This meta-analysis found that broad-spectrum antibiotic therapy improved clinical outcomes for patients with active Crohn's disease compared with placebo. Limitations of the review, including the comprehensiveness of the search and the lack of consideration of the effect of individual trial quality on the results, mean that the reliability of the results is uncertain and they should be interpreted with caution.

To evaluate the effectiveness of broad-spectrum antimicrobial therapy in patients with Crohn's disease.

MEDLINE and Scopus were searched from 1966 to 3 May 2006 for relevant studies; the search terms were listed. Bibliographies of identified articles were also checked. Only English language articles were eligible for inclusion.

Study designs of evaluations included in the review

Randomised controlled trials were eligible for inclusion.

Specific interventions included in the review

Studies of any broad-spectrum antibiotic, including metronidazole, ciprofloxacin and cotrimoxazole were eligible for inclusion. Studies of antimycobacterial agents were not eligible for inclusion. The included studies evaluated metronidazole, cotrimoxazole and ciprofloxacin, and combinations of metronidazole and cotrimoxazole, and of metronidazole and ciprofloxacin. The dosage varied between trials and treatment duration ranged from 2 to 24 weeks. In some, but not all included trials, participants were treated with concomitant medication. Studies had to include a control arm that received placebo to be eligible for inclusion.

Participants included in the review

Participants of any age or either gender, with active luminal Crohn's disease, were eligible for inclusion. The mean age of participants in the included studies ranged from 27.5 to 43.5 years, and 42% were male.

Outcomes assessed in the review

Studies evaluating clinical improvement were eligible for inclusion. This was defined as response, remission, or improvement in clinical and/or laboratory findings. The included trials used different definitions of clinical improvement. Studies that evaluated microbial overgrowth, adverse effects of antibiotic therapy, or the efficacy of antibiotics in preventing post-operative recurrence of Crohn's disease were not eligible for inclusion.

How were decisions on the relevance of primary studies made?

Three reviewers independently made decisions on the relevance of studies identified by the search. Any disagreements were resolved by discussion and consensus.

The authors did not state that they assessed validity, although blinding was discussed in the text.

Three reviewers independently extracted the data from the individual trials. Results data were extracted as 2x2 tables.

How were the studies combined?

Mantel-Haenszel summary odds ratios (ORs) were calculated using a fixed-effect model. Funnel plots were used to investigate publication bias.

How were differences between studies investigated?

The Breslow-Day test was used to test for heterogeneity between studies.

Six trials (804 participants) were included in the review.

Five of the trials were double-blind; in the sixth trial the blinding procedure was not stated. Antibiotic therapy was associated with higher odds of clinical improvement (OR 2.26, 95% CI: 1.68, 3.04, p<0.001) with no evidence of heterogeneity. The funnel plot suggested an absence of publication bias.

Broad-spectrum antibiotics are associated with improved clinical outcomes in patients with Crohn's disease compared with placebo.

The review question and inclusion criteria were clearly stated. The search was restricted to two databases. Non-English language articles were omitted from the search, and the authors did not state that they attempted to identify unpublished studies. Validity, other than blinding, was not assessed. Each of these factors could have either introduced bias or reduced the reliability of the results of the review. An effort was made to minimise subjectivity in the study selection and data extraction phases of the review.

The meta-analysis was appropriately carried out, although the differing antibiotics were all combined in one analysis, and no subgroup analyses, stratified by drug type, were presented. There was no discussion of the appropriateness of analysing differing clinical outcomes as one category. Although there was no evidence of statistical heterogeneity, the authors did not investigate clinical heterogeneity, despite noting that there could be clinical reasons for differing responses to therapy (such as the site and duration of the disease, and the duration and type of therapy). The reliability of the authors' conclusions is uncertain given the limitations of the review.

Practice: The authors did not state any implications for practice.

Research: The authors stated that further trials are needed, but did not explain why this was so.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.