Detecting, Understanding, and Reducing Diabetes Belt Preventive Care Disparities

The economic costs of diabetes are huge with $176 billion in direct medical costs in 2012 and over 30% of Medicare expenditures spent on persons with diabetes. Of all healthcare expenditures attributed to diabetes, only 14% are spent to treat diabetes itself and the rest are spent to treat diabetic complications. Prevention of diabetic complications is important to reduce the economic burden of diabetes on national healthcare expenditures. American Diabetes Association (ADA) guidelines recommend preventive care to reduce the risk of these complications. Because these complications disproportionately affect the elderly, preventive care is also a special area of concern for the Centers of Medicare and Medicaid Services (CMS). Reducing health disparities and improving rates of preventive care utilization is national priority as reflected in Healthy People 2020 objectives to reduce health disparities and improve rates of diabetes preventive care (Objectives D4-D14). The "Diabetes Belt"h is a recently identified group of counties with especially high diabetes prevalence (>=11% compared to 8.5% average in the rest of the country). The Diabetes Belt is comprised of two medically disadvantaged populations, namely, low-income whites in Appalachia and blacks in the rural South. These populations have high mortality risk which is mainly attributable to high chronic disease burden. Our analysis of 2008 - 2010 BRFSS data shows that elderly persons (ages .>=65 years) with diabetes in the Diabetes Belt had almost 30% lower uptake of diabetes preventive services such as annual foot exam, annual eye exam, and diabetes self-management education compared to their counterparts in the rest of the country. They also had 12 - 23% higher rates of diabetes-related comorbidities such as heart attack, stroke, and health-related disability than those outside the Belt. The proposed study is a comprehensive research program that involves detecting, understanding, and reducing disparities in diabetes preventive care between the counties in the Diabetes Belt and the surrounding areas and within the Diabetes Belt itself. Detection involves measuring disparities in preventive service use and diabetic complications using Medicare claims data. Understanding involves identifying individual-level and county-level determinants of disparities. Finally, the third phase greducingh involves developing optimal policy options using Markov Decision Process (MDP) analysis. In addition to the traditional factors such as race/ethnicity, income, access to care, and healthcare supply, we will evaluate the role of public policy (e.g., Annual Wellness Visits) and local community partnerships in reducing disparities in diabetes preventive care use. In addition, we propose to use gefficiencyh in producing preventive services as a new gmetrich to measure disparities and to identify areas of targeted action.