Abstract
IMPORTANCE: It has been hypothesized that the association between acute-phase treatment with antipsychotic (vs placebo) and outcomes might be larger in individuals who were not recently treated compared to recently treated individuals. However, evidence is still lacking. OBJECTIVE: To compare the association between antipsychotic (vs placebo) acute-phase treatment and outcomes in individuals who were not recently treated to recently treated individuals. DATA SOURCES: The Yale University Open Data Access Project's database was searched from inception to April 16, 2024 (PROSPERO CRD42021224350). STUDY SELECTION: We included placebo-controlled antipsychotic acute-phase randomized clinical trials with participants with schizophrenia or schizoaffective disorder aged 18 years and older. Participants were divided into 2 groups: recently treated (treated with an antipsychotic before the start of randomized treatment) and not recently treated (not treated for ≥4 weeks before the start of randomized treatment). DATA EXTRACTION AND SYNTHESIS: Multiple linear regression models were implemented to estimate difference in mean difference (DMD), defined as mean difference (antipsychotic vs placebo) in not recently treated minus mean difference in recently treated, and 95% confidence intervals in each study. DMD and 95% confidence intervals were synthesized across studies using a random effects meta-analysis model. MAIN OUTCOMES AND MEASURES: The primary outcome was overall symptoms (Positive and Negative Syndrome Scale [PANSS] score after 6 weeks). RESULTS: A total of 470 trials were identified, of which 12 were eligible for the analysis. A total of 692 individuals were included in the not recently treated group and randomized to antipsychotic (n = 502) or placebo (n = 190), and 2089 individuals were included in the recently treated group and randomized to antipsychotic (n = 1513) or placebo (n = 576). Overall median (IQR) age of included individuals was 38 (30-48) years, and 998 individuals (35.9%) were female. No evidence of difference in the antipsychotic vs placebo outcomes was detected across the not recently treated and recently treated individuals (DMD, 0.8 PANSS points; 95% CI, -3.6 to 5.2). Not recently treated individuals had better outcomes with both the antipsychotic and placebo compared to recently treated individuals. CONCLUSIONS AND RELEVANCE: In this individual participant data meta-analysis, the association between antipsychotic (vs placebo) acute-phase treatment and outcomes may be similar in individuals with and without recent treatment. Individuals who were not recently treated had better outcomes with both the antipsychotic and placebo compared with recently treated individuals.