Adjusted Mortality Rates Are Lower For Medicare Advantage Than Traditional Medicare, But The Rates Converge Over Time
Adjusted Mortality Rates Are Lower For Medicare Advantage Than Traditional Medicare, But The Rates Converge Over Time
复制标题
DOI：
10.1377/hlthaff.2018.05390
复制
发表时间：
2019-04-01
期刊：
影响因子：
9.7
通讯作者：
Landon, Bruce E.
中科院分区：
文献类型：
--
作者：
Newhouse, Joseph P.;Price, Mary;Landon, Bruce E.
根据年龄、性别和医疗补助状况进行调整后，医疗保险优势项目的总体死亡率多年来一直低于传统医疗保险项目。因此，医疗保险优惠政策的有利选择问题引起了人们的广泛关注。共同研究设计用于估计医疗保险优势对利用和结果的因果影响，比较了在医疗保险优势登记之前和之后的新医疗保险优势受益人与选择保留传统医疗保险的受益人。尚未被研究的是最初选择参加医疗保险优惠计划的人群与选择传统医疗保险的人群的死亡率经历。在这项研究中，我们发现新加入Medicare Advantage的队列的调整死亡率最初远低于新加入传统Medicare的队列，但五年后差异显著减小。因此，常见的研究设计是有缺陷的，因为它假设在加入医疗保险优势后，死亡率风险的任何初始差异保持不变。换句话说，那些最初选择医疗保险优势的人在五年内会比传统的医疗保险受益人病得更重。如果观察时间超过五年，死亡率是否会完全趋同，这是一个需要进一步研究的课题。
Overall mortality rates, adjusted for age, sex, and Medicaid status, in Medicare Advantage have been below those in traditional Medicare for many years. Much attention has been paid to the resulting issue of favorable selection in Medicare Advantage. The common study design used to estimate causal effects of Medicare Advantage on utilization and outcomes compares new Medicare Advantage beneficiaries immediately before and after enrollment in Medicare Advantage with beneficiaries who choose to remain in traditional Medicare. What has not been studied is the mortality experience of a cohort that initially chooses enrollment in Medicare Advantage versus one that chooses traditional Medicare. In this study we found that the adjusted mortality rate of a cohort newly enrolled in Medicare Advantage was initially well below that of a cohort newly enrolled in traditional Medicare, but the difference markedly decreased after five years. As a result, the common study design is flawed because it assumes that any initial difference in mortality risk remains constant after enrollment in Medicare Advantage. In other words, those initially choosing Medicare Advantage become sicker relative to traditional Medicare beneficiaries over five years. Whether the mortality rates would fully converge if a period longer than five years were observed is a topic for further research.