Real-world comparison of all-cause hospitalizations, hospitalizations due to stroke and major bleeding, and costs for non-valvular atrial fibrillation patients prescribed oral anticoagulants in a US health plan

Aims: To compare the risk of all-cause hospitalization and hospitalizations due to stroke/systemic embolism (SE) and major bleeding, as well as associated healthcare costs for non-valvular atrial fibrillation (NVAF) patients initiating apixaban, dabigatran, rivaroxaban, or warfarin. Materials and methods: NVAF patients initiating apixaban, dabigatran, rivaroxaban, or warfarin were selected from the OptumInsight Research Database from January 1, 2013-September 30, 2015. Propensity score matching (PSM) was performed between apixaban and each oral anticoagulant. Cox models were used to estimate the risk of stroke/SE and major bleeding. Generalized linear and 2-part models were used to compare healthcare costs. Results: Of the 47,634 eligible patients, 8,328 warfarin-apixaban pairs, 3,557 dabigatran-apixaban pairs, and 8,440 rivaroxaban-apixaban pairs were matched. Compared to apixaban, warfarin patients were associated with a significantly higher risk of all-cause (hazard ratio [HR] = 1.30; 95% confidence interval [CI] = 1.21-1.40) as well as stroke/SE-related (HR = 1.60; 95% CI = 1.23-2.07) and major bleeding-related (HR = 1.95; 95% CI = 1.60-2.39) hospitalization; rivaroxaban patients were associated with a higher risk of all-cause (HR = 1.15; 95% CI = 1.07-1.24) and major bleeding-related hospitalization (HR = 1.71; 95% CI = 1.39-2.10); and dabigatran patients were associated with a higher risk of major bleeding hospitalization (HR = 1.46, 95% CI = 1.02-2.10). Warfarin patients had significantly higher major bleeding-related and total all-cause healthcare costs compared to apixaban patients. Rivaroxaban patients had significantly higher major bleeding-related costs compared to apixaban patients. No significant results were found for the remaining comparisons. Limitations: No causal relationships can be concluded, and unobserved confounders may exist in this retrospective database analysis. Conclusions: This study demonstrated a significantly higher risk of hospitalization (all-cause, stroke/SE, and major bleeding) associated with warfarin, a significantly higher risk of major bleeding hospitalization associated with dabigatran or rivaroxaban, and a significantly higher risk of all-cause hospitalization associated with rivaroxaban compared to apixaban. Lower major bleeding-related costs were observed for apixaban patients compared to warfarin and rivaroxaban patients.