Correlates and predictors of antipsychotic drug polypharmacy in real-life settings: Results from a nationwide cohort study

Abstract Reasons for using antipsychotic polypharmacy (APP) in routine clinical practice, despite a potentially unfavorable risk-benefit ratio, are poorly understood. This research aimed to determine (1) if severe courses of schizophrenia were associated with APP and (2) if a schizophrenia-related acute event would predict a switch to APP in the short term. Observational prospective data (at baseline and 6 months) were drawn from a French nationwide cohort (“ Cohorte Générale Schizophrénie ”), which included 1859 inpatients and outpatients with schizophrenia. APP was defined as the prescription of ≥ 2 antipsychotic drugs (there being different active substances). Early-onset schizophrenia, legal guardianship, higher lifetime maximal severity of illness and comorbid antisocial personality were used as proxies for severe courses of schizophrenia. Schizophrenia-related acute events included hospitalization and recent suicide attempts. Logistic regression models were used to determine (1) whether the use of APP at baseline (vs. monotherapy) was associated with a severe course of schizophrenia or not, independent of acute events, and (2) if a switch to APP at 6 months (vs. remaining on monotherapy) was associated with acute events, independent of severe courses of schizophrenia. Increased odds of APP use at baseline were independently associated with legal guardianship (OR = 1.6; 95%CI = 1.3, 2.0) and higher lifetime maximum severity of illness (OR = 1.3; 95%CI = 1.2, 1.5). A switch to APP at 6 months was predicted by a hospitalization occurring since baseline (OR = 6.1; 95%CI = 3.9, 9.4). In routine clinical practice, APP is more likely prescribed to patients with severe courses of illness, possibly indicating the difficulty to manage these patients.