Use of Oral Anticoagulation in Eligible Patients Discharged With Heart Failure and Atrial Fibrillation.

Abstract

Stroke prophylaxis in patients with atrial fibrillation (AF) and heart failure (HF) in the era of direct oral anticoagulants is not well characterized. Using data from American Heart Association Get With The Guidelines-AFIB, we sought to evaluate oral anticoagulation (OAC) use at discharge among AF patients with concomitant HF. AF patients with a diagnosis of HF hospitalized from January 2013 to March 2017 were included. We compared patient characteristics and use of OAC at discharge among patients with reduced (redundant ejection fraction [EF], EF≤40%), borderline (40%<EF<50%), and preserved (EF≥50%) EF using multivariable mixed logistic regression models. Among 10 883 patients with AF and HF, 1790 (16.4%) had a reported contraindication to anticoagulation and were excluded from further analysis. Among 9093 patients eligible for OAC, 3499 (38.5%) had HF with reduced EF, 1062 (11.7%) had HF with borderline EF, and 4532 (49.8%) had HF with preserved EF. The median CHA2DS2-VASc score was 5 (Q1, Q3; 3, 6) among all patients and higher among those with HF with preserved EF than HF with reduced EF (5 [4, 6] versus 4 [3, 5]; P<0.0001). The proportion of eligible patients discharged on OAC was 94.9%, with 43.6% discharged on warfarin and 50.7% discharged on direct oral anticoagulants. A higher proportion of patients with HF with reduced EF and HF with borderline EF were discharged on direct oral anticoagulants than with HF with preserved EF, but the difference was small (52.8%, 53.1% versus 48.5%, respectively; P=0.0002). EF group was not significantly associated with a patient's OAC use at discharge. In the context of American Heart Association Get With The Guidelines-AFIB, a quality improvement program, the rate of use of OAC at discharge in eligible AF patients with HF was almost 95%. To our knowledge, these rates represent some of the highest use of appropriate anticoagulation for patients in a national registry to date.