The efficiency of hospital-based clusters: evaluating system performance using data envelopment analysis.

Background: The rapid increase in the number of hospitals becoming members of multihospital systems in recent decades has led to the formation of local and regional clusters that have the potential to function as regional systems, a model long advocated as a policy strategy for improving health system performance. Purpose: This study addresses both cluster efficiency and the hierarchical configuration with which hospitals are grouped into clusters. Methodology/Approach: This study uses 2004 data from the American Hospital Association Annual Survey multihospital system designations updated to 2005. Efficiencies are measured using data envelopment analysis. Principal Findings: The data envelopment analysis results show that 20 clusters or 5.8% of the sample of 343 clusters are highly efficient; the remaining 323 or 94.2% of the clusters received lesser efficiency scores, averaging 0.73 on the data envelopment analysis measure. The study found the number of beds in the primary hospitals and the percentage of hospitals in the clusters that were urban, two of three variables that reflect patterns of regional model service configurations, to be significantly correlated with cluster efficiency. Conclusion: Results suggest that many hospital clusters have evolved service configurations that are consistent with historically conceptualized regional organizational forms and that the particular regional pattern of distributing service capacities across cluster members might contribute to measured performance. The study also confirms the applicability of data envelopment analysis for assessing the performance of complex, multiunit organizations. [ABSTRACT FROM AUTHOR]