Abstract P205: Anticoagulation Volume Does Not Predict Facility Performance of Time in Therapeutic INR Range of Patients With Atrial Fibrillation: A National VA Study

Abstract

Introduction: Effectiveness of warfarin for stroke prevention in atrial fibrillation (AF) is proportional to the time in INR therapeutic range of 2.0-3.0. Time in therapeutic range (TTR) ≥ 58% is a validated threshold at which warfarin has benefit over aspirin, while TTR < 40% may cause net harm due to hemorrhage. Variation of TTR across facilities may be due to systematic or facility-based factors, including patient volume. We evaluated the relationship of anticoagulation facility patient volume and TTR. Methods: We performed a national cohort study of Veterans Health Administration data using centralized medical and administrative records. We identified 48,192 patients with at least two outpatient ICD-9 diagnoses of AF who were prescribed warfarin from 94 VA anticoagulation facilities in FY 2003-04. The primary outcome was facility TTR, which was mean TTR calculated for FY04 from all patient-level TTRs of that facility. Patient-level TTR was calculated using the modified Rosendaal method, using linear imputation of INR values up to a maximum of 56 days. Results: Excluding two outliers with TTR < 1%, the mean TTR across 92 facilities was 54.6% ± 6.1% (95% CI 43.7%-63.0%). Only 30 of 92 facilities (33%) had a mean TTR ≥ 58%, and 2 (2.2%) had a mean TTR < 40%. Facility volume, as measured by the number of identified patients per facility, was not related to TTR (R-squared 2.6%) although facility INR test frequency was related to TTR (R-squared 29%). After multivariate adjustment for facility volume and facility prevalence of diabetes, heart failure, hypertension, coronary disease, stroke/TIA, and mean patient age, INR test frequency was significantly associated with TTR ≥ 58% (p < 0.001), while facility volume remained insignificant (p=0.55). For every three additional INR tests per month, the adjusted odds of achieving TTR ≥ 58% increased 5.33 (p < 0.001). Conclusions: We found systematic variation of facility TTR in a national health care system. Facility volume did not predict quality of anticoagulation. However, the facility's mean INR frequency is strongly associated with its overall quality of anticoagulation of patients with AF. Standardization and improvement of warfarin management across facilities may improve quality and prevent adverse outcomes.