Nonhierarchical vs hierarchical modeling to measure hospital performance

Editor's Note

In this study, nonhierarchical modeling to measure hospital performance frequently misclassified average-quality hospitals as low quality, and hierarchical modeling misclassified low-quality hospitals as average.

At low hospital case volumes, hierarchical modeling missed 90.6% of low-quality hospitals, and nonhierarchical modeling missed 65.3%. However, 38.9% of hospitals classified as low-quality using nonhierarchical modeling were actually average quality, compared to 5.3% using hierarchical modeling.

The validity of basing reimbursement on hospital performance is grounded in the accuracy of performance measurement. Assuming that the consequences of misclassifying an average-quality hospital as low quality outweigh those of misclassifying a low-quality hospital as average, hierarchical modeling may be the better choice for performance measurement, the authors say.

Impact of Risk Adjustment, Hospital Volume, and Hospital Performance