Improving statistical inference when interest focuses on the identification of extreme random effects in clustered data

当兴趣集中于识别集群数据中的极端随机效应时，改进统计推断

• 批准号: 10179473
• 负责人: Charles E McCulloch
• 金额: $ 26.49万
• 依托单位: UNIVERSITY OF CALIFORNIA, SAN FRANCISCO
• 依托单位国家: 美国
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-09-30 至 2024-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10179473
• 关键词: Address; Area; Computer software; Custom; Data; Goals; Grant; Health; Hospital Planning; Hospitals; Impaired cognition; Individual; Intervention; Investigation; Lead; Medicine; Methodology; Methods; Modeling; Motivation; Outcome; Patients; Performance; Physicians; Predictive Value; ROC Curve; Research; Rewards; Sample Size; Scheme; Sensitivity and Specificity; Statistical Models; Tail; Techniques; Uncertainty; United States Centers for Medicare and Medicaid Services; Weight; base; deviant; health plan; improved; interest; novel; older patient; readmission rates; simulation

PROJECT ABSTRACT Statistical models that generate predicted random effects are widely used to evaluate the status of and rank patients, physicians, hospitals and health plans from longitudinal and clustered data. Predicted random effects have been proven to outperform simpler approaches such as standard regression models, on average. These predicted random effects are often used to identify extreme or outlying values, such as elderly patients with rapid declines in their health or poorly performing hospitals. When interest focuses on the extremes rather than performance on average, there has been no systematic investigation of best approaches. We propose to develop novel methods for prediction of extreme or outlying values and systematically evaluate their performance using theoretical calculations, simulations and examples. Merely predicting extreme or outlying values is rarely sufficient and decision rules for identifying extreme values in a statistically rigorous manner are also needed. We will develop such approaches and provide easy- to-use software to implement the recommended methods.

项目摘要 产生预测随机效应的统计模型被广泛用于评估和排名 纵向和聚类数据的患者，医生，医院和健康计划。预测随机效应 事实证明，平均而言，诸如标准回归模型等更简单的方法。 这些预测的随机效应通常用于识别极端或外部值，例如老患者 他们的健康状况快速下降或表现不佳的医院。当兴趣关注极端时 比表现平均而言，最佳方法没有系统的投资。我们建议 开发用于预测极端或外围价值的新方法，并系统地评估其性能 使用理论计算，模拟和示例。 仅仅预测极端或外围价值很少是识别极端的决策规则 还需要以统计上严格的方式值。我们将开发此类方法并提供简单 - 使用软件以实现推荐的方法。