Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus(mrsa) infections in surgical wounds(review)

Transcription

1 Cochrane Database of Systematic Reviews Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus(mrsa) infections in surgical wounds (Review) GurusamyKS,KotiR,ToonCD,WilsonP,DavidsonBR GurusamyKS,KotiR,ToonCD,WilsonP,DavidsonBR. Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus(mrsa) infections in surgical wounds. Cochrane Database of Systematic Reviews 2013, Issue 8. Art. No.: CD DOI: / CD pub2. Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus(mrsa) infections in surgical wounds(review) Copyright 2013 The Cochrane Collaboration. Published by John Wiley& Sons, Ltd.

2 T A B L E O F C O N T E N T S HEADER ABSTRACT PLAIN LANGUAGE SUMMARY SUMMARY OF FINDINGS FOR THE MAIN COMPARISON BACKGROUND OBJECTIVES METHODS RESULTS Figure Figure DISCUSSION AUTHORS CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES CHARACTERISTICS OF STUDIES DATA AND ANALYSES Analysis 1.1. Comparison 1 Linezolid versus vancomycin, Outcome 1 Eradication of MRSA ADDITIONAL TABLES APPENDICES CONTRIBUTIONS OF AUTHORS DECLARATIONS OF INTEREST SOURCES OF SUPPORT INDEX TERMS i

3 [Intervention Review] Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus (MRSA) infections in surgical wounds Kurinchi Selvan Gurusamy 1, Rahul Koti 1, Clare D Toon 2, Peter Wilson 3, Brian R Davidson 1 1 Department of Surgery, Royal Free Campus, UCL Medical School, London, UK. 2 Public Health, West Sussex County Council, Chichester, UK. 3 Department of Microbiology & Virology, University College London Hospitals, London, UK Contact address: Kurinchi Selvan Gurusamy, Department of Surgery, Royal Free Campus, UCL Medical School, Royal Free Hospital Rowland Hill Street, London, NW3 2PF, UK. kurinchi2k@hotmail.com. Editorial group: Cochrane Wounds Group. Publication status and date: New, published in Issue 8, Citation: Gurusamy KS, Koti R, Toon CD, Wilson P, Davidson BR. Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus (MRSA) infections in surgical wounds. Cochrane Database of Systematic Reviews 2013, Issue 8. Art. No.: CD DOI: / CD pub2. Background A B S T R A C T Methicillin-resistant Staphylococcus aureus (MRSA) infection after surgery is usually rare, but incidence can be up to 33% in certain types of surgery. Postoperative MRSA infection can occur as surgical site infections (SSI), chest infections, or bloodstream infections (bacteraemia). The incidence of MRSA SSIs varies from 1% to 33% depending upon the type of surgery performed and the carrier status of the individuals concerned. The optimal antibiotic regimen for the treatment of MRSA in surgical wounds is not known. Objectives To compare the benefits and harms of various antibiotic treatments in people with established surgical site infections (SSIs) caused by MRSA. Search methods In February 2013 we searched the following databases: The Cochrane Wounds Group Specialised Register; The Cochrane Central Register of Controlled Trials (CENTRAL); Database of Abstracts of Reviews of Effects (DARE); NHS Economic Evaluation Database; Health Technology Assessment (HTA) Database; Ovid MEDLINE; Ovid MEDLINE (In-Process & Other Non-Indexed Citations); Ovid EMBASE; and EBSCO CINAHL. Selection criteria We included only randomised controlled trials (RCTs) comparing one antibiotic regimen with another antibiotic regimen for the treatment of SSIs due to MRSA. All RCTs irrespective of language, publication status, publication year, or sample size were included in the analysis. Data collection and analysis Two review authors independently decided on inclusion and exclusion of trials, and extracted data. We planned to calculate the risk ratio (RR) with 95% confidence intervals (CI) for comparing the binary outcomes between the groups and mean difference (MD) with 95% CI for comparing the continuous outcomes. We planned to perform the meta-analysis using both a fixed-effect and a randomeffects model. We performed intention-to-treat analysis whenever possible. 1

