Clozapine use and risk of infections in patients with schizophrenia: a population-based cohort study in Hong Kong

Background: Infection is the leading natural cause of mortality in patients with schizophrenia, and use of antipsychotics might be associated with an increased risk of infections. This study aimed to examine the association of clozapine use with the incidence of various infectious diseases. Methods: In this population-based cohort study, we used territory-wide electronic health records from the Hong Kong Hospital Authority. Individuals aged 18 years or older with a diagnosis of schizophrenia (ICD-9-CM code 295) based on records from Jan 1, 2002, to Dec 31, 2023 were considered for inclusion. We included patients who used clozapine or olanzapine continuously for at least 90 days and excluded those who had previously used fewer than two other antipsychotic medications, to identify a cohort of patients with long-term antipsychotic use who had treatment resistance. Olanzapine was chosen as a comparator due to its similar chemical structure and often overlapping indication with clozapine. The primary outcome was the occurrence of any infectious disease, identified by a range of ICD-9-CM codes, with subtypes analysed as secondary outcomes. Propensity score-based inverse probability of treatment weighting was applied to balance covariates. Weighted hazard ratios (HRs) for infection in patients using clozapine versus those using olanzapine were estimated by Cox proportional hazards models. Weighted absolute differences in incidence rate were also estimated. People with lived experience of schizophrenia were not involved in the design or conduct of the study. Findings: We identified 53 092 individuals with clozapine or olanzapine initiation between Jan 1, 2004 and Dec 31, 2023, with 2002–03 used as a wash-out period to exclude individuals having used the two antipsychotics before. After exclusion of participants without a schizophrenia diagnosis and who did not meet the other selection criteria, the final cohort consisted of 11 051 patients (1450 using clozapine and 9601 using olanzapine). Overall, 6031 (54·6%) patients were female and 5020 (45·4%) were male. Mean age was higher in the olanzapine group (45·33 years [SD 14·55]) than in the clozapine group (40·59 years [12·26]). No ethnicity data were available. The incidence of infection overall was increased in patients using clozapine (weighted incidence rate 7·26 per 100 person-years) compared with those using olanzapine (6·00 per 100 person-years; weighted HR 1·25 [95% CI 1·13–1·39]). The weighted absolute difference in incidence rate was 1·27 infections per 100 person-years (95% CI 0·55–2·04). The risk was notably higher in older age ranges, with weighted HRs rising from 1·24 (1·08–1·42) in those aged 18–44 years (weighted absolute difference 0·81 per 100 person-years [0·13–1·55]) to 1·41 (1·15–1·72) in those aged 45–54 years (2·23 per 100 person-years [0·82–3·86]), and 1·45 (1·14–1·84) in those aged 55 years or older (4·70 per 100 person-years [1·29–8·45]). Secondary analyses identified upper and lower respiratory tract infections and gastrointestinal infections as primary contributors to the observed increased incidence associated with clozapine versus olanzapine use. Interpretation: Clinicians should balance the therapeutic benefits of clozapine with infection control measures, including regular monitoring and preventive strategies. Funding: InnoHK initiative of the Innovation and Technology Commission of Hong Kong. Translation: For the Chinese translation of the abstract see Supplementary Materials section.