Major bleeding modifiers in anticoagulated patients with atrial fibrillation and heart failure: a systematic review.

Abstract

Patients with atrial fibrillation (AF) and heart failure (HF) often require oral anticoagulation (OAC) but have renal dysfunction, anaemia, and polypharmacy. Evidence on bleeding-risk modifiers in patients with both AF and HF has not been synthesised specifically.We conducted a systematic review of studies reporting major bleeding modifiers in anticoagulated patients with atrial fibrillation and heart failure. Modifier domains, outcome definitions, study design features, and antithrombotic-treatment contexts were extracted. Structured narrative synthesis was the primary approach; only contrasts meeting the prespecified pooling criteria, with sufficiently comparable adjusted estimates, were evaluated in exploratory random-effects meta-analyses.Twenty-three studies provided 93 effect rows. Prominent clinically monitorable domains reported across the included studies were severe renal dysfunction (Cockcroft-Gault-based renal function <30 vs ≥80 mL/min: hazard ratio [HR] 5.77, 95% confidence interval [CI] 3.16-10.52), prior bleeding (HR 2.48, 1.44-4.28), inappropriate or unrevised anticoagulant dosing, anaemia, and concomitant antiplatelet therapy, with more limited evidence for nonsteroidal anti-inflammatory drug exposure. In a separate exploratory random-effects meta-analysis of the antithrombotic-treatment context, any direct oral anticoagulant (DOAC) versus warfarin or other non-DOAC oral anticoagulation was associated with lower major bleeding (pooled HR 0.69, 95% CI 0.55-0.87; k=6; I2=75%; τ2=0.05552). This was contextual safety evidence rather than a clinical modifier.This systematic review identifies clinically monitorable major bleeding modifier domains in anticoagulated patients with atrial fibrillation and heart failure. Quantitative synthesis was limited to a small number of comparable comparisons and should be interpreted as exploratory.