4 Main results We included one trial involving 59 people hospitalised because of MRSA SSIs. Thirty participants were randomised to linezolid (600 mg either intravenously or orally every 12 hours for seven to 14 days) and 29 to vancomycin (1 g intravenously every 12 hours for seven to 14 days). The type of surgical procedures that were performed were not reported. The trial reported one outcome, which was the eradication of MRSA. The proportion of people in whom MRSA was eradicated was statistically significantly higher in the linezolid group than in the vancomycin group (RR 1.80; 95% CI 1.20 to 2.68). Authors conclusions There is currently no evidence to recommend any specific antibiotic in the treatment of MRSA SSIs. Linezolid is superior to vancomycin in the eradication of MRSA SSIs on the basis of evidence from one small trial that was at high risk of bias, but the overall clinical implications of using linezolid instead of vancomycin are not known. Further well-designed randomised clinical trials are necessary in this area. P L A I N L A N G U A G E S U M M A R Y Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus (MRSA) infections in people with surgical wounds Some people having surgery develop wound infections. These are usually caused by bacteria. Most of these wound infections heal naturally, or after treatment with routine antibiotics. However, some bacteria are resistant to routine antibiotics, for example, methicillinresistant Staphylococcus aureus (MRSA). MRSA infection after surgery is rare, but can occur in wounds (surgical site infections, or SSIs), the chest, or bloodstream (bacteraemia). MRSA SSIs occur in 1% to 33% of people having surgery, depending on the type of surgery concerned; they can be life-threatening and cause extended hospital stays. We do not know what is the best antibiotic treatment for a person who has an MRSA SSI. We aimed to resolve this uncertainty by performing a thorough search of the medical literature for studies that compared different antibiotic treatments for MRSA SSIs. We included only randomised controlled trials, as, if conducted properly, these provide the best information. We did not limit our search for trials according to language or year of publication. Two review authors independently identified the trials and extracted information. We identified only one trial that compared different antibiotic treatments for MRSA SSIs. This trial had 59 participants who were hospitalised because of their MRSA SSIs. Thirty participants in this trial received an antibiotic called linezolid, which can be taken in tablet form or given as injections through the vein (intravenously). The remaining participants received another antibiotic called vancomycin, which can be only given intravenously. The type(s) of surgical procedures that the participants had were not specified. MRSA was eradicated in more people who received linezolid than vancomycin. It would be helpful if this finding were confirmed by other studies. This trial did not report on other features of eradicating MRSA with these antibiotics, such as: 1. whether the wound healed quickly; 2. length of hospital stay; 3. quality of life, and 4. whether the benefits of treatment outweighed any unwanted side-effects of the medicine. Overall, the quality of available evidence was poor. Currently, we cannot recommend any particular antibiotic for treating MRSA SSIs. Linezolid appears to be better than vancomycin for eradication of MRSA SSIs according to low-quality evidence from this one small trial, but the wider implications of this treatment are not known. Further well-designed randomised clinical trials are necessary to identify the best antibiotic treatment for MRSA SSIs. 2

5 S U M M A R Y O F F I N D I N G S F O R T H E M A I N C O M P A R I S O N [Explanation] Antibiotic therapy for the treatment of methicillin- resistant Staphylococcus aureus(m RSA) infections in surgical wounds Patient or population: people with MRSA infection in surgical wounds Settings: secondary or tertiary care Intervention: linezolid Comparison: vancomycin Outcomes Illustrative comparative risks(95% CI) Relative effect (95% CI) Assumed risk Vancomycin Corresponding risk Linezolid Eradication of M RSA 483 per per 1000 (579 to 1000) RR1.8 (1.2 to 2.68) No of Participants (studies) 59 (1 study) Quality of the evidence (GRADE) verylow 1,2,3 *The basis for theassumedrisk was the control group risk in the study. Thecorrespondingrisk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI) CI: Confidence interval; RR: Risk ratio GRADE Working Group grades of evidence High quality: further research is very unlikely to change our confidence in the estimate of effect M oderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate Lowquality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate Very low quality: we are very uncertain about the estimate 1 This trial was of high risk of bias 2 This was the only trial,so we could not assess inconsistency 3 There was only trial,so we were unable to assess reporting bias 3

