Discontinuation of Anticoagulants and Occurrence of Bleeding and Thromboembolic Events in Vitamin K Antagonist Users with a Life-limiting Disease

Background Data on risks and benefits of long-term anticoagulants in patients with a life-limiting disease are limited. This cohort study aims to describe (dis)continuation of anticoagulants and incidences of bleeding and thromboembolic events in vitamin K antagonist (VKA) users with a life-limiting disease. Methods Data from five Dutch anticoagulation clinics were linked to data from Statistics Netherlands and the Netherlands Cancer registry. Prevalent VKA users diagnosed with a pre-specified life-limiting disease between January 1, 2013 and December 31, 2019 were included and followed until December 31, 2019. Bleeding and thromboembolic events were identified by hospitalization data. Cumulative incidences of anticoagulant discontinuation, accounting for death as competing risk, and event rates for both anticoagulant exposed and unexposed person-years (PYs) were determined. Results Among 18,145 VKA users (median age 81 years [IQR: 74-86], 49% females, median survival time 2.03 years [95%CI: 1.97-2.10]), the most common life-limiting diseases were heart disease (60.0%), hip fracture (18.1%), and cancer (13.5%). One year after diagnosis, the cumulative incidence of anticoagulant discontinuation was 14.0% (95%CI: 13.5-14.6). Over 80% of patients continued anticoagulant therapy until the last month before death, with median 14 days between discontinuation and death. Event rates per 100 PYs (95%CI) were comparable during anticoagulant use and after discontinuation for bleeding 2.6 (2.4-2.8) versus 2.1 (1.5-2.8), venous thromboembolism 0.2 (0.1-0.2) versus 0.4 (0.2-0.7), and arterial thromboembolism 3.1 (2.9-3.3) versus 3.3 (2.6-4.2). Conclusion Most VKA users with a life-limiting disease continued anticoagulant treatment during their last phase of life, with similar rates of bleeding and thromboembolic events during use and after discontinuation.