Excess risk of bleeding in patients with venous thromboembolism on direct oral anticoagulants during initial and extended treatment versus population controls

BACKGROUND: The risk of major bleeding from anticoagulant treatment is influenced by both the treatment and the patient's baseline risk, which is often disregarded. OBJECTIVES: To determine the excess bleeding risk in venous thromboembolism (VTE) cases during initial (0 to 6 months) and extended (6 months to 5 years) treatment compared to matched population controls without VTE or anticoagulant treatment, overall, and stratified by sex and age. METHODS: Cancer-free patients with VTE treated with direct oral anticoagulants from 2014 to 2020, along with propensity score-matched controls, were identified from nationwide Swedish registers. Excess risk of major bleeding was assessed using the incidence rate difference (IRD) calculated by subtracting the control bleeding rate from the case bleeding rate. RESULTS: The matched cohort comprised 36,115 VTE cases and 36,115 controls. During initial treatment, 388 VTE cases (1.07%) and 103 controls (0.29%) experienced bleeding, IRD: 2.19 (95% confidence interval 1.89-2.49) per 100 person-years. Following rematching at 6 months, 139 cases (0.70%) and 214 controls (1.08%) experienced bleeding, IRD: 0.70 (0.52-0.89). During initial treatment, females had a higher excess bleeding risk than males, with male IRD: 1.73 (1.34-2.12) and female IRD: 2.69 (2.23-3.15). Excess bleeding risk was highest in the oldest patient population. In extended treatment, excess bleeding was not dependent on sex-male IRD: 0.60 (0.35-0.85), female IRD: 0.81 (0.54-1.08)-and did not increase with age. CONCLUSION: The excess bleeding risk from anticoagulant treatment was high during initial treatment, particularly among females and the elderly, but lower and not influenced by sex or age during extended treatment.