Kim Nancy, Bernheim Susannah M, Ott Lesli S, Han Lein, Spivack Steven B, Xu Xiao, Volpe Mark, Liu Alex, Krumholz Harlan M
*Center for Outcomes Research and Evaluation, Yale-New Haven Hospital †Section of General Internal Medicine, Department of Internal Medicine, Yale University School of Medicine, New Haven, CT ‡Centers for Medicare & Medicaid Services, Baltimore, MD §Department of Obstetrics, Gynecology and Reproductive Sciences ∥Physician Associate Program, Yale University School of Medicine, New Haven, CT ¶Cipher Health, New York, NY #Section of Cardiovascular Medicine **Robert Wood Johnson Clinical Scholars Program, Department of Internal Medicine, Yale University School of Medicine ††Department of Health Policy and Administration, Yale School of Public Health, New Haven, CT.
Med Care. 2015 Jun;53(6):542-9. doi: 10.1097/MLR.0000000000000361.
Understanding both cost and quality across institutions is a critical first step to illuminating the value of care purchased by Medicare. Under contract with the Centers for Medicare and Medicaid Services, we developed a method for profiling hospitals by 30-day episode-of-care costs (payments for Medicare beneficiaries) for acute myocardial infarction (AMI).
We developed a hierarchical generalized linear regression model to calculate hospital risk-standardized payment (RSP) for a 30-day episode for AMI. Using 2008 Medicare claims, we identified hospitalizations for patients 65 years of age or older with a discharge diagnosis of ICD-9 codes 410.xx. We defined an AMI episode as the date of admission plus 30 days. To reflect clinical care, we omitted or averaged payment adjustments for geographic factors and policy initiatives. We risk-adjusted for clinical variables identified in the 12 months preceding and including the AMI hospitalization. Using combined 2008-2009 data, we assessed measure reliability using an intraclass correlation coefficient and calculated the final RSP.
The final model included 30 variables and resulted in predictive ratios (average predicted payment/average total payment) close to 1. The intraclass correlation coefficient score was 0.79. Across 2382 hospitals with ≥ 25 hospitalizations, the unadjusted mean payment was $20,324 ranging from $11,089 to $41,897. The mean RSP was $21,125 ranging from $13,909 to $28,979.
This study introduces a claims-based measure of RSP for an AMI 30-day episode of care. The RSP varies among hospitals, with a 2-fold range in payments. When combined with quality measures, this payment measure will help profile high-value care.
了解各机构的成本和质量是阐明医疗保险所购买医疗服务价值的关键第一步。根据与医疗保险和医疗补助服务中心签订的合同，我们开发了一种方法，通过急性心肌梗死（AMI）30天护理期间的费用（医疗保险受益人的支付金额）对医院进行剖析。
我们开发了一种分层广义线性回归模型，以计算AMI 30天护理期间的医院风险标准化支付（RSP）。利用2008年医疗保险理赔数据，我们确定了年龄在65岁及以上、出院诊断为ICD-9编码410.xx的患者的住院情况。我们将AMI护理期间定义为入院日期加上30天。为反映临床护理情况，我们忽略或平均了地理因素和政策举措的支付调整。我们对AMI住院前12个月（包括住院期间）确定的临床变量进行了风险调整。利用2008 - 2009年的合并数据，我们使用组内相关系数评估测量可靠性，并计算最终的RSP。
最终模型包含30个变量，预测比率（平均预测支付/平均总支付）接近1。组内相关系数得分是0.79。在2382家住院病例≥25例的医院中，未调整的平均支付为20324美元，范围从11089美元到41897美元。平均RSP为21125美元，范围从13909美元到28979美元。
本研究引入了一种基于理赔数据的AMI 30天护理期间RSP测量方法。各医院的RSP有所不同，支付范围相差两倍。当与质量指标相结合时，这种支付指标将有助于剖析高价值医疗服务。
