Prevalence, correlates, tolerability-related outcomes, and efficacy-related outcomes of antipsychotic polypharmacy: a systematic review and meta-analysis
Mikkel Højlund,Ole Köhler-Forsberg,Anton T Gregersen,Christopher Rohde,Angelina I Mellentin,Simon J Anhøj,Adam F Kemp,Nina B Fuglsang,Anne Christine Wiuff,Louise Nissen,Marc A Sørensen,Nanna M Madsen,Christina B Wagner,Armaghan Agharazi,Cecilie Søndergaard,Marie Sandmark,Jana Reinhart,Christoph U Correll
DOI: https://doi.org/10.1016/s2215-0366(24)00314-6
IF: 77.056
2024-11-18
The Lancet Psychiatry
Abstract:Summary Background Antipsychotic polypharmacy remains a clinical reality, despite an increased risk of adverse events and little evidence of additional efficacy compared with antipsychotic monotherapy. In this systematic review and meta-analysis, we aimed to provide a comprehensive assessment of antipsychotic polypharmacy prevalence, trends, and correlates across mental disorders. Methods We searched MEDLINE and Embase from Jan 1, 2009 to April 30, 2024, for any original study (observational and interventional) reporting antipsychotic polypharmacy prevalence in populations with mental disorders or use of antipsychotics, regardless of age or diagnosis. Relevant studies before May 1, 2009, were identified from two previous systematic reviews of antipsychotic polypharmacy prevalence. Pooled antipsychotic polypharmacy prevalence was estimated using random-effects meta-analysis. Using subgroup and mixed-effects meta-regression analyses, we sought to identify relevant correlates of antipsychotic polypharmacy. People with lived experience were not involved in the project. This review is registered with PROSPERO (CRD42022329953). Findings We analysed 517 studies with 599 individual timepoints reporting on 4 459 149 individuals (mean age 39·5 years [range 6·4–86·3]; data on sex and ethnicity were infrequently reported). Most studies included patients with schizophrenia spectrum disorders (SSDs; k=270, 52%). Overall, 24·8% (95% CI 22·9–26·7) of the populations received antipsychotic polypharmacy, ranging from 33·2% (30·6–36·0) in people with SSDs to 5·2% (4·0–6·8) in people with dementia. Antipsychotic polypharmacy prevalence varied by region from 15·4% (95% CI 12·9–18·2) in North America to 38·6% (27·7–50·6) in Africa. Overall antipsychotic polypharmacy prevalence increased significantly from 1970 to 2023 (β=0·019, 95% CI 0·009–0·029; p=0·0002) and was higher in adults than in children and adolescents (27·4%, 95% CI 25·2–29·8 vs 7·0%, 4·7–10·3; p<0·0001) and among inpatients than among outpatients (31·4%, 27·9–35·2 vs 19·9%, 16·8–23·3; p<0·0001). Compared with antipsychotic monotherapy, antipsychotic polypharmacy was associated with an increased risk of relapse (relative risk [RR] 1·42, 95% CI 1·04–1·93; p=0·028), psychiatric hospitalisation (1·24, 1·12–1·38; p<0·0001), worse global functioning (standardised mean difference [SMD] –0·31, 95% CI –0·44 to –0·19; p<0·0001), and more adverse events, including extrapyramidal symptoms (RR 1·63, 95% CI 1·13–2·36; p=0·0098), dystonia (5·91, 1·20–29·17; p=0·029), anticholinergic use (1·91, 1·55–2·35; p<0·0001), higher side-effect scores (SMD 0·33, 95% CI 0·24–0·42; p<0·0001), longer corrected QT interval (0·24, 0·23–0·26; p<0·0001), and greater all-cause mortality risk (RR 1·19, 95% CI 1·00–1·41; p=0·047). Interpretation The prevalence of antipsychotic polypharmacy has increased globally over the past 50 years and is particularly high in patients with SSDs. Prescription of antipsychotic polypharmacy is associated with greater illness severity and poorer outcomes than is antipsychotic monopharmacy but does not resolve these issues. Furthermore, antipsychotic polypharmacy is associated with higher side-effect burden, including all-cause mortality. Funding None.
psychiatry