Temporal trends in the use of direct oral anticoagulants prior to stroke admission

BACKGROUND AND AIMS: This population-based cohort study measures temporal trends in direct oral anticoagulants (DOAC) use before hospital admission among intracranial haemorrhage (ICH) and acute ischemic stroke (AIS) hospitalizations. METHODS: Consecutive patients with ICH and AIS in British Columbia (B.C.), Canada between 2015 and 2023 were identified using ICD-10 diagnosis codes from Canada's Discharge Abstract Database. Linkages were made with outpatient prescriptions. We used Cox regression with cause-specific hazard models to explore differences in hospitalization outcomes by DOAC use adjusting for patient demographics and stroke treatment. RESULTS: 562/5434 (10.3%) of ICH and 3002/35,666 (8.4%) AIS discharges had DOAC exposure prior to hospital admission. Apixaban was the most common DOAC (49.7%) followed by rivaroxaban (41.4%), dabigatran (8.2%) and edoxaban (0.7%). The proportion of patients on DOACs increased every year among ICH patients by 9% (IRR 1.09, 95% CI 1.05-1.13; from 6.3% to 12.0%) and AIS by 10% (IRR 1.10, 95% CI 1.08-1.12; from 5.1% to 10.1%). After adjustment for patient demographics, health authority location in BC, patient comorbidities, ambulance use and reperfusion treatment, ICH patients admitted on DOACs were more likely to die in hospital (HR 1.27, 95% CI 1.06-1.51) and be transferred to long term care (HR 1.36, 95% 1.04-1.76). CONCLUSIONS: The proportion of patients on DOACs at admission doubled between 2015 and 2023. Patients on DOACs were older, had more comorbidities. Patients with ICH on DOAC prior to admission were more likely to have died during their hospitalization or be discharged to long-term care.