Methodological limitations in case mix hospital reimbursement, with a proposal for change.

We have found five methodological limitations in the creation and implementation of the diagnosis related group (DRG) patient classification system, which is used to define a hospital's case mix. There are four methodological limitations in the system that Klastorin and Watts have proposed to identify hospital peer groups. We conclude that the effects of these limitations should be sought, and we propose studies to measure their extent. We also propose that these two approaches can be combined to create an improved hospital reimbursement program that accurately measures differences between hospitals caused by case mix and peer group characteristics.