Placebo effects in adult and adolescent patients with schizophrenia: combined analysis of nine RCTs

OBJECTIVE: To examine characteristics of placebo responders and seek optimal criteria of early improvement with placebo for predicting subsequent placebo response in patients with schizophrenia. METHOD: Data of 672 patients with schizophrenia randomized to placebo in nine double-blind antipsychotic trials were analyzed. Multiple logistic regression analyses were conducted to examine associations between placebo response at week 6 (i.e., a ≥ 25% reduction in the Positive and Negative Syndrome Scale [PANSS] score) and gender, age, study locations, baseline PANSS total or Marder 5-Factor scores, and per cent PANSS score reduction at week 1. Predictive power of improvement at week 1 for subsequent response was investigated; sensitivity and specificity of incremental 5% cutoff points between 5% and 25% reduction in the PANSS total score at week 1 were calculated. RESULTS: Per cent PANSS total score reduction at week 1 and lower PANSS Marder disorganized thought scores at baseline were significantly associated with subsequent placebo response. A 10% reduction in a per-protocol analysis or a 15% reduction in last-observation-carried-forward analysis in the PANSS total score at week 1 showed the highest predictive power. CONCLUSION: These findings are informative to identify potential placebo responders at the earliest opportunity for optimal trial design for schizophrenia.