Abstract 096: Association of Body Mass Index With Bleeding in Patients Taking Warfarin for Atrial Fibrillation or Venous Thromboembolism

Abstract

Background: Most patients taking warfarin for atrial fibrillation (AF) and venous thromboembolism (VTE) have a target International Normalized Ratio (INR) between 2-3 to reduce risk of bleeding and thromboembolic events. Body Mass Index (BMI) is not included in traditional bleed risk scores, but may be an indicator of bleeding risk in warfarin patients. Methods: Using data from the multi-site Michigan Anticoagulation Quality Improvement Initiative (MAQI 2 ) Registry, we identified all AF/VTE patients , separated them into three cohorts: BMI < 20 (underweight), BMI 20-25 (normal weight) and BMI >25 (overweight). Bleeding events in these cohorts were identified and stratified into severity according to International Society of Thrombosis and Hemostasis criteria. Results: Of 6,054 patients, 4,766 (78.7%) had a BMI of > 25. These patients were generally younger, with higher prevalence of hypertension. The HAS-BLED scores were slightly lower in overweight AF patients (2.6 vs 2.8; p=0.04); otherwise no difference between groups. The overall minor, major, and life threatening bleeding rates were 22.8/27.7; 4.3/3.7; and 1.2/0.7 (per 100 patient years) in AF and VTE patients, respectively. A higher proportion of females were underweight for both indications, and AF patients were older. More underweight and normal weight AF and VTE patients had a bleeding history compared to overweight patients. Bleeding outcomes are listed in Table. Comparisons were made with Poisson regression analysis. Conclusion: In a large, unselected cohort of warfarin treated patient from a multi-site registry, minor bleeding was more common in underweight and normal weight AF patients; major and life-threatening bleeding was more common in underweight and normal weight VTE patients. Since the majority of patients were overweight, further studies are needed to determine if reasons for bleeding differ between patients based on BMI in order to guide quality improvement efforts.