One of the main determinants of establishing the optimal treatment duration of patients with venous thromboembolism (VTE) is the risk of major bleeding during long-term anticoagulant therapy. The 6-variable VTE-BLEED score was recently developed to enable estimation of this bleeding risk. This study aimed at externally validating VTE-BLEED. This was a post-hoc study of the randomised, double-blind, double-dummy, Hokusai-VTE study that compared edoxaban versus warfarin for treatment of VTE. VTE-BLEED was calculated in all 8,240 study patients. The numbers of adjudicated major bleeding events during 'stable anticoagulation', i. e. occurring after day 30, in patients with low (total score <2 points) and high risk of bleeding (total score ≥2 points) were compared for the overall study population, patients randomised to edoxaban or warfarin, and for important patient subcategories. During 'stable' anticoagulation, major bleeding occurred in 1.02 % (40/3,903) and 0.82 % (32/3,899) of patients treated with warfarin and edoxaban, respectively. For the overall study population, the risks of bleeding in the low and high risk groups were 0.51 % and 2.03 %, respectively, for an odds ratio (OR) of 4.04 (95 % confidence interval [CI]: 2.51-6.48). ORs were 5.04 (95 %CI: 2.62-9.69) and 3.09 (95 %CI: 1.54-6.22) for warfarin and edoxaban, respectively. VTE-BLEED was consistently able to identify patients at a 2.5- to 11-fold higher bleeding risk across all the predefined subcategories, as well as for the treatment period between day 30 to day 180, and beyond day 180. In conclusion, patients identified as high risk by VTE-BLEED had a four-fold increased risk of bleeding during the chronic phase of treatment.
Keywords: Anticoagulation; bleeding; clinical prediction rule; deep vein thrombosis; edoxaban; pulmonary embolism; warfarin.