Efficacy and acceptability of interventions to reduce antipsychotic polypharmacy: A systematic review and meta-analysis of randomized clinical trials

Background: Antipsychotic polypharmacy (APP) is frequent but evidence-based guidelines on reducing APP to antipsychotic monotherapy (APM) are sparse. We aimed to systematically review clinical interventions randomizing patients to reducing APP to APM versus continuing APP. Methods: Systematic literature review searching Medline and Embase (latest search January 10, 2024) for randomized clinical trials (RCTs) studying interventions comparing individuals randomized to reduction of APP to APM with individuals continuing on APP. Two independent reviewers performed the literature screening, data extraction, and risk of bias assessment (RoB2). We performed random effects meta-analyses on the main outcome all-cause discontinuation/”acceptability” of the treatment strategy and secondary outcomes change in psychopathology, functional level, and side effects. Results: The search identified 4672 hits, whereof 8 trials (N = 1204, 6 patient-level RCTs and 2 cluster-RCTs) were included, primarily in patients with schizophrenia. All trials were associated with high risk of bias. Compared to APP continuation, reduction to APM was associated with no significant change in all-cause discontinuation (studies = 6, n = 455, RR = 1.48, 95%CI = 0.74–2.95, I² = 78 %) or inefficacy-related discontinuation (studies = 5, n = 351, RR = 1.60, 95%CI = 0.46–5.55, I² = 70 %). Patients randomized to APM showed a trend towards greater reduction in psychopathology (studies = 5, n = 244, SMD = -0.24, 95%CI = -0.49, 0.02, I² = 0 %) but no difference in functional level nor side effects. The cluster-RCTs found that interventions at the departmental level can result in lower rates of APP. Conclusion: Although switching patients from APP to APM can be a viable approach, too few RCTs exist on this important topic. Clinicians need to evaluate potential benefits and risks of APP and APM on an individual basis. Prospero registration: CRD42022329955.