Real world data show that oral anticoagulant (OAC) is prescribed in approximately 40% of patients with atrial fibrillation (AF) and a low thromboembolic (TE) risk–CHA2DS2-VASc score 0 [male] or 1 [female]. Guidelines recommend against OAC in such patients because the TE risk is outweighed by the bleeding risk. Determinants of the decision to prescribe OAC for patients with a low TE risk are poorly understood.

To identify patient characteristics and reasons for clinicians to prescribe OAC in AF despite a low TE risk.

Patient characteristics associated with OAC in the GARFIELD-AF registry were assessed using logistic regression analysis [OR (95%)]. One-year all-cause mortality, ischemic stroke or systemic embolism, and major bleeding were reported. Subsequently, a sample of physicians was questioned through a web-based survey to identify factors that may influence their decision-making. Total Unduplicated Reach and Frequency analysis was used.

Of the 52014 patients included in GARFIELD-AF, 2123 had a low risk and 950 (45%) of them had OAC. Permanent [OR=2.3(1.5–3.6)] or persistent [OR=3.1(2.2–4.4)] vs incident AF and increasing age>65 years [OR=1.3(1.2–1.5) per 10-y increment] were associated with OAC use, antiplatelet therapy [OR=0.08(0.07–0.11)] and female gender [OR=0.7(0.6–0.9)] with no OAC use. 1-y all-cause mortality (14 vs 20), ischemic stroke or systemic embolism (6 vs 5) and major bleeding (4 vs 3) were low with and without OAC. In the physician survey (n=229), an enlarged left atrium was the most frequently cited reason to use OAC (reach: 59.8%). Other factors were cardioversion/ablation procedures, rheumatic heart disease and “fear” of stroke. Approximately 83.8% of physicians would tend to prescribe OAC if any of these factors is present (Table 1).

Several factors not including in the CHA2DS2-VASc score influence physicians’ decision making toward OAC use in low TE risk patients, against current guidelines.