PERSISTENCE OF ORAL ANTICOAGULANT USE AMONG CANADIAN ATRIAL FIBRILLATION PATIENTS: ONE-YEAR RESULTS FROM THE PROSPECTIVE SPRINT-AF REGISTRY

Abstract

Despite the proven benefit of oral anticoagulation (OAC) in reducing the risk of stroke among atrial fibrillation (AF) patients, there is lingering concern that persistence with these life-saving medications is suboptimal in routine clinical practice. From a prospective registry, we sought to examine: (i) the one-year persistence rates to warfarin and non-vitamin K oral anticoagulants (NOAC) and (ii) factors associated with OAC persistence. We enrolled 2,498 AF patients from 132 community-based sites in Canada (November 2013 to March 2016) in a prospective registry. Use of OAC at one year was ascertained by telephone follow-up. Multivariable logistic regression was performed to examine for factors associated with continued OAC use (vs. non-use) at one-year. Among patients who remained treated with OAC at one-year, a multivariable logistic regression was performed to examine for factors associated with use of the same OAC agent vs. being switched to another OAC agent. Among the 2,010 patients treated with OAC at baseline, one-year data on OAC use were obtained for 1,387 (69%) patients. In this cohort, 124 patients (8.9%) were no longer treated with OAC at one-year. The rates of OAC discontinuation of warfarin- and NOAC-treated patients were 5.1% and 10.5%, respectively (P<0.05). The distribution of OAC use at one-year, stratified by baseline OAC use, was shown in the figure. On multivariable regression analysis, use of NOAC and concomitant use of antiplatelet agents were associated with increased likelihood of OAC discontinuation at one-year (odds ratio (OR) 2.22, 95% CI 1.35-3.70 and OR 2.50, 95% CI 1.54-4.00). Among patients who remained treated with OAC at one-year, the odds of staying on the same OAC agent were higher for patients treated with factor Xa inhibitors when compared to those treated with warfarin at baseline (OR 1.64, 95% CI 1.05-2.54 (rivaroxaban vs. warfarin) and OR 5.26, 95% CI 2.22-12.5 (apixaban vs. warfarin)). Similar results were obtained when our analysis was restricted to patients treated with the same OAC for >90 days prior to their baseline visit. In this contemporary Canadian AF registry, the one-year persistence rates of OAC were high irrespective of the type of agent prescribed. Among patients who experienced a change in their OAC use, warfarin-treated patients were more likely to switch to NOACs instead of stopping OAC. NOAC-treated patients were more likely to discontinue use of OAC altogether rather than being switched to warfarin.