Abstract TP517: Characteristics and Outcomes in Patients With Atrial Fibrillation (af) Receiving Direct Oral Anticoagulants (doacs) in off-Label Dose

Abstract

Objective: We evaluated adherence to FDA-approved dosing criteria for patients with AF who initiated dabigatran and rivaroxaban, and the impact of off-label dosing on ischemic stroke and bleeding events. Methods: We identified 17,622 and 37,452 Humana Medicare beneficiaries age >=65 years with AF who initiated dabigatran or rivaroxaban, respectively, during 2010-2016. Patients were excluded if they had pulmonary embolism, deep vein thrombosis, or underwent hip surgery within the previous 6 weeks, or were not enrolled in Humana Medicare for at least 1 year prior to initiating the index drug. Stroke, major bleeding, and medication termination in patients who were either ‘potentially over-dosed’ (i.e., receive the standard dose but were eligible for low dose), or ‘potentially under-dosed’ (i.e., received the low dose but were not eligible for low dose) were identified. Multivariable methods controlled for differences in characteristics of patients prescribed low and standard dose within eligibility category. Results: Overall, 1734 (9%) and 10,045 (27%) of patients who received dabigatran and rivaroxaban met criteria for low dose, respectively. Of those, 1,161 (67%) and 4,868 (48%) received the standard dose (“potentially over-dosed”). In contrast, 2154 (13%) and 6,480 (23%) of patients eligible for standard dose dabigatran and rivaroxaban received a lower dose, respectively (“potentially under-dosed”). Increased age, female sex, black race, low weight, bleeding history, and heart failure were associated with receipt of lower than recommended dose for patients receiving dabigatran or rivaroxaban. There was modest evidence of increased risk of intracranial hemorrhage in patients overdosed with dabigatran (p=0.07). Among rivaroxaban users, medication termination was significantly higher for rivaroxaban users who were potentially overdosed (Hazard Ratio [HR] = 1.05; p=0.04); and significantly lower among rivaroxaban users who were potentially underdosed (HR = 0.96; p=0.03). No other significant outcomes differences were noted. Conclusions: Significant proportions of patients receive off-label dosing of dabigatran and rivaroxaban. Nevertheless, we found little evidence of significant impact on clinical outcomes.