Temporal patterns and risk factors of clozapine-associated neutropenia: a nationwide population-based cohort study

Abstract

Background Clozapine is the most effective antipsychotic for treatment-resistant schizophrenia, but neutropenia concerns continue to limit its use. Large real-world evidence on temporal risk patterns, re-initiation, and patient-level determinants remains limited. Methods We conducted a nationwide cohort study using linked claims and laboratory data from Taiwan's National Health Insurance database (2014–2023). New clozapine users and re-initiators after a ≥42 day interruption were identified. Neutropenia was classified as isolated minor, serious with treatment cessation, or serious without cessation, using absolute neutrophil count thresholds when available and white blood cell counts otherwise. Incidence rates were estimated per 1000 person-years. Temporal changes were examined using joinpoint regression, and risk factors using multivariable cause-specific Cox models with time-varying medication exposures. Findings Among 12,810 new users, isolated minor neutropenia occurred at 10.5 per 1000 person-years, and serious neutropenia leading to clozapine cessation at 1.7 per 1000 person-years. Joinpoint regression identified early inflection points at week 14 for isolated minor neutropenia, week 15 for serious neutropenia with cessation, and week 6 for serious neutropenia without cessation. Older age and lower baseline white blood cell count were consistently associated with higher risks. Concomitant medication associations were generally stronger for minor than serious events. Re-initiators showed risk patterns comparable to new users, and absolute neutrophil count-based sensitivity analyses yielded consistent results. Interpretation Clozapine-associated neutropenia was uncommon and concentrated early after initiation, with risk stratified by baseline haematological status and age. These findings support risk-informed monitoring strategies. Funding National Health Research Institutes, Taiwan.