Criteria for considering studies for this review

Search methods for identification of studies

We will search The Cochrane Hepato‐Biliary Group Controlled Trials Register, theCochrane Central Register of Controlled Trials (Gluud 2006) in The Cochrane Library (CENTRAL), MEDLINE, EMBASE and Science Citation Index Expanded (Royle 2003). We have given the preliminary search strategies in Table 1 with the time span for the searches. As the review progresses, we will improve them if needed.

References of the identified studies will also be searched for identifying further studies.

Data collection and analysis

Study selection and extraction of data
Both authors independently of each other will identify the studies for inclusion. We will also list the excluded studies with the reasons for the exclusion. We will resolve any differences in opinion through discussion.

Both authors will independently extract the following data:
(1) Year and language of publication.
(2) Country.
(3) Year of study.
(4) Inclusion and exclusion criteria.
(5) Sample size.
(6) Population characteristics such as age and sex.
(7) Open or laparoscopic cholecystectomy.
(8) Peri‐operative antibiotics.
(9) Drain used or not.
(10) Outcomes (mentioned above).
(11) Methodological quality (described below).

Any unclear or missing information will be sought by contacting the authors of the individual trials. If there is any doubt whether the trials share the same patients ‐ completely or partially (by identifying common authors and centres), the authors of the trials will be contacted to clarify whether the report has been duplicated.

Differences between authors will be resolved by discussion until we reach agreement.

Assessment of methodology quality
The authors will assess the methodological quality of the trials independently, without masking of the study names. The reviewers will follow the instructions given in Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2005). Due to the risk of overestimation of intervention effects in randomised trials with inadequate methodological quality (Schulz 1995; Moher 1998; Kjaergard 2001), we will look at the influence of methodological quality of the trials on the trial results by evaluating the reported randomisation and follow‐up procedures in each trial. If information is not available in the published trial, we will contact the authors in order to assess the trials correctly. We will assess generation of allocation sequence, allocation concealment, blinding, and follow‐up.

Generation of the allocation sequence

• Adequate, if the allocation sequence was generated by a computer or random number table. Drawing of lots, tossing of a coin, shuffling of cards, or throwing dice will be considered as adequate if a person who was not otherwise involved in the recruitment of participants performed the procedure.

• Unclear, if the trial was described as randomised, but the method used for the allocation sequence generation was not described. (The authors will be contacted and attempts will be made to find out the allocation method.)

• Inadequate, if a system involving dates, names, or admittance numbers were used for the allocation of patients. These studies are known as quasi‐randomised and will be excluded from the review.

Allocation concealment

• Adequate, if the allocation of patients involved a central independent unit, on‐site locked computer, or sealed envelopes.

• Unclear, if the trial was described as randomised, but the method used to conceal the allocation was not described.

• Inadequate, if the allocation sequence was known to the investigators who assigned participants or if the study was quasi‐randomised. Such trials will be excluded from the review.

Blinding
Double‐blinding will not be assessed since we expect that there will be no double‐blind trials. However, we will record whether any of the outcomes were assessed by a blinded observer or blinded assessor.

Follow‐up

• Adequate, if the numbers and reasons for dropouts and withdrawals in all intervention groups were described or if it was specified that there were no dropouts or withdrawals.

• Unclear, if the report gave the impression that there had been no dropouts or withdrawals, but this was not specifically stated.

• Inadequate, if the number or reasons for dropouts and withdrawals were not described.

Statistical methods
We will perform the meta‐analyses according to the recommendations of The Cochrane Collaboration (Higgins 2005). We will use the software package RevMan Analyses (RevMan 2003). For dichotomous variables, we will calculate the relative risks with 95% confidence interval. We will use a random‐effects model (DerSimonian 1986) and a fixed‐effect model (Demets 1987). In case of discrepancy between the two models we will report both results; otherwise we will report only the results from the fixed‐effect model.

We will perform subgroup analyses depending on the methodological quality of the trials in order to compare the intervention effect in trials with adequate methodological quality to that of trials with unclear or inadequate methodological quality. Heterogeneity will be explored by chi‐squared test with significance set at P value 0.10, and the quantity of heterogeneity will be measured by I² (Higgins 2002).

We will adopt the 'available‐case analysis' (Higgins 2005). The analysis will be performed on an intention‐to‐treat basis (Newell 1992). In case we find 'zero‐event' trials, we will perform a sensitivity analysis with and without empirical continuity correction factors as suggested by Sweeting et al (Sweeting 2004).

Subgroup analysis
We will perform the following subgroup analyses:
‐ trials with high (adequate allocation concealment) compared to trials with low methodological quality (unclear allocation concealment).
‐ trials using blinded outcome assessment to trials with blinded outcome assessment.
‐ laparoscopic versus open cholecystectomy.
‐ trials that use routine antibiotic prophylaxis (for surgery) compared to those that do not use routine antibiotic prophylaxis.

Bias exploration
We will use a funnel plot to explore publication bias and other bias (Egger 1997; Macaskill 2001). Asymmetry in funnel plot of study size against treatment effect will be used to identify this bias. We will perform linear regression approach described by Egger (Egger 1997) to determine the funnel plot asymmetry.