Dabigatran Persistence and Outcomes Following Discontinuation in Atrial Fibrillation Patients from the GLORIA-AF Registry.

Prospective studies evaluating persistence to nonvitamin K antagonist oral anticoagulants in patients with atrial fibrillation are needed to improve our understanding of drug discontinuation. The study objective was to evaluate if and when patients with newly diagnosed atrial fibrillation stop dabigatran treatment and to report outcomes following discontinuation. Patients prescribed dabigatran in diverse clinical practice settings were consecutively enrolled and followed for 2 years. Dabigatran persistence over time, reasons for discontinuation, and outcomes post discontinuation were assessed. Of 4,859 patients, aged 70.2 ± 10.4 years, 55.7% were male. Overall 2-year dabigatran persistence was 70.9% (95% confidence interval [CI] 69.6 to 72.2). Persistence probability was lower in the first 6-month period (83.7% [82.7 to 84.8]) than in subsequent periods for patients on dabigatran at the start of each period (6 to 12 months, 92.5% [91.6 to 93.3]; 12 to 18 months, 95.1% [94.3 to 95.8]; 18 to 24 months, 96.3% [95.6 to 96.9]). Of 1,305 patients (26.9%) who discontinued dabigatran, adverse events were reported as the reason for discontinuation in 457 (35.0%). Standardized stroke incidence rate post discontinuation (per 100 patient-years) in patients discontinuing without switching to another oral anticoagulant was 1.76 (95% CI 0.89 to 2.76) and 1.02 (95% CI 0.43 to 1.76) in those who switched, consistent with the expected benefit of remaining on treatment. Patients persistent with treatment at 1 year had >90% probability of remaining persistent at 2 years suggesting clinical interventions to improve persistence should be focused on the early period following treatment initiation.