OBJECTIVES: Prescribing oral anticoagulation is a guideline recommended approach to control the risk of atrial fibrillation (AF)-associated stroke. A study investigating commercially insured patients with AF found that cardiologists were more likely to initiate oral anticoagulation than other physicians. In this study, we aimed to compare the rates of oral anticoagulant prescriptions for patients with risk-stratified AF between cardiologists and other physicians at the national level.

METHODS: We queried the National Ambulatory Medical Care survey database from 2013 to 2015. All adults (age ≥18 years old) diagnosed with AF were included. Self-reported physician specialty was categorized as cardiology versus other specialties. Oral anticoagulants included warfarin or direct acting oral anticoagulants. CHADS score was calculated for each patient to stratify estimated risk of AF-associated stroke. Logistic regression model was used to determine the association of physician specialty with anticoagulation use, controlling for CHADS_­ score, age, sex, race/ethnicity and health insurance status.

RESULTS: The number of outpatient visits for AF increased from 13.2 million (95% Confidence Interval [CI] 9.1-17.3) in 2012 to 17 million (95% CI 12.6-21.5) in 2015. Overall, 78.8% of patients in moderate risk group and 74.3% of patients in high risk group did not receive any anticoagulants. Among high risk patients, cardiologists were less likely to prescribe anticoagulants compared to other physicians (20.1% vs. 29.9%; adjusted odds ratio [OR] 0.58, 95% CI 0.34-0.97). Anticoagulation prescription rates were not significantly different between cardiologists and other physicians for low risk (26.2% vs. 10.2%; OR 3.56, 95% CI 0.85-14.92) and moderate risk (20.4% vs. 21.9%; OR 1.10, 95% CI 0.48-2.49) patients.

CONCLUSIONS: Prescription patterns for oral anticoagulation, particularly for high AF-associated stroke risk, continue to vary between physician specialties. Identifying reasons for variability of prescription patterns may help in improving quality of adherence to guideline recommendations.