Service duplication within urban hospital clusters.

Background: Approximately 80% of multihospital system member hospitals in U.S. urban areas are clustered with other same-system member hospitals located in the same market area. A key argument for clustering is the potential for reducing service duplication across cluster members.

Purpose: The aim of this study is to examine the effects of characteristics of hospital clusters on service duplication within 339 hospital clusters in U.S. metropolitan statistical areas and adjacent counties in 2002.

Methodology/approach: Ordinary least squares regression is used to estimate the relationship between cluster characteristics in 1998 and duplicated services per cluster member in 2002.

Findings: Duplication is higher in hospitals clusters with higher case mix index and higher bed size range. Duplication is lower in hospital clusters with more members, for-profit ownership, and more geographic dispersion.

Practice Implications: Increases in the size of hospital clusters allow more opportunities for service rationalization. For-profit clusters may be innovators in rationalization activity, and they should be studied in this regard. Clusters with a higher case mix, lower geographic dispersion, and hub-and-spoke design (with high bed-size range) may find service reallocation less feasible.