6 B A C K G R O U N D Description of the condition Methicillin-resistant Staphylococcus aureus (MRSA), that is isolates of the bacterium Staphylococcus aureus that are not susceptible to the antibiotic methicillin, were first discovered in 1961 (Barber 1961; Jevons 1961; Knox 1961), and outbreaks of MRSA have been reported since the 1970s (Klimek 1976; O Toole 1970). MRSA infection is associated with significant mortality and morbidity. In the European Union member states plus Norway and Iceland, MRSA infections cause an estimated one million extra hospital stays and cost an estimated EUR 600 million annually (ECDC 2009a), while in the USA, they cause an estimated 125,000 hospitalisations annually (Kuehnert 2005). Although there has been a decrease in the incidence of MRSA in some countries, such as the USA (Kallen 2010), probably as a result of measures to combat MRSA infections (ECDC 2009b), there has also been an increase in incidence of infections in Nordic countries (Skov 2005). Methicillin (meticillin is the International Nonproprietary Name but we have used methicillin in this review in line with the way we speak) resistance is a marker of resistance to some beta-lactam antibiotics (i.e. the penicillin and cephalosporin group of antibiotics, which are some of the most commonly-used antibiotics for patients) (Otter 2011). In addition to beta-lactam antibiotics, MRSA may be resistant to many other commonly-used antibiotics such as erythromycin, clindamycin, gentamycin, ciprofloxacin, and fusidic acid (Otter 2011). So, even though methicillin is not commonly used at present, methicillin resistance is an indicator of resistance to a wide range of antibiotics. Currently, there are concerns that farm animals, which do not exhibit the symptoms and signs of MRSA infection and with which farmers are in regular contact, may become reservoirs of methicillin-resistant organisms and a source of a major epidemic of MRSA in the community in developed countries (Wulf 2008). MRSA infection after surgery is usually rare, but incidence can be up to 33% in certain types of surgery (Sanjay 2010). Postoperative MRSA infection can occur as surgical site infections (SSI), chest infections, or bloodstream infections (bacteraemia) (Fraser 2010; Reddy 2007; Sanjay 2010). The incidence of MRSA SSIs varies from 1% to 33% depending upon the type of surgery performed and the carrier status of the individuals concerned (Harbarth 2008; Reddy 2007; Ridgeway 2005; Sanjay 2010; Shukla 2009). Emergency surgery, contaminated surgery, immunosuppression, and presence of co-morbidities such as diabetes in patients are all associated with a higher incidence of MRSA SSIs (Harbarth 2008). The role of MRSA SSIs in development of MRSA bacteraemia is not known, however, it would be logical to expect deep SSIs and organ SSIs to have the potential to cause it. Rates of MRSA bacteraemia vary between types of surgery, and, while it is usually rare, infection rates can reach 5% in some types of surgery (Sanjay 2010). MRSA bacteraemia is associated with a 30-day mortality of about 15% to 38% (Lamagni 2011; Lewis 2011; Kang 2012; Wang 2010), and a one-year mortality of about 55% (Kaye 2008),compared to a 6% 30-day mortality in methicillin-sensitive Staphylococcus aureus bacteraemia (Kang 2012), and a 31% 30-day mortality in bacteraemia due to other organisms (Melzer 2013). Surgery that involves indwelling implants and prostheses is prone to the development of biofilms (layers of micro-organisms) that can increase antimicrobial resistance (Donlan 2002). Indwelling drains and catheters can also predispose individuals to infections. This means that people undergoing surgery with indwelling implants, prostheses, drains, and catheters are more prone to infection with Staphylococcus aureus, which has a very high capacity to develop biofilms, although the methicillin resistant form itself is not associated with a higher capacity to form biofilms (Sanchez 2013). There is currently no evidence that operations involving prostheses are more prone to MRSA infection than operations not involving prostheses. MRSA SSIs have been shown to be associated with increased mortality in people undergoing cardiac surgery (Reddy 2007): those with MRSA SSIs have an in-hospital mortality rate of 12.9% compared to a rate of 3% in those without infection. People who developed MRSA SSIs stayed longer in hospitals than those who did not develop them (Chemaly 2010; Shukla 2009). Horan et al published the Centers for Disease Control and Prevention s (CDC) criteria for SSI (Horan 1992), which have been summarised in Table 1. Description of the intervention The Oxford English Dictionary defines an antibiotic substance as, One of a class of substances produced by living organisms that is capable of destroying, or inhibiting the growth of, micro-organisms, and is especially used for therapeutic purposes. Synthetic organic compounds that have similar properties are also called antibiotics (OED 2011). A variety of antibiotics, including betalactams (penicillin derivatives, cephalosporins), glycopeptide antibiotics (e.g. vancomycin, teicoplanin), clindamycin, trimethoprim-sulfamethoxazole (TMP-SMX), tetracyclines (doxycycline or minocycline), linezolid, daptomycin, telavancin, rifampicin, gentamycin, and fluoroquinolone, can be used for the treatment of MRSA, since, despite the resistance of MRSA to some antibiotics, it is sensitive to one or more of the above antibiotics (Liu 2011). Different antibiotics are administered in different ways, with the common routes being oral, intravenous, and topical (surface) administration (Liu 2011). Antibiotics can be given as a single agent or in combinations, preoperatively, during operations, postoperatively, or in a combination of the above (Liu 2011). How the intervention might work 4

7 The mechanisms of action vary for different types of antibiotics, but in general terms they either destroy MRSA or prevent the bacteria from dividing (i.e. reproducing). Why it is important to do this review A systematic review that addressed the use of different microbial treatments for the treatment of people colonised with MRSA either nasally, or at extra-nasal sites, concluded that there was insufficient evidence to support use of topical or systemic antimicrobial therapy for eradicating nasal or extra-nasal MRSA (Loeb 2003).Thus far, there has been no systematic review comparing antibiotics for the treatment of post-operative MRSA infection. Information to inform clinical guidelines for surgical centres that treat people with MRSA is needed. O B J E C T I V E S To compare the benefits and harms of various antibiotic treatments in people with established surgical site infections (SSIs) caused by MRSA. Types of interventions We considered comparisons of different antibiotic treatments (and regimens) and comparisons of the same antibiotic with different regimens. We planned to include a combination of antibiotics as a complex intervention, that is, consider the combination as a package. A complex intervention is one that can be divided into one or more individual components, but where the component that causes the treatment effect is not known. Co-interventions were allowed where they were performed equally across trial groups. We considered antibiotic administration prior to development of SSI as a co-intervention, but, provided that participants in all arms of a trial received the same antibiotic prior to development of SSIs, trials addressing prophylactic antibiotic administration were included in the review. We included only trials in which the trial groups differed systematically only in the antibiotic regimens used. We excluded trials where antibiotics used as prophylaxis for the prevention of MRSA SSI in people undergoing surgery (but not as treatment for patients who developed post-operative MRSA) as these interventions are covered in a different review (Gurusamy 2013b). Types of outcome measures M E T H O D S Criteria for considering studies for this review Types of studies We included randomised clinical trials (RCTs), irrespective of their use of blinding, language of publication, publication status, date of publication, study setting, or sample sizes. We planned to include cluster-randomised clinical trials if the effect estimate was available after adjusting for clustering effect. We excluded quasi-randomised studies because of the risk of selection bias in these trials. Types of participants Any person having undergone a surgical procedure (irrespective of age, or whether their surgery involved implants or prostheses, or whether they had impaired immunity or any other co-morbidities or disease manifestations, such as cancer, diabetes, or jaundice, that could affect wound healing) with established MRSA-associated postoperative SSIs. All types of surgery were eligible, so surgical wounds could be on any part of the body (for example: the head, trunk, and the limbs). People with non-surgical wounds will be covered in another review (Gurusamy 2013a). Primary outcomes 1. All cause mortality. 2. Other serious adverse events: defined as any event that is life-threatening; requires inpatient hospitalisation; results in a persistent or significant disability; or any important medical event, which might jeopardise the patient or requires intervention to prevent it (ICH-GCP 1996) (at maximal followup) e.g. rates of bacteraemia, wound dehiscence. 3. Quality of life (as defined by trial authors). Secondary outcomes 1. Total length of hospital stay. 2. Use of healthcare resources (e.g. hospital visits). 3. Eradication of MRSA (at least three consecutive wound or discharge samples negative for MRSA separated by a time interval of at least two days between samples, or as defined by trial authors) at maximal follow-up. 4. Time to complete wound healing (as defined by trial authors). Search methods for identification of studies Electronic searches 5

