Systematic review with meta-analysis: placebo rates in induction and maintenance trials of Crohn's disease

Background Minimising placebo response is essential for drug development. Aim To conduct a meta-analysis to determine placebo response and remission rates in trials and identify the factors affecting these rates. Methods MEDLINE, EMBASE and CENTRAL were searched from inception to April 2014 for placebo-controlled trials of pharmacological interventions for Crohn's disease. Placebo response and remission rates for induction and maintenance trials were pooled by random-effects and mixed-effects meta-regression models to evaluate effects of study-level characteristics on these rates. Results In 100 studies containing 67 induction and 40 maintenance phases and 7638 participants, pooled placebo remission and response rates for induction trials were 18% [95% confidence interval (CI) 16-21%] and 28% (95% CI 24-32%), respectively. Corresponding values for maintenance trials were 32% (95% CI 25-39%) and 26% (95% CI 19-35%), respectively. For remission, trials enrolling patients with more severe disease activity, longer disease duration and more study centres were associated with lower placebo rates, whereas more study visits and longer study duration was associated with higher placebo rates. For response, findings were opposite such that trials enrolling patients with less severe disease activity and longer study duration were associated with lower placebo rates. Placebo rates varied by drug class and route of administration, with the highest placebo response rates observed for biologics. Conclusions Placebo rates vary according to whether trials are designed for induction or maintenance and the factors influencing them differ for the endpoints of remission and response. These findings have important implications for clinical trial design in Crohn's disease.