This project will test the underlying hypothesis that hospital behavior and performance are influenced by the spatial configuration of the local hospital market. Hospital performance has typically been examined with cross-sectional approaches that assume each hospital is independent. Physicians and hospital administrators, however, are acutely aware of the locations of alternative institutions. The spatial configuration of institutional providers has been studied by medical geographers, but their focus has been on time and distance within a market area. This project will combine the cross section and spatial configuration approaches to explicitly include hospital interdependencies in the analysis of hospital behavior. The spatial framework will be applied to three broad analytic questions: (1) What factors influence the acquisition, availability, and use of specialized clinical services and facilities; (2) What factors underlie the observed negative relation between hospital volume of specific patients and mortality; and (3) What factors influence the effectiveness of regulatory strategies in altering the distribution of services among hospitals. The spatial framework methodology will also be used to test the concept that some services and facilities are used by hospitals in a competitive fashion, while others serve as complements. The project draws upon research in economics, sociology, geography, quality assessment, hospital administration and organizational behavior. Multivariate regression models with data from all U.S. community hospitals will be used to examine factors that influence the distribution and use of specialized clinical services in 1972 and 1982. These data will be linked to information on the number of patients with selected procedures or medical diagnoses and their outcomes. The validity of the results of these studies will be tested with a series of localized case studies.
