Estimating Medicaid Costs for Cardiovascular Disease: A Claims-Based Approach
Haber, S. G., & Gilman, B. (2005, December). Estimating Medicaid Costs for Cardiovascular Disease: A Claims-Based Approach. Presented at American Public Health Association Annual Meeting, Philadelphia, PA.
Cardiovascular diseases, including coronary heart disease, stroke, hypertension, and congestive heart failure, are leading sources of mortality and morbidity in the United States. The Medicaid program serves a number of populations that are at high risk for these diseases --racial/ethnic minorities, elderly and disabled people, and people with low incomes. Rapid increases in Medicaid expenditures have strained state and federal budgets. Because these diseases are preventable, understanding their impacts on Medicaid expenditures can identify the value of reducing program costs through prevention efforts. We will describe our methodology for estimating per capita and total Medicaid costs associated with cardiovascular diseases by applying econometric techniques to standardized Medicaid claims data (the Medicaid Analytic Extract, or MAX, files). Beginning in 1998, MAX data are available for all states from the Centers for Medicare and Medicaid Services (CMS). Estimates will be presented for 3-6 states. In addition to statewide averages, we will present estimates for subpopulations based on age, gender, race/ethnicity, Medicare dual eligibility status, and local area of residence. We will review the benefits and challenges of using claims data to develop cost-of-illness estimates. Compared to survey data, claims data are more complex and expensive to work with. In addition, MAX data do not provide a complete picture of Medicaid utilization in states with high rates of enrollment in capitated managed care plans. However, claims-based cost estimates have several advantages over survey-based estimates, including the ability to support state-specific cost and prevalence estimates, as well as estimates for subpopulations of interest. Survey-based estimates typically assume that variations in the cost of cardiovascular disease across states and across subpopulations are explained primarily to disease prevalence, rather than the cost per person treated. The results of these Medicaid claims analyses will be used to validate this assumption.