The effects of antipsychotic discontinuation or maintenance on the process of recovery in remitted first-episode psychosis patients – A systematic review and meta-analysis of randomized controlled trials

Abstract

Background The optimal duration of antipsychotic treatment following remission of first-episode psychosis (FEP) is uncertain, considering potential adverse effects and individual variability in relapse rates. This study aimed to investigate the effect of antipsychotic discontinuation compared to continuation on recovery in remitted FEP patients. Methods CENTRAL, MEDLINE (Ovid), Embase, and PsycINFO databases were searched on November 2, 2023, with no language restrictions. RCTs evaluating antipsychotic discontinuation in remitted FEP patients were selected. The primary outcome was personal recovery, and secondary outcomes included functional recovery, global functioning, hospital admission, symptom severity, quality of life, side effects, and employment. Risk of bias was assessed using the Cochrane risk-of-bias tool 2, and the certainty of evidence was evaluated with GRADE. Meta-analysis used a random-effect model with an inverse-variance approach. Results Among 2185 screened studies, 8 RCTs (560 participants) were included. No RCTs reported personal recovery as an outcome. Two studies measured functional recovery, and discontinuation group patients were more likely to achieve functional recovery (RR 2.19; 95% CIs: 1.13, 4.22; I2 = 0%; n = 128), although evidence certainty was very low. No significant differences were found in hospital admission, symptom severity, quality of life, global functioning, or employment between the discontinuation and continuation groups. Conclusions Personal recovery was not reported in any antipsychotic discontinuation trial in remitted FEP. The observed positive effect of discontinuation on functional recovery came from an early terminated trial and an RCT followed by an uncontrolled period. These findings should be interpreted cautiously due to very low certainty of evidence.