Abstract 18404: Warfarin Persistence in Atrial Fibrillation: The Michigan Anticoagulation Quality Improvement Initiative Experience

Abstract

Introduction: Warfarin reduces the risk of stroke in patients with atrial fibrillation (AF), but increases the risk of bleeding. In clinical trials, persistence with warfarin therapy ranges from 75-79% at one year, but few studies offer real-world estimates of warfarin persistence. There is limited data demonstrating the relationship between CHADS2 and HAS-BLED scores with adherence to warfarin. Methods: New patients treated with warfarin for stroke prevention in AF were followed at seven anticoagulation clinics involved in the Michigan Anticoagulation Quality Improvement Initiative (MAQI2) collaborative. Patients were excluded if they had additional indications for warfarin such as artificial valve, DVT, or PE. Risk factors for CHADS2 and HAS-BLED scores and reasons for warfarin discontinuation were abstracted from patient charts. Kaplan-Meier (KM) curves were constructed to characterize drug therapy discontinuation rates at one year. Cox proportional hazard regression and logistic regression analyses were performed to assess the effects on warfarin persistence by CHADS2 and HAS-BLED scores. Results: Of 1,973 eligible patients, 153 (7.8%) switched to non-warfarin therapy (e.g. dabigatran). In the remaining 1820 patients, one year persistence is 63% based on KM estimates. One year persistence by low, intermediate, or high CHADS2 score was 29%, 58%, and 73%, respectively. One year persistence for low, intermediate, or high HAS-BLED score was 41%, 66% and 76%, respectively. Warfarin persistence at one year was associated with an increasing CHADS2 (OR 1.244, CI 1.045-1.482) but not associated with HAS-BLED score (OR 1.016, CI 0.839-1.231) in multivariate analysis. Conclusions: A significant proportion of patients taking warfarin for SPAF discontinue therapy by one year. In the multivariate analysis, patients at higher risk of stroke, but not bleeding, are more likely to remain on warfarin therapy. Efforts to improve warfarin persistence need further investigation.