OBJECTIVES: To compare adherence, persistence, discontinuation and switching rates among Medicare beneficiaries with non-valvular atrial fibrillation (NVAF) or venous thromboembolism (VTE) using direct oral anticoagulants (DOACs).

METHODS: Retrospective observational design using Medicare Part-D claims from 2015-2018. Inclusion-exclusion criteria applied to identify NVAF and VTE patients using dabigatran, rivaroxaban, apixaban, edoxaban and warfarin. Adherence, persistence, time to non-persistence and time to discontinuation assessed in those who didn't switch the index drug in the follow-up period (365 days). Switching rates assessed in those who switched the index drug at least once in the follow-up period. Descriptive statistics conducted for all outcomes, and comparisons made using t-tests, chi-square, and ANOVA. Logistic regression conducted to compare the odds of being adherent to each DOAC.

RESULTS: Of all the DOACs, highest adherence was reported among apixaban users (PDC = 76.88) and highest non-persistence and discontinuation rates were reported among warfarin users. NVAF cohort using dabigatran was significantly less adherent vs. rivaroxaban (OR:0.90) and apixaban (OR:0.84); apixaban users were significantly more adherent vs. rivaroxaban (OR:0.91) and edoxaban (OR: 1.42); rivaroxaban users were significantly more adherent vs. edoxaban (OR: 1.37). VTE cohort using apixaban was significantly more adherent vs. dabigatran (OR: 0.86), rivaroxaban (OR: 0.91) and edoxaban (OR: 0.38). Majority of switches were reported away from dabigatran and towards apixaban. For both cohorts, 24.35% of dabigatran, 13.94% of rivaroxaban, 29.43% of edoxaban and 11.19% of warfarin patients switched to apixaban. Despite the better utilization outcomes reported for apixaban users, Medicare plans covered rivaroxaban more favorably with the lowest mean amount paid by the patient (NVAF: $76; VTE: $59) and the highest mean amount covered by the plans (NVAF: $359; VTE: $326).

CONCLUSIONS: Strategies to incentivize Medicare plans to consider adherence, persistence, discontinuation and switching as metrics of utilization outcomes are needed to contain the high Medicare spending on DOACs.