Many rural hospitals are in poor financial condition and low occupancy often contributes to this state. At the heart of the occupancy issue is whether the rural hospitals capture the market that is most readily available. In other words, do patients whose closest hospital is a rural hospital choose that hospital for care? Research supports the loss of those with complex care needs to larger, more distant hospitals. Research also suggests that urban hospitals are in competition with rural hospitals for patients with non-complex conditions. The research proposed here is a policy study which focuses on rural residents of Iowa and their choices of hospitals, with attention given to identifying bypassing behavior for non-complex conditions. The study's long-term objective is to present policy implications of rural resident's hospital decisions. Such decisions have implications for the facilities and for policymakers who are concerned with access, quality of care, and cost of care within the political reality that rural hospitals often hold a sacrosanct position in their communities.
The specific aim of this study is to answer the research question: What patient and hospital determinants explain local and non-local hospital use by rural residents of Iowa? The purpose of the research is four-fold: 1) to explore the prevalence and explanation for rural residents bypassing their local hospitals to get care in hospitals that are more distant; 2) to describe the rural residents who use local and non-local hospitals and the hospitals these patients select for care; 3) to determine what patient and hospital characteristics explain local and non-local hospital use; and 4) to determine what patient and hospital characteristics explain local and non-local hospital use by those with complex and non-complex conditions. This study is cross-sectional with descriptive, exploratory, and explanatory components. Patients are the unit of analysis. The study uses patient origin data to observe all general acute care hospital use by rural residents hospitalized in Iowa in 1988. This includes about 170,000 hospitalizations in Iowa's 126 hospitals. American Hospital Association and geographic data will augment the patient origin data. Variables will be limited to those cited by the literature and available from this data. Univariate, bivariate, and multivariate techniques will be used in the analysis. The use of regression, hierarchical log-linear, and logit is anticipated.
