Antipsychotic exposure and infection risk in people with schizophrenia spectrum disorders during the COVID-19 pandemic: a Danish nationwide registry study.

Background: Infection risk and mortality are increased in schizophrenia spectrum disorders, which was corroborated during the COVID-19 pandemic. However, evidence is lacking regarding the additional impact of antipsychotic drugs, and the highly debated safety of clozapine treatment during large-scale infection outbreaks. Therefore, we aimed to investigate risk of COVID-19 and non-COVID respiratory infections during exposure to antipsychotics.

Methods: We used several nationwide Danish registers (National Prescription Registry, National Hospital Registry, Psychiatric Research Register, Microbiology Database, Vaccination Registry, Cause of Death Registry, and Database for Labour market Research) to investigate all individuals aged 18 years or older with a schizophrenia spectrum disorder (ICD-10: F20-F29) living in Denmark between Jan 1 and March 1, 2020. Antipsychotic exposure groups were defined as prevalent-users and incident-users. The full observation period was March 1, 2020 to Dec 31, 2021. Antipsychotic exposure was defined in a time-varying manner and compared with non-exposure. Risk was calculated for mild infection outcomes (positive SARS-CoV-2 PCR and anti-infective drug prescriptions) and severe infection outcomes (hospitalisation and death) related to COVID-19 and non-COVID-19 respiratory infections. Outcomes were adjusted for demographics, socio-economic factors, and comorbidity.

Findings: Of 85 083 individuals (44 293 men [52·1%] and 40 790 women [47·9%], median age 45·8 years [IQR 31·1-60·2]) with pre-existing schizophrenia spectrum disorders, 30 984 had antipsychotic exposure periods. Ethnicity data were not available. During antipsychotic exposure compared with non-exposed periods, assessing mild infection outcomes, risk of a positive SARS-CoV-2 test was decreased (hazard ratio 0·91 [95% CI 0·85-0·97]) and risk of redeeming an anti-infective drug was not statistically significantly different (1·01 [0·97-1·06]). For severe infection outcomes, COVID-19-related hospitalisation risk was increased (1·28 [1·07-1·52]) although COVID-19-related death was not statistically significantly increased (1·24 [0·82-1·86]). For non-COVID-19 respiratory infections, risk was increased both for hospitalisation (1·61 [1·44-1·79]) and death (1·61 [1·18-2·21]). Specifically, COVID-19 hospitalisation risk was increased in individuals older than 70 years, and non-COVID-19 hospitalisation risk increased in individuals older than 40 years and death risk in age groups of 50-59 years and 70-79 years. Based on homogeneity testing, no apparent excess risk of any outcome was observed with clozapine exposure compared with other antipsychotics.

Interpretation: During antipsychotic exposure compared with unexposed periods, risk of severe infection outcomes increases. It seems reasonable to initiate infection countermeasures, such as pneumococcal vaccination, in people older than 40 years with schizophrenia spectrum disorders, who commence or are treated with antipsychotics. We do not suggest the avoidance of specific antipsychotics but rather adherence to treatment guidelines and a call for increased vigilance regarding this at-risk group.

Funding: Mental Health Services of the Capital Region of Denmark.