| Abstract: | Objective: To compare the risk of antipsychotic-related seizure (ARS) by identifying seizures first diagnosed within 12 months after starting new antipsychotics, using a 12-year total population health claims database from Taiwan. Methods: Seizure events were identified through emergency department visits or hospitalization with a diagnosis of convulsion (ICD-9-CM: 780.3) or epilepsy (ICD-9-CM: 345). Subjects had an ICD-9-CM diagnosis of schizophrenia, bipolar disorders, or major depressive disorders. Incidence rates of ARS were calculated by person-years of exposure. The ARS risk, adjusted for patient characteristics and medical conditions, of individual antipsychotics versus risperidone was examined by high-dimensional propensity score stratification analyses, followed by sensitivity analyses. Results: The overall 1-year incidence rate of ARS was 9.6 (95% CI, 8.8-10.4) per 1,000 person-years (550 ARS events among 288,397 new antipsychotic users). First-generation antipsychotics were marginally associated with a higher ARS risk than second-generation antipsychotics (adjusted hazard ratio [aHR] = 1.34; 95% CI, 0.99-1.81; P = .061). Most antipsychotics, first-or second-generation, had comparable ARS risks versus risperidone. Notably, clozapine (aHR = 3.06; 95% CI, 1.40-6.71), thioridazine (aHR = 2.90; 95% CI, 1.65-5.10), chlorprothixene (aHR = 2.60; 95% CI, 1.04-6.49), and haloperidol (aHR = 2.34; 95% CI, 1.48-3.71) had higher ARS risks than risperidone, whereas aripiprazole (aHR = 0.41; 95% CI, 0.17-1.00; P = .050) had a marginally lower ARS risk. Sensitivity analyses largely confirmed such findings. Conclusions: Higher vigilance for ARS is warranted during use of clozapine, chlorprothixene, thioridazine, and haloperidol. The possible lower ARS risk associated with aripiprazole can be clinically significant but needs to be confirmed by larger-scale systematic studies. The comparative ARS risks of antipsychotics supplement empirical knowledge for making judicious choices in prescribing antipsychotics. |
