The Extent of Role Differentiation among Hospitals

The case mix of 65 western Pennsylvania hospitals is studied from Hospital Utilization Project data comprising a quarter of a million patient records, and five measures are constructed to describe the inpatient population on the basis of diagnoses and surgical procedures. The case-mix variation defined by these measures is then analyzed in a series of regressions on selected institutional characteristics and on demographic and locational variables, revealing urban vs. rural location to be an important determinant. While size, number of facilities and services, and teaching status of a hospital are shown to be correlated with case mix, they are found to explain too small a portion of the variation to be satisfactory surrogates for case mix.