8 In February 2013 we searched the following electronic databases to identify reports of relevant randomised clinical trials The Cochrane Wounds Group Specialised Register (searched 28 February 2013); The Cochrane Central Register of Controlled Trials (CENTRAL) (2013, Issue 1); The Database of Abstracts of Reviews of Effects (DARE) (2013, Issue 1); NHS Economic Evaluation Database (2013, Issue 1); Health Technology Assessment (HTA) Database (2013, Issue 1); Ovid MEDLINE (1948 to February Week ); Ovid MEDLINE (In-Process & Other Non-Indexed Citations, 28 February 2013); Ovid EMBASE (1974 to 2013 Week 08); EBSCO CINAHL (1982 to 21 February 2013) We used the following search strategy in The Cochrane Central Register of Controlled Trials (CENTRAL): #1 MeSH descriptor Methicillin Resistance explode all trees #2 MeSH descriptor Staphylococcal Infections explode all trees #3 MeSH descriptor Staphylococcus aureus explode all trees #4 (#2 OR #3) #5 (#1 AND #4) #6 MeSH descriptor Methicillin-Resistant Staphylococcus aureus explode all trees #7 (methicillin NEXT resistan*) or (meticillin NEXT resistan*) or MRSA #8 (#5 OR #6 OR #7) #9 MeSH descriptor Wound Infection explode all trees #10 MeSH descriptor Sepsis explode all trees #11 MeSH descriptor Soft Tissue Infections explode all trees #12 MeSH descriptor Surgical Wound Dehiscence explode all trees #13 surg* NEAR/5 infect* #14 surg* NEAR/5 wound* #15 surg* NEAR/5 site* #16 surg* NEAR/5 incision* #17 surg* NEAR/5 dehisc* #18 wound* NEAR/5 dehisc* #19 (deep NEXT infection*) or deep sepsis or (infected NEXT collection*) #20 (#9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19) #21 MeSH descriptor Anti-Bacterial Agents explode all trees #22 MeSH descriptor Cephalosporins explode all trees #23 MeSH descriptor Tetracycline explode all trees #24 MeSH descriptor Penicillins explode all trees #25 antibiotic* or penicillin* or beta-lactam* or cephalosporin* or clindamycin or trimethoprim* or tetracycline* or doxycycline or minocycline or linezolid or vancomycin or daptomycin or telavancin or rifampicin or gentamycin or gentamicin or fluoroquinolone #26 (#21 OR #22 OR #23 OR #24 OR #25) #27 (#8 AND #20 AND #26) The search strategies for Ovid MEDLINE, Ovid EMBASE and EBSCO CINAHL can be found in Appendix 1. We combined the Ovid MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity- and precision-maximising version (2008 revision) ( Lefebvre 2011). We combined the EMBASE search with the Ovid EMBASE filter developed by the UK Cochrane Centre (Lefebvre 2011). We combined the CINAHL searches with the trial filters developed by the Scottish Intercollegiate Guidelines Network ( SIGN 2011). We did not restrict studies with respect to language, date of publication or study setting. We searched the metaregister of Controlled Trials (mrct) ( and the ICTRP (International Clinical Trials Registry Platform) portal maintained by World Health Organization ( on 11 December The meta-register includes the ISRCTN Register and the NIH ClinicalTrials.gov Register amongst others. The ICRTP portal includes these trial registers, along with trial registry data from a number of countries. Searching other resources We searched the references of included trials to identify further relevant trials. We also contacted experts in MRSA infection to identify further relevant trials. Data collection and analysis We performed the systematic review following the instructions in the Cochrane Handbook for Systematic Reviews of Intervention (Higgins 2011a). Selection of studies Two review authors (KG or CT and RK) independently identified trials for inclusion by screening the titles and abstracts. We obtained full texts for references selected by at least one of the reviewers. We made the final selection based on full texts. We have listed excluded studies with reasons for their exclusion (Characteristics of excluded studies). We resolved any differences through discussion. Data extraction and management Two review authors (KG and RK) independently extracted the following data from trial reports: 1. Year, and language, of publication. 2. Country in which the trial was performed. 6

