Appl Health Econ Health Policy 3:107-114.
Stanford Prevention Research Center, Stanford University Medical School, Stanford, CA, USA
Population-based risk adjustment, as applied to reimbursement in managed care settings, may reduce pressures for adverse selection by managed care organisations. Using insurance claims data from 184 340 plan members, we compared the performance of three risk-adjustment methods. We present a model for measuring the impact of risk adjustment on the likelihood that individual members will be at risk for adverse selection. These results are compared with resource allocation based on age/sex. The predictive ability of alternative allocation schemes increased from an R(2) of 1.2% for age-sex allocation to 11.4% based on risk adjustment using diagnostic cost groups. However, the impact of risk adjustment on the proportion of members at risk for adverse selection was small. At an absolute threshold loss of $US2400 per year, 8.3% to 8.6% of members were at risk for adverse selection compared with 9.3% based on age-sex allocation. The limited impact of risk adjustment on the likelihood of adverse selection suggests that other strategies for reducing adverse selection may be required.
Please log in/register to access.