Oral Anticoagulation and Risk of Adverse Clinical Outcomes in Venous Thromboembolism

Importance: Over the past decade, there has been a considerable shift in the use of pharmacologic agents for venous thromboembolism (VTE), with direct oral anticoagulants replacing warfarin as the drugs of choice for VTE recurrence prevention; however, evidence from head-to-head comparison studies remains limited. Objective: To compare the effectiveness and safety of 3 common oral anticoagulants (apixaban, rivaroxaban, and warfarin) in patients with VTE. Design, Setting, and Participants: This population-based cohort study used Medicare and 2 commercial insurance databases from 2016 up to 2024 to identify patients 18 years and older who initiated an oral anticoagulant following VTE and had at least 1 year of continuous insurance enrollment before the index date. Exposure: Initiation of apixaban, rivaroxaban, or warfarin within 30 days after VTE discharge. Main Outcomes and Measures: The primary effectiveness outcome was hospitalization for recurrent VTE. The primary safety outcome was major bleeding. Patients were followed up from treatment initiation until outcome occurrence, treatment discontinuation/switch, disenrollment, death, or end of available data. Propensity score-matching weights were used to adjust for confounding. Weighted Cox proportional hazard models estimated weighted hazard ratios (HRs) and 95% CIs. Results: Among 163 593 eligible individuals (mean SD age, 71.4 13.5 years; 56.7% female), 58.5% initiated apixaban, 25.7% initiated rivaroxaban, and 15.8% initiated warfarin. Overall, 3270 hospitalizations for recurrent VTE and 4229 hospitalizations for bleeding events occurred. Compared with warfarin, patients taking apixaban (HR, 0.67; 95% CI, 0.61-0.75) and rivaroxaban (HR, 0.77; 95% CI, 0.69-0.87) had a lower risk of recurrent VTE. Apixaban showed a further decrease in risk compared with rivaroxaban (HR, 0.87; 95% CI, 0.78-0.96). Patients taking apixaban also had a lower risk of major bleeding compared with warfarin (HR, 0.70; 95% CI, 0.64-0.76) and rivaroxaban (HR, 0.69; 95% CI, 0.63-0.75). No difference in bleeding risk was observed between rivaroxaban and warfarin (HR, 1.02; 95% CI, 0.92-1.12). These findings were consistent across subgroups defined by age, sex, cancer, chronic kidney disease, bleeding history, and frailty. Conclusions and Relevance: In this cohort study of patients with VTE who initiated an oral anticoagulant, apixaban was associated with a lower risk of VTE recurrence and major bleeding compared with rivaroxaban and warfarin. These results provide evidence to guide the selection of appropriate initial oral anticoagulant regimens for adult patients with VTE.