Social Determinants Matter For Hospital Readmission Policy: Insights From New York City
Social Determinants Matter For Hospital Readmission Policy: Insights From New York City
复制标题
DOI：
10.1377/hlthaff.2020.01742
复制
发表时间：
2021-04-01
期刊：
影响因子：
9.7
通讯作者：
Haberman, Merle
中科院分区：
文献类型：
--
作者：
Baker, Matthew C.;Alberti, Philip M.;Haberman, Merle
本研究评估了个人社会风险因素变量和健康社会决定因素 (SDOH) 措施对医疗保险和医疗补助服务中心 (CMS) 医院再入院减少计划 (HRRP) 中使用的再入院率和处罚的影响。使用纽约市 2012-16 年度出院数据，我们通过在四个地理层面构建 SDOH 评分和衡量个人层面的社会风险，增强 CMS 的心脏病发作、心力衰竭和肺炎再入院模型，从而预测 HRRP 处罚。在模型中包含额外的 SDOH 评分，尤其是那些使用最精细的地理数据构建的评分，以及社会风险因素变量，会极大地影响对治疗高 SDOH 评分患者比例最高的医院的预计处罚。即使我们根据《21 世纪治愈法案》将同侪群体分层纳入 HRRP 模型中，绩效仍出现改善。模型准确性的微小改进与预计性能的大幅变化相关。我们的结果表明，CMS 在 HRRP 再入院模型中持续遗漏相关患者和地理数据，从而错误地将 SDOH 和社会风险因素影响的处罚分配给了高风险患者比例最大的医院。
This study assessed the impact of individual social risk factor variables and social determinants of health (SDOH) measures on hospital readmission rates and penalties used in the Centers for Medicare and Medicaid Services (CMS) Hospital Readmissions Reduction Program (HRRP). Using 2012-16 hospital discharge data from New York City, we projected HRRP penalties by augmenting CMS's readmission model for heart attack, heart failure, and pneumonia with SDOH scores constructed at each of four geographic levels and a measure of individual-level social risk. Including additional SDOH scores in the model, especially those constructed with the most granular geographic data, along with social risk factor variables substantially affects projected penalties for hospitals treating the highest proportion of patients with high SDOH scores. Improved performance occurred even after we included peer-group stratification in the HRRP model pursuant to the 21st Century Cures Act. Small improvements in model accuracy were associated with substantial shifts in projected performance. Our results suggest that CMS's continued omission of relevant patient and geographic data from the HRRP readmission model misallocates penalties attributable to SDOH and social risk factor effects to hospitals with the largest share of high-risk patients.