Abstract 13828: Real World Bleeding Risks in Non-valvular Atrial Fibrillation Patients With Heart Failure: Contemporary EHR Results Among Prescribed Apixaban, Dabigatran, Rivaroxaban and Warfarin

Introduction: With the availability of four non-Vitamin K antagonist oral anticoagulants (NOACs) for patients with non-valvular atrial fibrillation (NVAF), insights into how these medications perform in ‘real-world’ settings, particularly among heart failure (HF) patients with increased risk for stroke and bleeding events, would be helpful. This study compared bleeding risk among newly anticoagulated NVAF patients, overall and in an HF subgroup. Methods: Using Humedica de-identified EHR data, newly anticoagulated adult NVAF patients with treatment initiation (index) between 2013 and 2015 were identified. Based on index prescription, patients were assigned to apixaban, rivaroxaban, dabigatran, and warfarin cohorts and followed until the earliest of drug switch, bleeding event (major or clinically-relevant non-major), last encounter, or 180 days post-index. For the HF subgroup, patients had to have an HF diagnosis in the 12 months pre-index. EHR prescription data did not allow accounting for discontinuation. Cox proportional hazard analysis was used to estimate bleeding hazard ratios (HR) for each NOAC compared to warfarin, adjusting for age, sex, region, stroke risk, and baseline comorbidities. Results: A total of 118,300 NVAF patients were selected, of which 17,127 were HF patients. Mean age ranged from 71.9 for dabigatran to 74.6 for warfarin. Mean CHADS 2 score was similar across cohorts. Overall, initiation of apixaban (HR: 0.78; 95% CI: 0.73-0.82) and dabigatran (HR: 0.71; 95% CI: 0.66-0.77) was associated with statistically significant lower risk of bleeding compared to warfarin, while rivaroxaban (HR: 1.09; 95% CI: 1.04-1.13) was associated with higher risk of bleeding compared to warfarin. Results were similar in the HF subgroup with the risk of bleeding lower for apixaban (HR: 0.78; 95% CI: 0.67-0.90) and dabigatran (HR: 0.67; 95% CI: 0.54-0.82) and higher for rivaroxaban (HR: 1.19; 95% CI: 1.07-1.32). Conclusion: Among NVAF patients, both overall and those with HF, initiating treatment with apixaban and dabigatran was associated with significantly lower risk of any bleeding event compared to initiating warfarin, whereas initiating with rivaroxaban was associated with higher bleeding risk compared with warfarin.