Department of Health Care Policy, Harvard Medical School, Boston, MA, USA
This study used 1992 and 1993 data from private employers to compare the performance of various risk adjustment methods in predicting the mental health and substance abuse expenditures of a nonelderly insured population. The methods considered included a basic demographic model, Ambulatory Care Groups, modified Ambulatory Diagnostic Groups and Hierarchical Coexisting Conditions (a modification of Diagnostic Cost Groups), as well as a model developed in this paper to tailor risk adjustment to the unique characteristics of psychiatric disorders (the “comorbidity” model). Our primary concern was the amount of unexplained systematic risk and its relationship to the likelihood of a health plan experiencing extraordinary profits or losses stemming from enrollee selection. We used a two-part model to estimate mental health and substance abuse spending. We examined the R2 and mean absolute prediction error associated with each risk adjustment system. We also examined the profits and losses that would be incurred by the health plans serving two of the employers in our database, based on the naturally occurring selection of enrollees into these plans. The modified Ambulatory Diagnostic Groups and comorbidity model performed somewhat better than the others, but none of the models achieved R2 values above .10. Furthermore, simulations based on actual plan choices suggested that none of the risk adjustment methods reallocated payments across plans sufficiently to compensate for systematic selection.
Please log in/register to access.