Abstract 16520: Treatment Persistence With New Oral Anticoagulants for Non-Valvular Atrial Fibrillation: The Quebec Observational Study
Introduction: Continuous use of oral anticoagulant therapy is essential to reduce the risk of stroke in patients with non-valvular atrial fibrillation (NVAF). To date, no study has compared real world treatment persistence patterns between users of new oral anticoagulants (NOACs) and warfarin, before and after the introduction of NOACs in 2011
Objective: To compare treatment persistence between NOACs and warfarin users with NVAF.
Methods: A retrospective cohort of patients 40 years of age or older with a first inpatient or outpatient diagnosis for NVAF during 2006-14 was identified using the healthcare databases of the Régie de I′assurance-maladie du Québec (RAMQ). Patients were followed until the emigration date, end of prescription drug coverage, death or December 31, 2014. Treatment persistence was defined from dispensed prescriptions that follow each other with a gap of no more than 20% of the duration of the previous prescription. Kaplan-Meier curves were used to describe the cumulative probability of persistence of the initiating treatment. Cox proportional hazards models were used to estimate hazard ratios (HRs) of discontinuation of NOACs relative to VKAs, after adjustment for baseline confounders.
Results: The cohort included 154,101 patients with NVAF, of which 74,000 were diagnosed before January 2011 (54,000 on VKA) and 70,119 after (22,458 on VKA and 16,637 on NOAC). Patients initially treated with VKAs were slightly older and had more comorbidity, including prior bleeding events, cardiovascular diseases as well as other chronic and acute conditions. At twelve months following treatment initiation, persistence was 36% for patients on NOACs compared with 14% for patients on VKAs before and after January 2011. A similar pattern was seen for rivaroxaban versus VKA. After adjustment, the probability of NOAC treatment discontinuation was half that of VKA (HR 0.50; 95% CI: 0.49-0.51). Sensitivity analyses using a longer grace period of 60 days to define persistence confirmed a higher one-year persistence of 80% for NOACs versus 60% for VKAs (HR of discontinuation 0.50; 95% CI: 0.48-0.52).
Conclusions: This real world study shows that twice as many patients initiated on NOACs, including rivaroxaban, remain on therapy than those initiating VKAs.
Author Disclosures: S. Suissa: Research Grant; Modest; Bayer, Boehringer-Ingelheim, Bristol- Myers-Squibb. J. Coulombe: None. T. Evers: Employment; Modest; Bayer Pharma AG. C. Renoux: None.
- © 2016 by American Heart Association, Inc.