9 3. Year(s) of conduct of the trial. 4. Inclusion and exclusion criteria. 5. Sample sizes. 6. Type(s) of surgery. 7. Details of antibiotic treatment including dose(s), route(s), frequency, and duration(s). 8. Outcomes (as described above). 9. Risk of bias (as described below). We sought further information from trial authors when sufficient information was not available in the report. In future, if multiple reports exist for a trial, we will examine all the reports for information (there were no multiple reports for the included trial). If there is any doubt about whether the trials share the same participants - completely or partially (by identifying common authors and centres) - we plan to contact the trial authors to check whether the trial report has been duplicated, and to seek clarification for any unclear or missing information. We planned to resolve any differences in opinion through discussion with PW and BRD. Assessment of risk of bias in included studies We followed the guidance in the Cochrane Handbook for Systematic Reviews of Intervention to assess risk of bias in included studies (Higgins 2011b). According to empirical evidence (Kjaergard 2001; Moher 1998; Schulz 1995; Wood 2008), the risk of bias of the trials was assessed on the following domains: Sequence generation 1. Low risk of bias: the method used was either adequate (e.g. computer-generated random numbers, table of random numbers), or unlikely to introduce confounding. 2. Uncertain risk of bias: there was insufficient information to assess whether the method used was likely to introduce confounding. 3. High risk of bias: the method used was improper and likely to introduce confounding (e.g. quasi-randomised studies). Such studies were excluded. 1. Low risk of bias: blinding was performed adequately in the participants and personnel, or the outcome measurement was not likely to be influenced by lack of blinding. 2. Uncertain risk of bias: there was insufficient information to assess whether the type of blinding used was likely to induce bias on the estimate of effect. 3. High risk of bias: no blinding or incomplete blinding, and the outcome or the outcome measurement was likely to be influenced by lack of blinding. Blinding of outcome assessors 1. Low risk of bias: blinding was performed adequately, or the outcome measurement was not likely to be influenced by lack of blinding. 2. Uncertain risk of bias: there was insufficient information to assess whether the type of blinding used was likely to induce bias on the estimate of effect. 3. High risk of bias: no blinding or incomplete blinding, and the outcome or the outcome measurement was likely to be influenced by lack of blinding. Incomplete outcome data 1. Low risk of bias: the underlying reasons for missing data were unlikely to make treatment effects depart from plausible values, or proper methods have been employed to handle missing data. 2. Uncertain risk of bias: there was insufficient information to assess whether the missing data mechanism, in combination with the method used to handle missing data, was likely to induce bias on the estimate of effect. 3. High risk of bias: the crude estimate of effects (e.g. complete case estimate) was clearly biased due to the underlying reasons for missing data, and the methods used to handle missing data were unsatisfactory. Allocation concealment 1. Low risk of bias: the method used was unlikely to induce bias on the final observed effect (e.g. central allocation). 2. Uncertain risk of bias: there is insufficient information to assess whether the method used was likely to induce bias on the estimate of effect. 3. High risk of bias: the method used was likely to induce bias on the final observed effect (e.g. open random allocation schedule). Blinding of participants and personnel Selective outcome reporting 1. Low risk of bias: the trial protocol was available and all of the trial s pre-specified outcomes of interest in the review have been reported, or if the trial protocol was not available, the primary outcomes in this review were reported. 2. Uncertain risk of bias: there was insufficient information to assess whether the magnitude and direction of the observed effect was related to selective outcome reporting. 3. High risk of bias: not all of the trial s pre-specified primary outcomes were reported. We consider trials that are classified as being at low risk of bias in all the above domains to be low bias-risk trials. All other remaining trials will be considered as being high bias-risk trials. 7

