ABSTRACT
Background: The patterns of direct oral anticoagulant (DOAC) selection and switching to a different oral anticoagulant (OAC) in patients with atrial fibrillation (AF) are unknown. Objectives: To describe temporal patterns in first DOAC prescriptions, estimate the incidence, and identify predictors of switching to a different OAC within 1 year in OAC-naive AF patients. Methods: In this retrospective cohort study, using a near-nationwide prescription registry (IQVIA, the Netherlands), we determined the number of patients per month initiated on each DOAC and identified predictors of switching within 1 year with robust Poisson regression. Results: We included 94,874 patients. From November 2015 to November 2019, the monthly use of apixaban (n = 366 to n = 1066, +191%), rivaroxaban (n = 379 to n = 868, +129%), and edoxaban (n = 2 to n = 305, +15,150%) increased, whereas that of dabigatran decreased (n = 317 to n = 179, -44%). In the 66,090 patients with ≥1 year of available calendar time, 7% switched to a different OAC within 1 year. Strong predictors of switching to a different DOAC were using dabigatran (adjusted risk ratio [aRR], 3.33; 95% CI, 3.02-3.66) or edoxaban (aRR, 1.56; 95% CI, 1.34-1.82) rather than apixaban and using a standard DOAC dose (aRR, 2.54; 95% CI, 2.23-2.88). Strong predictors of switching to a vitamin K antagonist were using rivaroxaban (aRR, 1.36; 95% CI, 1.19-1.54 vs apixaban) and using a standard DOAC dose (aRR, 1.49; 95% CI, 1.26-1.77). Conclusion: In the Netherlands, factor Xa inhibitors are increasingly being selected for OAC-naive AF patients. Seven percent of patients switch to a different OAC within 1 year, and the initial DOAC type and dose are strong predictors of switching.
ABSTRACT
Background For most patients with newly diagnosed atrial fibrillation (AF), direct oral anticoagulants (DOACs) are preferred over vitamin K antagonists. However, there is concern that the lack of monitoring may impair therapy adherence and therefore the anticoagulant effect. Objective To assess 1-year DOAC nonadherence in patients with AF and a treatment indication of at least 1 year in the Dutch health care setting, and to identify predictors of nonadherence. Methods We performed a near-nationwide historical cohort study in patients with a novel DOAC indication for AF. Data were obtained from a pharmacy database, covering 65% of all outpatient prescriptions dispensed in the Netherlands. The 1-year nonadherence was assessed by the proportion of days covered; the threshold was set at <80%. Robust Poisson regression analyses were performed to identify predictors of nonadherence. Results A total of 46,211 patients were included and the 1-year nonadherence was 6.5%. We identified male sex (risk ratio [RR] 1.23, 95% confidence interval [CI]: 1.15-1.33), younger age (age ≥60 to <70 years: RR: 1.15, 95% CI: 1.00-1.33, age <60 years: RR: 2.22, 95% CI: 1.92-2.57; reference age ≥85 years), a reduced DOAC dose (RR: 1.10, 95% CI: 1.00-1.22), a twice-daily dosing regimen (RR: 1.21, 95% CI: 1.12-1.30), and treatment with apixaban (RR: 1.16, 95% CI: 1.06-1.26, reference rivaroxaban) or dabigatran (RR: 1.25, 95% CI: 1.14-1.37) as independent predictors of 1-year nonadherence. Conclusion One-year nonadherence to DOACs was low yet relevant in patients with AF newly prescribed a DOAC. Understanding the predictors for nonadherence may help identify patients at risk.