Implementing Risk-Adjustment for Medicaid

Can a risk adjustment model be developed which allows state Medicaid programs to adequately adjust for the risk of enrollees? Researchers refined the Disability Payment System (DPS), compared it with other risk assessment systems, and developed a model that will work well for AFDC recipients and children with special health care needs, as well as adults with disabilities. In addition, they developed methods to measure and adjust for diagnostic discovery (the tendency of HMOs to report diagnostic information on encounter data more intensively than it was reported in the fee-for-service (FFS) data on which the payment weights are estimated), assembled and developed coding guidelines and a training curriculum for important diagnoses, and developed formulae and payment factors for newly eligible recipients based on analysis of eleven-state databases developed previously. Finally, the researchers constructed a variety of biased sub-samples to simulate enrollment in HMOs and analyzed the extent to which state-specific weights produce simulated plan payments different from those estimated from the eleven-state data set. The objective of the project was to develop tools that state Medicaid programs need to successfully implement risk-adjusted payment systems and to demonstrate the applications of these tools in Colorado and Missouri.