Association between preadmission antipsychotic medication use and 30-day mortality in acute stroke patients in various adjusted models.

Model 1: adjusted for age group and calendar period. Overall estimates adjusted for gender in all models.†Model 2: further adjusted for type of stroke, former stroke, pre-stroke drug use (lipid-lowering drugs, antihypertensive drugs, antidiabetic drugs, platelet inhibitors) and education level.‡Model 3: further adjusted for year of admission, severity of stroke, and modified Charlson’s index (cerebrovascular disorders excluded). No interaction between genders in model 3 (p = 0.79).