Heart failure and major haemorrhage in people with atrial fibrillation.

backgroundHeart failure (HF) is not included in atrial fibrillation (AF) bleeding risk prediction scores, reflecting uncertainty regarding its importance as a risk factor for major haemorrhage. We aimed to report the relative risk of first major haemorrhage in people with HF and AF compared with people with AF without HF ('AF only').

methodsEnglish primary care cohort study of 2 178 162 people aged ≥45 years in the Clinical Practice Research Datalink from January 2000 to December 2018, linked to secondary care and mortality databases. We used traditional survival analysis and competing risks methods, accounting for all-cause mortality and anticoagulation.

resultsOver 7.56 years median follow-up, 60 270 people were diagnosed with HF and AF of whom 4996 (8.3%) had a major haemorrhage and 36 170 died (60.0%), compared with 8256 (6.4%) and 34 375 (27.2%), respectively, among 126 251 people with AF only. Less than half those with AF were prescribed an anticoagulant (45.6% from 2014 onwards), although 75.7% were prescribed an antiplatelet or anticoagulant. In a fully adjusted Cox model, the HR for major haemorrhage was higher among people with HF and AF (2.52, 95% CI 2.44 to 2.61) than AF only (1.87, 95% CI 1.82 to 1.92), even in a subgroup analysis of people prescribed anticoagulation. However, in a Fine and Gray competing risk model, the HR of major haemorrhage was similar for people with AF only (1.82, 95% CI 1.77 to 1.87) or HF and AF (1.71, 95% CI 1.66 to 1.78).

conclusionsPeople with HF and AF are at increased risk of major haemorrhage compared with those with AF only and current prediction scores may underestimate the risk of haemorrhage in HF and AF. However, people with HF and AF are more likely to die than have a major haemorrhage and therefore an individual's expected prognosis should be carefully considered when predicting future bleeding risk.