Abstract 12992: Regional Variation in Ischemic Stroke Rates and Oral Anticoagulant Use Among the Non-Valvular Atrial Fibrillation Medicare Population

Abstract

Introduction: The prevalence and economic burden of non-valvular atrial fibrillation (NVAF) among Medicare beneficiaries are high. Examining regional variation in healthcare resource utilization and treatment may show where there is potential for reducing costs and improving quality. Objectives: To identify regional variation in the prevalence of NVAF, OAC use, and ischemic stroke rates and model the impact of improved oral anticoagulation (OAC) scenarios on ischemic stroke rates and associated costs in the Medicare population.

Method: This was a retrospective analysis of 5% sample of Medicare Part A and B (2007-9) and 100% Part A claims (2009) of NVAF patients. The 5% data was used to identify NVAF patients with 1 inpatient, ER claim, or 2 outpatient claims on different dates of service. The incremental cost of ischemic stroke was calculated using the annualized costs of stroke patients over 3 years from the stroke date and subtracting the patient costs observed in the year prior to the stroke. The 100% sample was used to examine NVAF prevalence, OAC use, and stroke rate at the state and county levels. A model evaluating the outcomes of ischemic stroke rates and associated costs was developed to predict the scenarios of treating OAC non-users or treating warfarin users with more efficacious OAC therapy (assuming a 24% RRR).

Results: The nationwide prevalence of NVAF in the Medicare population is 8.4%, ranging from 5.4% in Pike, KY to 11% in Waupaca County, WI. The nationwide OAC use was 45.6%, ranging from 21.8% in Kauai, HI to 60.6% in Jones, SD. The nationwide ischemic stroke rate was 2.9%, ranging from 1.5% in Cochise, AZ to 4.5% in Roane, TN. The national ischemic stroke rate of NVAF warfarin takers is 31% lower than that of the non-warfarin takers (p value <0.0001). The model predicted a $2.19 per member per month (PMPM) savings attributed to ischemic stroke if 80% of OAC non-users were placed on an OAC (warfarin or more efficacious OAC). A predicted $0.29 PMPM could be saved if 50% of current warfarin users were treated with a more efficacious OAC.

Conclusions: Significant regional variation in the prevalence, OAC treatment, and stroke rate for NVAF patients was found. Improvements in therapy and decreasing regional variation may lead to improved outcomes and lower costs.