Understanding hospital value: provider, hospital and community effects

摘要 美国的政策制定者和支付者现在集中关注医疗保健价值，通常定义为 每花费一美元所实现的质量，作为同时提高护理质量和减少或 稳定成本。2017年，所有付款人中有34%的付款是通过基于价值的模式进行的，而只有 41%是通过传统的收费服务模式。住院治疗占美国医疗保健的三分之一 它拥有最长期的基于价值的模型。然而，迄今为止， 住院病人护理方案的结果好坏参半或令人沮丧。如果不清楚 因素有助于医院和社区提供高价值的护理，基于价值的支付运动可能不会 成功这一更新建议扩展了第一次资助的R 01中成功完成的工作， 与再入院相关的医院和社区因素（质量和成本的一个具体例子）。期间 该基金已经在《新英格兰杂志》等期刊上发表了14篇文章， 医学和JAMA以及174次引用，我们开发了一个强大的数据基础设施， 超过70家医院和社区机构。在这项研究中，我们将建立在这些数据的基础上， 基础设施，以探索与整体价值相关的医生，医院和社区因素， 住院医疗在目标1中，我们将使用医疗保险和医疗补助服务中心（CMS）的星星评级 由我们的团队开发的作为主要质量成果的衡量标准（它汇总了57项衡量标准的绩效 死亡率，再入院，安全性，经验，有效性和成像的使用），以及CMS医疗保险 每项福利措施的支出是我们的主要成本结果。在目标2中，我们将使用其他质量衡量标准， 例如对死亡率的性能，或特定条件下的总体质量，以及其他成本指标， 作为我们团队开发的CMS特定条件风险标准化支付措施。在这两个目标中，我们 将探讨提供者、医院和社区因素对结果的影响，以确定 它们有助于医疗保健价值并介导患者因素。在目标1-2中，我们将使用 已经在CMS、美国国立卫生研究院、人口普查局、 美国医院协会，城市健康仪表板，县健康排名和其他。在目标3中， 我们将通过直接调查高价值和低价值医院的具体做法来确定新的预测因素， 现有数据集中没有：例如，质量基础设施、董事会和工作人员 参与、电子健康记录能力、数据基础设施、社区协调等。我们 将评估这些新的预测因子与目标1和目标2中使用的结果之间的关联。结束时 在赠款期间，我们希望对医疗保健价值有一个细致入微的理解， 临床医生和政策制定者，以改善所有患者的医疗保健服务。

Abstract Policymakers and payers in the US are now focusing intensively on healthcare value, commonly defined as quality achieved per dollar spent, as a means of simultaneously improving quality of care and reducing or stabilizing costs. In 2017, 34% of payments across all payers were through value-based models, while only 41% were through traditional fee-for-service models. Inpatient care accounts for one third of all US health expenditures and has had the most long-standing value-based models. To date, however, value-based programs for inpatient care have had mixed or discouraging results. Without a clear understanding of what factors help hospitals and communities provide high value care, the value-based payment movement may not succeed. This renewal proposal extends work successfully done in the first funded R01, which explored hospital and community factors associated with readmission (one specific example of quality and cost). During that grant, which has already generated 14 publications in journals such as The New England Journal of Medicine and JAMA and 174 citations, we developed a robust data infrastructure that links 6.8 million hospitalizations to over 70 hospital and community factors. In this study, we will build upon that data infrastructure to explore practitioner, hospital and community factors associated with the overall value of inpatient healthcare. In Aim 1, we will use the Centers for Medicare & Medicaid Services (CMS) Star Ratings measure developed by our team as our main quality outcome (which aggregates performance on 57 measures of mortality, readmission, safety, experience, effectiveness, and use of imaging), and the CMS Medicare Spending per Beneficiary measure as our main cost outcome. In Aim 2 we will use other measures of quality, such as performance on mortality, or overall quality for specific conditions, and other measures of cost, such as the CMS condition-specific risk-standardized payment measures developed by our team. In both aims, we will explore the influence of provider, hospital and community factors on outcomes to determine the degree to which they contribute to healthcare value and mediate patient factors. In Aims 1-2 we will use measures already available in a number of datasets from CMS, the National Institutes of Health, the Census Bureau, the American Hospital Association, the City Health Dashboard, the County Health Rankings and others. In Aim 3, we will identify new predictors by directly surveying high and low value hospitals about specific practices that are not available in existing datasets: for instance, aspects of quality infrastructure, Board and staff engagement, electronic health record capabilities, data infrastructure, community coordination and others. We will assess the association of those new predictors with the outcomes used in Aims 1 and 2. By the end of the grant period, we expect to have developed a nuanced understanding of healthcare value that will enable clinicians and policymakers to improve healthcare delivery for all patients.