Evaluating diagnosis-based case-mix measures: How well do they apply to the VA population?

BACKGROUND. Diagnosis-based ease-mix measures are increasingly used for pro vider profiling, resource allocation, and capitation rate setting. Measures developed in one setting may not adequately capture the disease burden in other settings. OBJECTIVES. TO examine the feasibility of adapting two such measures, Adjus ted Clinical Groups (ACGs) and Diagnostic Cost Groups (DCGs), to the Depart ment of Veterans Affairs (VA) population. RESEARCH DESIGN. A 60% random sample of veterans who used health care servi ces during FY 1997 was obtained from VA inpatient and outpatient administra tive databases. A split-sample technique Tvas used to obtain a 40% sample ( n = 1,046,803) for development and a 20% sample (n = 524,461) for validatio n. METHODS. Concurrent ACG and DCG risk adjustment models, using 1997 diagnose s and demographics to predict FY 1997 utilization (ambulatory provider enco unters, and service days the sum of a patients inpatient and outpatient vis it days), were fitted and cross-validated. RESULTS. patients were classified into groupings that indicated a populatio n with multiple psychiatric and medical diseases. Model R-squares explained between 6% and 32% of the variation in service utilization. Although repar ameterized models did better in predicting utilization than models with ext ernal weights, none of the models was adequate in characterizing the entire population. For predicting service days, DCGs were superior to ACGs in mos t categories, whereas ACGs did better at discriminating among Veterans who had the lowest utilization. CONCLUSIONS. Although "off-the-shelf" case-mix measures perform moderately well when applied to another setting, modifications may be required to accu rately characterize a population's disease burden with respect to the resou rce needs of all patients.