• BACKGROUND
    • Direct oral anticoagulants (DOACs) are recommended for patients with atrial fibrillation (AF) given their improved safety profile. Suboptimal adherence to DOACs remains a significant concern among individuals with AF. However, the extent of adherence to DOACs following a cardiovascular or bleeding event has not been fully evaluated.
  • OBJECTIVE
    • To evaluate the pattern of adherence trajectories of DOACs after a cardiovascular or bleeding event and to investigate the sociodemographic and clinical predictors associated with each adherence trajectory by using claims-based data.
  • METHODS
    • This retrospective study was conducted among patients with AF prescribed with DOACs (dabigatran/apixaban/rivaroxaban) between July 2016 and December 2017 and who were continuously enrolled in the Texas-based Medicare Advantage Plan. Patients who experienced a cardiovascular or bleeding event while using the DOACs were further included in the analysis. The sample was limited to patients who experienced a clinical event such as a cardiovascular or bleeding event while using the DOACs. The clinical events considered in this study were cardiovascular (stroke, congestive heart failure, myocardial infarction, systemic embolism) and bleeding events. To assess adherence patterns, each patient with a DOAC prescription was followed up for a year after experiencing a clinical event. The monthly adherence to DOACs after these events was evaluated using the proportion of days covered (PDC). A group-based trajectory model incorporated the monthly PDC to classify groups of patients based on their distinct patterns of adherence. Predictors associated with each trajectory were assessed using a multinomial logistic regression model, with the adherent trajectory serving as the reference group in the outcome variable.
  • RESULTS
    • Among the 694 patients with AF who experienced clinical events after the initiation of DOACs, 3 distinct adherence trajectories were identified: intermediate nonadherent (30.50%), adherent (37.7%), and low adherent (31.8%); the mean PDC was 0.47 for the intermediate nonadherent trajectory, 0.93 for the adherent trajectory, and 0.01 for low adherent trajectory. The low-income subsidy was significantly associated with lower adherence trajectories (odds ratio [OR] = 4.81; 95% CI = 3.07-7.51) and with intermediate nonadherent trajectories (OR = 1.57; 95% CI = 1.06-2.34). Also, nonsteroidal anti-inflammatory drug use was significantly associated with lower adherence trajectories (OR = 5.10; 95% CI = 1.95-13.36) and intermediate nonadherent trajectories (OR = 3.17; 95% CI = 1.26-7.93). Other predictors significantly associated with both nonadherent trajectories are type of DOACs (OR = 0.53; 95% CI = 0.35-0.79), presence of coronary artery disease (OR = 1.89; 95% CI = 1.01-3.55), and having 2 or more clinical events (OR = 1.65; 95% CI = 1.09-2.50).
  • CONCLUSIONS
    • Predictors identified provide valuable insights into the suboptimal adherence of DOACs among Medicare Advantage Plan enrollees with AF, which can guide the development of targeted interventions to enhance adherence in this high-risk patient population.