10 Measures of treatment effect For dichotomous variables, we calculated the risk ratio (RR) with 95% confidence interval (CI). Risk ratio calculations do not include trials in which no events occurred in either group, whereas risk difference calculations do. We planned to report the risk difference where the results obtained by using this association measure were different from risk ratio in terms of interpretation of information or arriving at conclusions. For continuous variables, we planned to calculate the mean difference (MD) with 95% CI for outcomes that are measured or can be converted to the same units (such as length of hospital stay), and standardised mean difference (SMD) with 95% CI for outcomes (such as quality of life), where different assessment scales might have been used. For time-to-event outcomes such as survival at maximal follow-up, we planned to calculate the hazard ratio (HR) with 95% CI. Unit of analysis issues The unit of analysis were individual people undergoing surgical procedures. We performed an intention-to-treat analysis whenever possible (Newell 1992). We imputed missing data for binary outcomes using various scenarios such as best-best, best-worst, worst-best, worst-worst scenarios (Gurusamy 2009). In the best-best scenario, the outcomes of people with missing data in both groups were assumed to be good. In the best-worst scenario, the outcomes of people with missing data in intervention group were assumed to be good while the outcomes were assumed to be bad for those with missing data in the control group. The worst-best scenario was the opposite of the best-worst scenario, i.e. the outcomes of people with missing data in the intervention group were assumed to be bad while the outcomes were assumed to be good for those in the control group with missing data. In the worst-worst scenario, the outcomes of people with missing data in both groups were assumed to be bad. For continuous outcomes, we planned to use available-case analysis. We planned to impute the standard deviation from P values according to the instructions in the Cochrane Handbook for Systematic Reviews of Intervention, and planned to use the median for meta-analysis when the mean was not available (Higgins 2011c). If it was not possible to calculate the standard deviation from the P value or the confidence intervals, we planned to impute the standard deviation as the highest standard deviation in the other trials included under that outcome, fully recognising that this form of imputation would decrease the weight of the study for calculation of mean differences, and bias the effect estimate to no effect in case of standardised mean difference (Higgins 2011c). For time-to-event outcomes, if the hazard ratio and 95% confidence intervals were not reported, we planned to obtain the logarithm of hazard ratios (ln(hr)) and the standard error (SE) of ln(hr) according to the methods described by Parmar Assessment of heterogeneity We planned to explore heterogeneity by means of the Chi 2 test, with significance set at P value 0.10, and measure the quantity of heterogeneity by I 2 (Higgins 2002). We also planned to use overlapping of confidence intervals on the forest plot to determine heterogeneity (Schunemann 2009). Thresholds for the interpretation of I 2 can be misleading. A rough guide to interpretation follows (Deeks 2011): 1. 0 to 40%: might not be important % to 60%: may represent moderate heterogeneity % to 90%: may represent substantial heterogeneity % to 100%: may represent considerable heterogeneity. The decision to classify the overlapping I 2 levels will be based on the overlapping of confidence intervals, and whether the heterogeneity was in the direction of effect or just in the magnitude of effect. Dealing with missing data Assessment of reporting biases We planned to use visual asymmetry on a funnel plot to explore reporting bias in the presence of at least 10 trials (Egger 1997; Macaskill 2001). We planned to perform the linear regression approach described by Egger 1997 to determine the funnel plot asymmetry. Selective reporting was also be considered as evidence of reporting bias. Data synthesis We planned to perform meta-analysis only if there was sufficient clinical homogeneity in terms of both participants included in the trials and of antibiotics used (i.e. similar class of antibiotics). We planned to perform meta-analyses using the RevMan 5 software package (RevMan 2011), and to follow the recommendations of The Cochrane Handbook for Systematic Reviews of Intervention ( Deeks 2011). We planned to use both a random-effects model (DerSimonian 1986), and a fixed-effect model in meta-analyses (DeMets 1987). Should there be discrepancy between the two models, we planned to report both results; otherwise we planned to report the results of the fixed-effect model. We planned to use the generic inverse variance method to combine hazard ratios for time-to-event outcomes. Summary of findings We have presented the Summary of findings table for all the primary and secondary outcomes that were reported (Summary of findings for the main comparison; Schünemann 2011). 8

11 Subgroup analysis and investigation of heterogeneity We planned to perform the following subgroup analyses: 1. Different antibiotics. 2. Operations involving prosthetic implants and indwelling devices. 3. Different types of surgery. 4. People with co-morbidities versus those without serious comorbidities. We planned to use the test for subgroup differences available via RevMan output to identify the differences between subgroups. Sensitivity analysis We planned to perform a sensitivity analysis by imputing data for binary outcomes using various scenarios such as best-best, bestworst, worst-best, worst-worst scenarios as described previously ( Gurusamy 2009). We also planned to perform a sensitivity analysis by testing the effect of removing trials at unclear or high risk of bias and trials in which the mean and the standard deviation were imputed from the analysis. R E S U L T S Description of studies Results of the search We identified a total of 489 unique references through electronic searches. We excluded 484 clearly irrelevant references through reading titles and abstracts. We retrieved five references in full for further assessment. We did not identify any additional references to trials by scanning reference lists of included trials. We did not identify any new trials through other searches. We excluded four references because of the reasons specified in the Characteristics of excluded studies. One trial met the inclusion criteria and was included in this review (Weigelt 2004). Since this was the only trial that met the inclusion criteria, no meta-analysis was performed. The reference flow chart is shown in Figure 1. 9

12 Figure 1. Study flow diagram 10

13 Included studies We included one trial involving 59 people hospitalised because of MRSA SSIs, randomised to either linezolid (600 mg either intravenously or orally every 12 hours for seven to 14 days) (30 people) or to vancomycin (1 g intravenously every 12 hours for seven to 14 days) (29 people) (Weigelt 2004). The average age of participants, the proportion of women and the type(s) of surgical procedures was not reported (Weigelt 2004). The only outcome reported was eradication of MRSA, which the trial authors defined as previously-infected participants being culture-negative for MRSA one week after the end of treatment. Further details of this trial are shown in Characteristics of included studies. Excluded studies We excluded four studies. Two trials were excluded because there were no data on people who underwent surgery (Lucasti 2008; Van Laethem 1988); one study involved animals (Jacqueline 2011); and in the fourth trial all MRSA episodes were treated with no randomisation in the choice of antibiotics (Philpott 2003). Risk of bias in included studies The only trial included in this review was at of high risk of bias. The risk of bias in the individual domains are shown in Figure 2. Figure 2. Risk of bias summary: review authors judgements about each risk of bias item for each included study. 11

