Risk of gastrointestinal rebleeding in direct oral anticoagulant-treated population

INTRODUCTION: We performed a retrospective analysis of rebleeding events and their association with direct oral anticoagulant (DOAC) discontinuation versus resumption in patients admitted with an index episode of gastrointestinal bleeding (GIB) while on oral anticoagulants. MATERIALS AND METHODS: We included patients ≥18 years between January 2018 and December 2022. The cohort was divided into two groups, initially based on anticoagulant management at discharge, subsequently, according to bleeding events during follow-up, which ended at the time of death, last clinical contact, time of the event, or within 6 months since last patient was included. RESULTS: A total of 120 patients were included. Those with HAS-BLED score > 3 were more likely discontinue DOAC (P = 0.0009). Significantly more patients with previous GIB had a rebleeding episode (29.6% vs. 6.4%, P = 0.002) and a 4-fold higher risk of rebleeding in the first 15.2 months (HR 4.070, CI [1.771-9.354], P = 0.001). A history of diabetes mellitus, malignancy, CHA2DS2-VASc, and HAS-BLED score > 3 was associated with an increased rebleeding risk, but without statistical significance. CONCLUSION: Previous episode of GIB is the most significant risk factor for recurrent bleeding in DOAC-treated patients. It leads to a four times higher risk of bleeding, particularly in the first 15 months. Patients aged ≥75, with a history of diabetes mellitus, malignancy, and HAS-BLED score > 3 should be closely followed-up. Timing of anticoagulation resumption and specific DOAC agent did not influence the rate of rebleeding, supporting their proven benefit.