Comparative Pharmacovigilance Analysis of Pulmonary Embolism Across Antipsychotics in VigiBase

BACKGROUND: Pulmonary embolism (PE) may be associated with antipsychotics or an underlying serious mental illness (SMI). This pharmacovigilance study in VigiBase studied the association between 6 antipsychotics and PE by using statistical analyses: (1) from VigiBase, or (2) developed by the authors based on fatal outcomes. METHODS: VigiBase PE data are included from the onset through July 1, 2025, for 6 antipsychotics (clozapine, olanzapine, risperidone, quetiapine, aripiprazole, and haloperidol). The disproportionality analyses used the information component (IC) and its confidence intervals (CIs). An IC=0 indicates no association. Logistic regression models explored PE fatal versus nonfatal outcomes. FINDINGS: Olanzapine had the highest IC=1.2 (1.1 to 1.3) with 592 PE cases (257 expected). Clozapine's IC=0.93 (0.85 to 1.0) with 1314 PE cases (688 expected). Other antipsychotics provided very low signals. The percentage of fatal outcomes was 41% (538/1313) on clozapine, 28% (161/579) on olanzapine but no clozapine, and 21% (225/1092) on other antipsychotics but no clozapine or olanzapine. In the logistic regression model of 2984 PEs, the adjusted odds ratios (ORs) for fatal outcomes were 1.5 (95% CI: 1.2-1.9) for olanzapine and 2.5 (95% CI: 2.1-3.0) for clozapine. IMPLICATIONS: This large sample from VigiBase, despite its limitations, including using voluntary reports, provided a PE signal that was statistically significant for olanzapine and clozapine, while not significant for other antipsychotics. Future studies will need to explore the contribution of SMI versus olanzapine and clozapine. Clinicians need to be aware that the proportion of fatal outcomes is around 40% in patients on clozapine and those on complex antipsychotic polytherapy.