Anticipated Impacts on Veterans Health Care. Changes in the purchased care program might affect the cost, quality, and accessibility of elective coronary revascularization procedures sponsored by VHA. The impact may be greatest among rural and highly rural Veterans as they are more reliant on purchased care. Background. VHA purchases 19% of the elective Coronary Artery Bypass Graft (CABG) and 22% of the elective Percutaneous Coronary Intervention (PCI) that it sponsors for veterans under age 65. Half of this care is provided to rural veterans. The effect of the VA pilot program of selective contracting on cardiac care has not been studied. VHA is undertaking a national program of selective contracting to contain purchased care cost. Selective contracting can improve quality and lower costs, but not all managed care organizations have succeeded in selecting high quality providers. Quality measurement is difficult when a plan relies exclusively on its own members'experience, as there may be too few observations for statistical confidence. Previous research has generated conflicting findings about the significance of the additional travel that is imposed by the reduction in the number of providers that results from selective contracting. Objectives. The cost-effectiveness of polices to select providers based on easily observed criteria will be evaluated. Policies include: use of a VA provider located within a reasonable travel distance, use of a VA provider that meets a minimum volume criterion, use of a non-VA provider that meets a minimum annual volume criterion, and use of a non-VA provider that reported a better than average 30-day risk-adjusted mortality in Hospital Compare. Supplementary project aims are to understand the effect of policy on components of cost (procedure, follow-up and patient incurred travel cost, including the value of patient time), outcomes (survival, probability of repeated procedure), and access (distance traveled and the rate of obtaining an elective procedure in patients with coronary artery disease). To estimate the discounts that might be negotiated with providers, cost of cardiac revascularization services provided under the pilot VHA contracting program, Project HERO, will be studied. The budget impact of cost-effective policies will be projected. Methods. Regional differences in the provision of non-emergency cardiac revascularization procedures to patients under age 65 will be used to estimate the impact of the policies on cost and survival. Multivariate cost and survival regressions will be used to estimate the difference between care that was consistent with each policy and care that was not policy concordant. These regressions will be risk adjusted for patient characteristics obtained from VA administrative data, the VA Corporate Date Warehouse, and the VA cardiac catheterization lab registry. Cost will include the full cost of travel, including veteran incurred travel cost and the value of patient time spentin transit (i.e., the societal perspective). Instrumental variables will be used to correct for selecton bias. Long-term outcomes will be estimated by applying the estimated hazard ratio to age-adjusted life-expectancy. The effect of travel distance on the probability that VHA patients with coronary artery disease obtain cardiac revascularization will be estimated to learn if the extra travel caused by selective contracting might prevent some patients from getting a procedure.
The cost of purchased care is growing and accounts for 10% of the VHA budget. It includes $170 million per year for Coronary Artery Bypass Grafts and Percutaneous Coronary Interventions. Much of this care can be scheduled, giving VHA the option of selecting providers with higher quality and lower cost. VHA is setting up a national program of selective contracting to contain costs. This program needs to know the best way to select providers. Cost and outcomes of current cardiac services will be studied to see how care that meets selection criteria differs from care that does not, while controlling for differences between patients. Criteria to be studied include: selecting the nearest VA facility, selecting a VA or non-VA facility that meets a minimum volume criterion, or selecting a non-VA provider with better than average outcomes. The direct cost of transportation and the value of Veteran time spent in transit will be considered. Cost-effectiveness analysis will identify the selective contracting policy that yields the best value for a given expenditure.