A method for constructing case-mix indexes, with application to hospital length of stay

This article presents the methodological development of an index for case-mix adjustment of hospital data exemplified by our construction of an index for studying length of stay. We describe the development and evaluation of this index, including internal and external validation procedures, and show an example of its use in a policy-relevant context by applying it to the analysis of length-of-stay differences between investor-owned and voluntary hospitals. Some advantages of this approach to adjusting for case mix are applicability to many hospital or patient output measurements/diagnostic scheme situations; usefulness in reducing heterogeneity in other case-mix adjustments, e.g., the Diagnosis-Related Group (DRG) approach; interpretation possibilities; production of a single score for each patient/hospital; statistical approach allowing more accurate and reliable interpretation of hospital and patient output measurements, ability to deal with hospital deaths; and consideration of the complete set of secondary diagnoses. We also suggest other possible uses of this approach.