Establishing Appropriate Peer Group Method for Comparing Healthcare Quality
Measuring and reporting healthcare facility performance via clinical measures of quality has become a major strategic initiative in improving the quality of healthcare. Establishing appropriate peer groups can help make equitable comparisons across hospital or healthcare systems, however, peer groups as commonly constructed have several undesirable characteristics. This study sought to develop a methodology for constructing peer groups for VA medical centers. Unlike peer groups based on traditional cluster analysis, this new methodology developed a peer group customized to each medical center by identifying the “nearest neighbor” medical centers, according to distance from each other and selected characteristics for comparison (i.e., patient complexity, patient reliance on VA care, size, community, and environment). Using VA data from FY05, investigators compared the nearest-neighbor peer groupings for 133 VAMCs by the selected characteristics.
Findings show that one of the advantages of the nearest-neighbor method is that the peer groups are more refined, reflecting the multi-dimensional diversity of healthcare providers. Moreover, the nearest-neighbor method incorporates the practical consideration that healthcare facilities or systems may have structural and patient-based differences that cannot be changed, but do affect financial or quality outcomes. Authors suggest that nearest-neighbor peer groups may be more appealing to some researchers and administrators than standard cluster analysis, and thus may strengthen organizational buy-in for financial and quality comparisons.
Byrne M, Daw C, Nelson H, Urech T, Pietz K, and Petersen L. A method to develop health care peer groups for quality and financial comparisons across hospitals. Health Services Research April 2009;44(2 Pt 1):577-592.
This study was partly funded by HSR&D. Drs. Daw, Urech, Pietz, and Petersen are part of HSR&D’s Houston Center for Quality of Care and Utilization Studies.