Aims: To examine the long-term persistence and safety of the non-vitamin-K-antagonist oral anticoagulants (NOACs) dabigatran (D), rivaroxaban (R) and apixaban (A) in patients with non-valvular atrial fibrillation (AF) treated in the framework of a well structured, nurse-based AF unit for initiation and follow-up of NOAC.
Methods: Retrospective clinical data were collected for 766 consequent patients from a single cardiology outpatient clinic incorporating the AF unit.
Results: The follow-up time, median (q1-q3), was 367 days (183-493) for D patients (n = 233), 432 days (255-546) for R patients (n = 282) and 348 days (267-419) for A patients (n = 251). No significant differences were found between the three groups with regard to age, sex, renal function, or CHA2DS2-VASc score. For all bleeding events the incidence rates per 100 patient-years of follow-up (95% confidence interval [CI], p-value) were reported more often for treatment with R (17.2, 12.7-22.8) than for D (7.0, 4.0-11.3, p = 0.001) and A (8.7, 5.2-13.6, p = 0.013). The differences remained significant after adjustment for clinically relevant variables. Discontinuation rates (n = 167) were lower for A (11.5, 7.5-16.8) than for D (30, 23.4-37.9, p < 0.001) and R (23.9, 18.6-30.1, p = 0.001), and were mainly attributed to drug-specific side effects and bleedings. The majority of discontinued patients (n = 142, 85%) proceeded with other types of oral anticoagulants.
Limitation: The main limitation of the study is the small patient population with a short follow-up time.
Conclusion: In a retrospective study at a single AF clinic, NOACs showed significantly different bleeding rates and varied discontinuation rates when compared to each other, related mainly to agent-specific side effects and bleedings. The majority of patients that discontinued proceeded with other types of oral anticoagulant.
Keywords: Apixaban; atrial fibrillation; dabigatran; persistence; rivaroxaban; safety.