Net clinical benefit of edoxaban versus no treatment in a ‘real world’ atrial fibrillation population: A modelling analysis based on a nationwide cohort study

Abstract

BackgroundIn non-valvular atrial fibrillation (AF), oral anticoagulation reduces the risk of thromboembolism such as stroke and systemic embolism (SSE), but increases the risk of major bleeding such as intracranial haemorrhage (ICH). The risk-benefit balance between SSE versus ICH can be expressed as the net clinical benefit (NCB); however, the risk of SSE and ICH varies according to clinical factors that can be assessed using CHADS2, CHA2DS2-VASc (both quantifying risk of stroke) and HAS-BLED (quantifying risk of major bleeding) scores, respectively. MethodsUsing established modelling based on event rates for thromboembolism and haemorrhage in the Danish nationwide cohort study, we tested the hypothesis that edoxaban has a superior NCB compared with warfarin. ResultsIn our overall model, compared to no treatment, warfarin had a NCB of 0.26 (95% CI 0.24,0.28) events prevented per 100 patient years, edoxaban 60mg daily a NCB of 0.71 [0.69,0.76], and edoxaban 30mg daily a NCB of 0.71 [0.0.68,0.73]. When compared to no treatment, both doses of edoxaban have superior NCB values than those of warfarin at all CHADS2 and CHA2DS2-VASc scores. At CHADS2 ≥2 and CHA2DS2-VASc ≥2, edoxaban 60mg dose had a better NCB than the 30mg dose or warfarin, when compared to no treatment. With HAS-BLED score ≥3, both doses of edoxaban had a positive NCB compared to warfarin, at CHADS2 or CHA2DS2-VASc ≥2. ConclusionOur modelling study suggests that both 30mg and 60mg doses of edoxaban have a favourable NCB compared to warfarin, and the degree of benefit differs according to CHADS2, CHA2DS2-VASc and HAS-BLED scores. At CHA2DS2-VASc score ≥2, both edoxaban doses were superior to warfarin, but compared to no treatment, the 60mg dose had a better NCB than the 30mg dose or warfarin.