14 Allocation The methods used for generation of random sequence and allocation concealment were not reported (Weigelt 2004). Blinding Blinding of participants, personnel and outcome assessors was not reported (Weigelt 2004). Incomplete outcome data Although there were post-randomisation drop-outs, imputations of outcomes under different scenarios did not alter the conclusions, so there was no attrition bias (Weigelt 2004). Sensitivity analysis Six participants were excluded from the report: four belonged to the linezolid group and two to the vancomycin group. We performed a sensitivity analysis by imputing the outcomes under various scenarios described in the Dealing with missing data and Sensitivity analysis sections. The sensitivity analysis did not change the results (Analysis 1.1). The worst-best scenario, in which the effect estimate was (RR 1.48; 95% CI 1.00 to 2.19), deserves a specific mention: although the confidence interval included 1, this was because of rounding. The actual lower and upper confidence intervals before rounding to two decimals were and The P value before rounding to two decimal places was So, even though it appears that there was no statistically significant difference between the two groups on the forest plot, the eradication of MRSA was significantly higher in the linezolid group than in the vancomycin group. Selective reporting Many of the important outcomes that would normally have been measured in such a trial were not reported, so there was selective reporting bias (Weigelt 2004). Reporting bias We did not explore reporting bias because of the presence of only one trial in this review. D I S C U S S I O N Other potential sources of bias There was no other bias in this trial (Weigelt 2004). Effects of interventions See: Summary of findings for the main comparison Antibiotic therapy for the treatment of methicillin-resistant Staphylococcus aureus (MRSA) infections in surgical wounds The only outcome (of interest for this review) reported in the trial was eradication of MRSA (Weigelt 2004). The proportion of people in whom MRSA was eradicated was significantly higher in the linezolid group than in the vancomycin group (RR 1.80; 95% CI 1.20 to 2.68) (Analysis 1.1). The interpretation did not change by calculating the risk difference. Since this was the only trial included in this review, questions about heterogeneity and fixed-effect model versus random-effects model were not relevant. Subgroup analysis Subgroup analysis was not performed because of the inclusion of only one trial. Summary of main results We included only one trial, with 59 participants, in this review. This trial was at high risk of bias, and found that the eradication of MRSA was significantly better with linezolid than with vancomycin (RR 1.80; 95% CI 1.20 to 2.68). MRSA was eradicated in approximately 87% of people after seven to 14 days treatment with linezolid compared with an eradication in approximately 49% of people after seven to 14 days treatment with vancomycin. This trial did not report on the clinical impact of the increase in the proportion eradicated with linezolid compared with vancomycin, or whether there were any serious adverse events related to the antibiotic treatment. Another review that investigated the safety of linezolid reported that people with neutropenia, Gram-positive infections, and foot infections who received treatment with linezolid showed similar serious adverse events to those treated with other antibiotics such as vancomycin, dalbavancin, and teicoplanin (Vinh 2009). This review reported that serious antibiotic-related adverse events occurred in fewer than 1% of the people taking linezolid (Vinh 2009); these serious adverse events included transient ischaemic attack (mini stroke), hypertension, severe vomiting, elevated liver enzymes, and thrombocytopenia (low 12

15 blood platelet levels). The authors of the review concluded that linezolid was generally safe and well tolerated (Vinh 2009). It is must also be noted that linezolid was administered either intravenously or orally, although the authors did not state the reasons that governed the choice of intravenous or oral preparations for individual participants. Previous research has shown that oral linezolid is well tolerated, and that bioavailability of linezolid in oral and intravenous forms is similar (Keel 2011). This appears to be an advantage of linezolid compared to vancomycin, which is administered intravenously. However, since complications, such as thrombocytopaenia and peripheral neuropathy, can occur with prolonged treatment, linezolid is not recommended for treatment that is longer than four weeks (Vinh 2009). However, further information on serious adverse events in surgical patients and further clinical information is necessary before we can recommend linezolid rather than vancomycin for treatment of MRSA wound infections. The incidence of MRSA SSIs varies from 1% to 33% depending upon the type of surgery performed and the carrier status of the individuals concerned (Harbarth 2008; Reddy 2007; Ridgeway 2005; Sanjay 2010; Shukla 2009). Considering the disease burden posed by MRSA SSIs, it is very surprising that we were able to identify only one trial that compared different antibiotic regimens in people with MRSA SSIs. One possible reason for the paucity of trials in this field may be that researchers usually make use of evidence from antibiotic comparisons for treatment of spontaneous skin and subcutaneous infections such as abscesses. However, the pathophysiology of spontaneous skin and subcutaneous infections is different from that of SSI. Spontaneous MRSA abscesses are likely to develop in people with other local or systemic illnesses that predispose them to such abscesses. While such systemic illnesses may predispose people to develop SSI (Watanabe 2008), the primary cause of SSIs is surgery. Another possible reason for the paucity of randomised clinical trials, is the use of evidence from observational studies. RCTs are currently considered to be the best design for evaluating interventions, and are clearly necessary in this field. Overall completeness and applicability of evidence This review included only one trial, in which the type of people undergoing surgery was not stated clearly, so we are unable to generalise the findings to all people undergoing surgery. Quality of the evidence As shown in the Summary of findings for the main comparison, the quality of evidence was very low. There are no insurmountable barriers to obtaining very high quality evidence in this area, unlike some other surgical comparisons where it is impossible, or unethical, to blind people. Potential biases in the review process We have performed a thorough search of the literature without any language or time of publication restrictions. In spite of this, we will not be able to identify any trials that were conducted, but not reported, in the pre-mandatory trial registration era. We included only trials in which MRSA infection and surgical treatment was mentioned. It is highly likely that several thousand trials have been conducted that investigated antibiotic treatments for all surgical site infections without assessing or reporting the methicillinresistance status of the organisms. Given the nature of the topic, one has to be pragmatic and accept that such trials are unlikely to be identified electronically, and that contacting the authors of every trial that has assessed antibiotic treatment for SSIs in order to obtain information about whether the trialists measured the MRSA status of the organisms, and to obtain the information in way that it could be used for this systematic review, would be very resource intensive. The reliability of information obtained in such a manner would also be a major concern. Agreements and disagreements with other studies or reviews This is the first systematic review on this topic. We agree with the trial author that linezolid is superior to vancomycin in the eradication of MRSA SSI (Weigelt 2004), although this conclusion is based on evidence from one small trial of high risk of bias and the overall clinical implications of using linezolid instead of vancomycin are not known. A U T H O R S C O N C L U S I O N S Implications for practice Currently, there is no evidence to recommend any specific antibiotic in the treatment of methicillin-resistant Staphylococcus aureus (MRSA) surgical site infections (SSIs). Linezolid appears to be superior to vancomycin in the eradication of MRSA SSIs on the basis of evidence from one small trial at high risk of bias, but the overall clinical implications of using linezolid instead of vancomycin is not known. Implications for research Further well-designed randomised clinical trials are necessary. Such trials should include mortality, morbidity, serious antibiotic-related adverse events, quality of life, and resource use as the basic outcomes. 13

