Robinson J C, Luft H S, Gardner L B, Morrison E M
School of Public Health, University of California, Berkeley 94720.
Inquiry. 1991 Summer;28(2):107-16.
Biased selection can threaten the viability of multiple choice health systems unless payments to particular plans are adjusted to offset risk differences among employees. We report the results of a study designed to predict medical care utilization and expenditures for groups of fee-for-service plan (FFS) and health maintenance organization (HMO) enrollees, using characteristics commonly available in the personnel files of large employers. Simulation analyses indicate that the six-equation, maximum likelihood model predicts well for groups of 1,000 or more. Additional data are required to reduce prediction errors for smaller groups. This new methodology potentially allows risk-rating of employer contributions to competing health plans, based on the expected utilization of the individuals choosing each plan.
除非对特定计划的支付进行调整以抵消员工之间的风险差异，否则有偏差的选择可能会威胁到多种选择的医疗系统的生存能力。我们报告了一项研究的结果，该研究旨在利用大型雇主人事档案中通常可用的特征来预测按服务收费计划（FFS）和健康维护组织（HMO）参保人群的医疗服务利用情况和支出。模拟分析表明，六方程最大似然模型对1000人或更多人群的预测效果良好。需要更多数据来减少对较小群体的预测误差。这种新方法有可能根据选择每个计划的个人的预期利用率，对雇主向竞争性健康计划的缴费进行风险评级。
