External Determinants of Non-Elderly Veterans' Demand of VA Health Care
The rapidly changing environment of the U.S. health care system, including federal and state health insurance reforms, rising out-of-pocket costs, and increased coverage limitations, is likely to impact Veterans' use of VA health care. In FY2010, approximately 30 percent of the 7.8 million enrolled Veterans, received care outside the VA or sought no care. While attention has been paid to dual coverage from Medicare, little is known about non-VA health care use among the non-elderly Veterans (age < 65 years), who comprise 55 percent of all enrollees generally uncovered by Medicare. A careful examination of key external factors is critical to better understand regional variations in VA health care demand and the potential impact from major ongoing and anticipated initiatives in the public and private health care sectors, such as the 2010 Patient Protection and Affordable Care Act (ACA). Of particular importance are low-income Veterans who comprise 60 percent of VA health care enrollment, and whose health care use may be most sensitive to out of pocket cost increases, availability of other insurance, economic factors and provider availability. Our proposed study will address this knowledge gap by developing a national model to estimate the sensitivity of VA demand for non-elderly Veterans to a range of external factors, including public assistance for uninsured (Medicaid), out-of-pocket costs for non-VA care, economic environment (unemployment) and non-VA provider availability. For Veterans enrolled for VA health care, VA demand will be measured by the total value ($) and volume of services of VA health care use. Data will be obtained from VA administrative databases covering inpatient care, outpatient visits and prescription drugs (FY 2008-2014). For Veterans with Medicaid and/or Medicare coverage, we will use data on non-VA care to examine reliance on VA (i.e., share of total care obtained in VA). Data on measures of external determinants, categorized into four distinct domains -- public policy, non-VA patient costs, economic environment, and local non-VA provider availability will be obtained from a range of patient- and area-level secondary data sources. We will quantify the improvement of the expanded model in projecting resource use by VAMC and VISN. We will test for systematic differences in demand across three high priority subpopulations whose VA use might be especially sensitive to external factors - rural residents, racial/ethnic minorities and women - and clinical cohorts with differing mix of inpatient and outpatient services: diabetes, congestive heart failure (CHF), serious mental illness and cancer. In the quasi-experimental setting of Massachusetts health reform, we will evaluate its impact on VA demand and reliance. As a secondary aim, we will evaluate the early impact of ACA insurance expansion on VA enrollment and utilization in FY2014. Our main aims are to (1) develop a base model of VA health care demand and reliance adjusted for patient case-mix, (2) extend the base model to include measures of external determinants, (3) apply the extended model to examine differential VA demand and reliance among rural residents, racial/ethnic minorities, women, and low income Veterans, and (4) evaluate the impact of Massachusetts health reform on VA demand and reliance. The proposed study responds to VA's strategic initiative to anticipate and proactively prepare for the needs of Veterans by the development of tools, analytical methods, data sources, and processes, such as predictive modeling, forecasting, business intelligence and data mining. We have partnered with the Office of the ADUSH for Policy and Planning and Office of Productivity, Efficiency and Staffing (OPES) to inform current operations models used for budget formulation, resource planning and the development of protocols for productivity standards for physicians and staffing (i.e., panel size).