Severity of Gastrointestinal Bleeding in Anticoagulant Therapy with Vitamin K Antagonist or Direct Oral Anticoagulant Therapy.

Abstract

It remains undetermined whether gastrointestinal bleeding (GIB) associated with direct oral anticoagulant (DOAC) or vitamin K antagonists (VKA) is more severe. In Danish nationwide registries, we identified patients with atrial fibrillation (AF) treated with anticoagulation with gastrointestinal bleeding (GIB). Logistic regression models were used to compare GIB severity between DOAC and VKA users, defined by 1. red blood cell transfusion or in-hospital death and 2. endoscopies. From 2012-2018, 6,784 anticoagulated AF patients with GIB were identified; 3,724 patients VKA users and 3,060 DOAC (apixaban, dabigatran or rivaroxaban) users. The proportion of upper GIB was 49% in VKA users and 40.1% in DOAC users. Blood transfusions were administered to 58.2% and 43.8-48.9% of patients with VKA and DOAC, respectively. Odds ratios (OR, adjusted for age, sex, and comorbidities) for severe GIB were significantly reduced for all DOACs compared to VKA (apixaban 0.71 (95% confidence interval (CI) 0.59-0.85), dabigatran 0.64 (95% CI 0.55-0.75) and rivaroxaban 0.69 (95% CI 0.59-0.81)). Odds for endoscopy was significantly reduced in apixaban (OR 0.77 (95% CI 0.65-0.90) and dabigatran (OR 0.81 (95% CI 0.71-0.93) compared to VKA, and borderline reduced for rivaroxaban (OR 0.88 (95% CI 0.77-1.02)). In patients with upper GIB, compared to VKA the use of gastroscopy was significantly reduced for apixaban (0.68 (95% CI 0.53-0.87)) but not for dabigatran or rivaroxaban. VKA use was associated with more severe GIB and more frequent endoscopies compared to DOAC. Future studies should identify subgroups at particular risk of severe GIB.