16 A C K N O W L E D G E M E N T S The authors would like to acknowledge the contribution of the Wounds Group Editors (Susan O Meara) and peer referees (Catriona McDaid, Gill Worthy, Patricia Danielsen, Roy Buffery) and copy-editor (Elizabeth Royle). R E F E R E N C E S References to studies included in this review Weigelt 2004 {published data only} Weigelt J, Kaafarani HM, Itani KM, Swanson RN. Linezolid eradicates MRSA better than vancomycin from surgical-site infections. American Journal of Surgery 2004;188(6): References to studies excluded from this review Jacqueline 2011 {published data only} Jacqueline C, Amador G, Desessard C, Le Mabecque V, Miegeville AF, Caillon J, et al. Assessment of linezolid, vancomycin, and rifampicin as monotherapy in comparison with combination of linezolid/rifampicin and vancomycin/ rifampicin in the treatment of methicillin-resistant Staphylococcus aureus acute osteomyelitis. Clinical Microbiology and Infection 2011;17(Supplement S4):S428. Lucasti 2008 {published data only} Lucasti C, Jasovich A, Umeh O, Jiang J, Kaniga K, Friedland I. Efficacy and tolerability of IV doripenem versus meropenem in adults with complicated intra-abdominal infection: a phase III, prospective, multicenter, randomized, double-blind, noninferiority study. Clinical Therapeutics 2008;30(5): Philpott 2003 {published data only} Philpott-Howard J, Burroughs A, Fisher N, Hastings M, Kibbler C, Mutimer D, et al. Piperacillin-tazobactam versus ciprofloxacin plus amoxicillin in the treatment of infective episodes after liver transplantation. Journal of Antimicrobial Chemotherapy 2003;52: Van Laethem 1988 {published data only} Van Laethem Y, Hermans P, De Wit S, Goosens H, Clumeck N. Teicoplanin compared with vancomycin in methicillin-resistant Staphylococcus aureus infections: preliminary results. Journal of Antimicrobial Chemotherapy 1988;21(Supplement A):81 7. Additional references Barber 1961 Barber M. Methicillin-resistant staphylococci. Journal of Clinical Pathology 1961 Jul;14: Chemaly 2010 Chemaly RF, Hachem RY, Husni RN, Bahna B, Rjaili GA, Waked A, et al. Characteristics and outcomes of methicillinresistant Staphylococcus aureus surgical-site infections in patients with cancer: a case-control study. Annals of Surgical Oncology 2010;17(6): Deeks 2011 Deeks JJ, Higgins JPT, Altman DG, on behalf of the Cochrane Statistical Methods Group and the Cochrane Bias Methods Group (Editors). Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version [updated March 2011]. The Cochrane Collaboration, Available from DeMets 1987 DeMets DL. Methods for combining randomized clinical trials: strengths and limitations. Statistics in Medicine 1987; 6(3): DerSimonian 1986 DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clinical Trials 1986;7(3): Donlan 2002 Donlan RM, Costerton JW. Biofilms: survival mechanisms of clinically relevant microorganisms. Clinical Microbiology Reviews 2002;15(2): ECDC 2009a ECDC. European Centre for Disease Prevention and Control. Annual epidemiological report on communicable diseases in Europe SUR Annual Epidemiological Report on Communicable Diseases in Europe.p 2009 (accessed on 17 June 2011). ECDC 2009b ECDC. European Centre for Disease Prevention and Control. The bacterial challenge: Time to react. A call to narrow the gap between multi-drug resistant bacteria in the EU and the development of new anti-bacterial agents TER The Bacterial Challenge Time to React.pdf 2009 (accessed on 17 June 2011). Egger 1997 Egger M, Davey SG, Schneider M, Minder C. Bias in metaanalysis detected by a simple, graphical test. BMJ 1997;315 (7109): Fraser 2010 Fraser S, Brady R, Graham C, Paterson-Brown S, Gibb A. Methicillin-resistant Staphylococcus aureus in surgical patients: identification of high-risk populations for the development of targeted screening programmes. Annals of the Royal College of Surgeons of England 2010;92(4):