Introduction: Non-vitamin K antagonist Oral Anticoagulation (NOAC) agents are approved for prevention of ischemic stroke in atrial fibrillation (Afib). NOACs offer both advantages (ease of use, reduced rate of intracranial hemorrhage) and disadvantages (cost, absence of reversal agent, less clinical experience) relative to warfarin. We tested the hypothesis that the decision to prescribe a NOAC might be influenced by non-medical factors such as prescriber subspecialty or patient demographics.
Methods: We performed an IRB-approved search of the Research Patient Data Registry at our major tertiary referral hospital for all Afib patients prescribed warfarin, dabigatran, rivaroxaban or apixaban between January 2013 and June 2014. We encoded data related to prescriber clinic, patient demographics (including home ZIP code as a marker of family income), intracerebral hemorrhage (ICH) and medical comorbidities.
Results: Of 4,261 individuals with Afib prescribed oral anticoagulants during the specified interval, 3450 (81%) received warfarin, 442 (10.4%) rivaroxaban, 304 (7.1%) dabigatran, and 65 (1.5%) apixaban per their last prescription during the period. In univariate analyses patients prescribed NOACs were younger (71.5 ± 11.4 vs. 76.6 ± 10.5 years, p < 0.001), more likely male (64.4% vs. 35.6%, p < 0.005), had lower CHADS2 scores (2.17 ± 1.43 vs. 2.68 ± 1.40, p<0.001) and resided in areas with higher median income ($86,384 ± 33,410 vs. $80,470 ± 30,390, p<0.001). Race and prior ICH did not differ. Among 2,013 prescriptions from a defined subset of cardiology, primary care and neurology clinics, NOACs were more commonly used by cardiologists (37.1%) than neurologists (18.9%) or primary care providers (17.1%, p<0.001). Multivariable analysis of this subset found that physician specialty, residential area income, age and CHADS2 score but not gender were independently associated with receiving a NOAC.
Conclusions: Younger, more affluent individuals treated by a cardiologist were likeliest to receive NOACs for prevention of Afib-related stroke. The data suggest that factors other than medical characteristics might influence the decision to use NOACs.