60 years of placebo-controlled antipsychotic drug trials in acute schizophrenia: Meta-regression of predictors of placebo response

Abstract

Objective: A recent meta-regression had shown that the degree of placebo response, which has increased over the decades, is the major predictor of drug-placebo differences in antipsychotic drug trials in acutely ill patients with schizophrenia. Drug response, however, had remained stable. In the current meta-regression we explored the factors that are associated with placebo-response.

Method: We searched multiple electronic databases, ClinicalTrials.gov and the FDA website for randomized, placebo-controlled, antipsychotic drug trials in patients with acute exacerbations of schizophrenia. The outcome was the degree of placebo response measured by the BPRS or PANSS change from baseline to endpoint. 26 patient-, design-, and drug-related potential predictors of placebo response were analyzed by univariable and multivariable meta-regressions.

Results: 167 double-blind randomized controlled trials with 28,102 participants were included. The mean PANSS change from baseline was 6.25 (95% CI 4.64,7.85). More recent publication year, larger study sample size, more study sites, use of the PANSS rather than the BPRS scale to measure response, shorter wash-out phases, shorter study duration, lower mean age and shorter duration of illness were associated with larger placebo response in univariable analyses. In a multivariable analysis only the number of study participants and mean participant age had an impact on placebo response.

Conclusions: The degree of placebo response is moderated by a number of design and patient-related factors. These explanatory variables of placebo response are only in part identical with those that moderated drug-placebo differences.

Keywords: Antipsychotics; Clinical trials; Meta-regression; Placebo response; Predictors.