Meta-Analysis: Placebo Response and Its Determinants in Functional Dyspepsia.

Abstract

INTRODUCTION\nPlacebo response rates are relatively higher in clinical trials of disorders of brain-gut interaction. However, placebo response in functional dyspepsia (FD) has not been well described. Minimizing placebo response is important in drug development. We therefore conducted a meta-analysis to determine placebo response in trials for FD and to identify factors affecting placebo response rates.\n\n\nMETHODS\nPubMed, EMBASE, and the Cochrane Central Register of Controlled Trials were searched to identify double-blinded randomized controlled trials (RCTs) comparing medication with placebo in patients with FD. Both symptom improvement and complete relief were considered as separate primary endpoints in the analysis. Proportions of placebo patients experiencing any symptom improvement or complete relief were calculated. Dropouts after randomization for any reason were assumed to represent treatment failure for data extraction and analysis. Placebo response was pooled by a random-effects model, and effects of trial characteristics on the magnitude of placebo response were evaluated.\n\n\nRESULTS\nIn 58 eligible placebo-controlled RCTs of FD from 52 selected citations, 6,732 of 17,890 participants in all trials received placebo. Pooled placebo response rates for symptom improvement and complete relief were 44.3% and 15.6%, respectively. The placebo response rate was lower when improvements were assessed for ≥8 weeks. Trials assessing complete symptom relief showed lower placebo response rates even in trials for <8 weeks.\n\n\nDISCUSSION\nOur systematic review and meta-analysis showed that pooled placebo response rates in double-blinded RCTs of FD depended on efficacy criteria. Trials assessing complete symptom relief showed stable low placebo response rates in short-term trials.