The Effects on Va Hospital Performance of a Prospective Resource Allocation System Based on a Case-Mix Weighted Measure of Workload (Veterans Administration).

The system by which resources are allocated to Veterans Administration (VA) hospitals includes explicit and implicit incentives. If a new allocation system changes the incentives, it is expected that hospital performance will be affected. In fiscal year 1981 the VA introduced a new resource allocation method which used a case-mix weighted measure of workload--Prospective Reimbursement Intradistrict Methodology (PRIME)--to four hospitals. The problem for this study was to determine whether VA hospital behavior had changed as a result of PRIME and , if so, how. The objectives were: (1) to develop an economic model that identified possible effects of PRIME, and (2) to document and measure these effects. Initially, the economic model specified that the objective function of VA hospitals is composed of two goods: (1) extra workload, and (2) "other preferred expenses" (the allocation of resources to other than productivity increasing inputs or expenses). The VA hospitals' objective function is constrained, however, by the discretionary budget. Analysis determined that PRIME would result in two simultaneously occurring changes: (1) a change in the relative prices of the two goods, and (2) a change in the means of measuring workload for purposes of reimbursement. These, it was hypothesized, would affect hospital behavior by increasing workload, the reimbursable expenses, and decreasing "other preferred expenses," the nonreimbursable expenses. Specific variables were identified and constructed to detect changes in workload. These were (1) case-mix index, (2) intensity of service index, (3) length of stay index, (4) one day lengths of stay, and (5) number of treated cases. These dependent variables, as well as the hospital control, trend, and treatment variables, were measured directly, or by their surrogates, in three experimental and three control hospitals for the pre- and post-treatment periods. The analysis led to three conclusions: (1) PRIME, the new resource allocation method, did alter hospital behavior in the expected direction; (2) hospital specific characteristics strongly influence workload levels; and (3) seasonal time and trend factors affect hospital behavior but to a lesser extent than do hospital specific characteristics.PhDHealth care managementUniversity of Michiganhttp://deepblue.lib.umich.edu/bitstream/2027.42/159800/1/8402345.pd