medical Letter 2009;7:47–52. [9] Scher KS. Studies on the duration of antibiotic administration for surgical prophylaxis. The American Surgeon 1997;63:59–62. [10] Page CP, Bohnen JM, Fletcher JR, McManus AT, Solomkin JS, Wittmann DH. Antimicrobial prophylaxis for surgical wounds. Guidelines for clinical care. Archives of Surgery (Chicago, IL: 1960) 1993;128:79–88. [11] Erman T, Yilmaz M, Demirhindi H, Arslan A. Postsurgical infection. comparative efficacy of intravenous cefoperazone/sulbactam and cefazoline in preventing surgical site infection after neurosurgery. Neurosurg Q 2007;17: 166–9. ´ [12] Bidzinski J, Barczewska M, Marchel A. Ceftriaxone (Rocephin Roche) in shortterm program of perioperational prophylaxis in neurosurgery. Neurologia i Neurochirurgia Polska 1993;27:55–61. [13] Zhao JZ, Wang S, Li JS, Yang J, Zang JT, He Q, et al. The perioperative use of ceftriaxone as infection prophylaxis in neurosurgery. Clinical Neurology and Neurosurgery 1995;97:285–9. [14] Arnaboldi L. Antimicrobial prophylaxis with ceftriaxone in neurosurgical procedures. A prospective study of 100 patients undergoing shunt operations. Chemotherapy 1996;42:384–90. [15] Horan TC, Andrus M, Dudeck M a. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. American Journal of Infection Control 2008;36:309–32. [16] Higgins JPT, Sally G. Cochrane handbook for systematic reviews of interventions, Version 5.1.0. The Cochrane Collaboration; 2011. [17] The Nordic Cochrane Centre. The Cochrane Collaboration. RevMan; 2011. [18] Wong GKC, Poon WS, Lyon D, Wai S. Cefepime vs. ampicillin/sulbactam and aztreonam as antibiotic prophylaxis in neurosurgical patients with external ventricular drain: result of a prospective randomized controlled clinical trial. Journal of Clinical Pharmacy and Therapeutics 2006;31:231–5. [19] Tai CC, Want S, Quraishi NA, Batten J, Kalra M, Hughes SPF. Antibiotic prophylaxis in surgery of the intervertebral disc. A comparison between gentamicin and cefuroxime. Journal of Bone and Joint Surgery British Volume 2002;84:1036–9. [20] Gurusamy K, Gluud C, Nikolova D, Davidson B. Assessment of risk of bias in randomized clinical trials in surgery. Br J Surg 2009;96:342–9. [21] Sackett DL. Commentary: Measuring the success of blinding in RCTs: don’t, must, can’t or needn’t? International Journal of Epidemiology 2007;36: 664–5. [22] Whitby M, Johnson BC, Atkinson RL, Stuart G. The comparative efficacy of intravenous cefotaxime and trimethoprim/sulfamethoxazole in preventing infection after neurosurgery: a prospective, randomized study. British Journal of Neurosurgery 2000;14:13–8. [23] Zhu XL, Wong WK, Yeung WM, Mo P, Tsang CS, Pang KH, et al. A randomized, double-blind comparison of ampicillin/sulbactam and ceftriaxone in the prevention of surgical-site infections after neurosurgery. Clinical Therapeutics 2001;23:1281–91. [24] Pons VG, Denlinger SL, Guglielmo BJ, Octavio J, Flaherty J, Derish PA, et al. Ceftizoxime versus vancomycin and gentamicin in neurosurgical prophylaxis. Neurosurgery 1993;33:537. [25] Nejat F, Tajik P, Khashab MEl, Kazmi SS, Khotaei GT. A randomized trial of ceftriaxone versus trimethoprim-sulfamethoxazole to prevent ventriculoperitoneal shunt infection. J Microbiol Immunol Infect 2008;41:112–7. [26] Pons VG, Denlinger SL, Guglielmo BJ, Octavio J, Flaherty J, Derish PA, et al. Ceftizoxime versus vancomycin and gentamicin in neurosurgical prophylaxis: a randomized, prospective, blinded clinical study. Neurosurgery 1993;33:416–22 [discussion 422–3]. [27] Page MGP. Cephalosporins in clinical development. Expert Opinion on Investigational Drugs 2004;13:973–85. [28] Institute for Clinical Systems Improvement. Antibiotic Prophylaxis for Surgical Site Infection Prevention in Adults Fourth Edition; 2010. ˜ [29] Núnez-Pereira S, Pellisé F, Rodríguez-Pardo D, Pigrau C, Sánchez JM, Bagó J, et al. Individualized antibiotic prophylaxis reduces surgical site infections by Gram-negative bacteria in instrumented spinal surgery. European Spine Journal 2011;20(Suppl. 3):397–402. [30] Reichert MCF, Medeiros EAS, Ferraz FAP. Hospital-acquired meningitis in patients undergoing craniotomy: incidence, evolution, and risk factors. American Journal of Infection Control 2002;30:158–64. [31] Korinek A-M, Baugnon T, Golmard J-L, van Effenterre R, Coriat PPL. Risk factors for adult nosocomial meningitis after craniotomy: role of antibiotic prophylaxis. Neurosurgery 2006;62(Suppl. 2):532–9. [32] Kourbeti IS, Jacobs AV, Koslow M, Karabetsos D, Holzman RS. Risk factors associated with postcraniotomy meningitis. Neurosurgery 2007;60:317–25 [discussion 325–6]. [33] Korinek A-M, Golmard J-L, Elcheick A, Bismuth R, Van Effenterre R, Coriat P, et al. Risk factors for neurosurgical site infections after craniotomy: a critical reappraisal of antibiotic prophylaxis on 4,578 patients. British Journal of Neurosurgery 2005;19:155–62. [34] McGirt MJ, Zaas A, Fuchs HE, George TM, Kaye K, Sexton DJ. Risk factors for pediatric ventriculoperitoneal shunt infection and predictors of infectious pathogens. Clinical Infectious Diseases 2003;36:858–62. [35] Christ W. Pharmacological properties of cephalosporins. Infection 1991;19(Suppl. 5):S244–52.
W. Liu et al. / Clinical Neurology and Neurosurgery 116 (2014) 13–19 [36] Fry DE, Barie PS. The changing face of Staphylococcus aureus: a continuing surgical challenge. Surgical Infections 2011;12:191–203. [37] Bratzler DW, Dellinger EP, Olsen KM, Perl TM, Auwaerter PG, Bolon MK, et al. Clinical practice guidelines for antimicrobial prophylaxis in surgery. American Journal of Health-system Pharmacy 2013;70:195–283. [38] Wilcox MH. The tide of antimicrobial resistance and selection. International Journal of Antimicrobial Agents 2009;34(Suppl. 3):S6–10.
19
[39] Kernodle DS, Classen DC, Burke JP, Kaiser AB. Failure of cephalosporins to prevent Staphylococcus aureus surgical wound infections. JAMA 1990;263: 961–6. [40] Lee S-O, Lee ES, Park SY, Kim S-Y, Seo Y-H, Cho YK. Reduced use of thirdgeneration cephalosporins decreases the acquisition of extended-spectrum beta-lactamse-producing Klebsiella pneumoniae. Infection Control and Hospital Epidemiology 2004;25:832–7.
