Methods Core C

核心C方法

• 批准号: 10712643
• 负责人: Alistair James O'MALLEY
• 金额: $ 31.49万
• 依托单位: DARTMOUTH COLLEGE
• 依托单位国家: 美国
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-08-01 至 2028-07-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10712643
• 关键词: Accounting; Address; Affect; Alzheimer' s disease care; Alzheimer' s disease diagnosis; Alzheimer' s disease related dementia; Area; Assisted Living Facilities; Attention; Automation; Caring; Chronology; Collaborations; Communities; Comparative Effectiveness Research; Computer software; Computing Methodologies; Data; Data Analytics; Detection; Development; Diagnosis; Diagnostic; Diagnostic Procedure; Diagnostic tests; Disease; Disparity; Environment; Experimental Designs; Goals; Health; Health Personnel; Health Services Accessibility; Health Status; Health system; Healthcare; Heterogeneity; Home; Hospitals; Human Resources; Inequity; Link; Machine Learning; Measurement; Measures; Medical; Medicare claim; Methods; Minority; Modeling; Network-based; Neurologist; Nursing Homes; Outcome; Output; Pathway Analysis; Patient-Focused Outcomes; Patients; Persons; Physicians; Provider; Quality of Care; Race; Research; Research Design; Risk; Risk Factors; Secure; Services; Specialist; Statistical Methods; Statistical Models; Structure; Time; Variant; Work; algorithmic bias; care outcomes; community based service; complex data; dementia care; electric impedance; health care availability; health care disparity; health disparity; health inequalities; improved; innovation; insight; interest; machine learning algorithm; medical specialties; member; mortality; network models; novel; patient population; patient responsibilities; provider networks; rurality; sex; socioeconomic disparity; statistics; synergism; trend

SUMMARY – CORE C The Statistics and Networks Methods Core (Core C) of this PPG will actively collaborate with each of the projects by helping them use statistical, network, and computational methods in their research; by advancing statistical and network methods to expand the scope of their accomplishments; and by helping them solve any unexpected problems that arise. Core C staff will collaborate with Project Leads on study designs, and will supervise Core B programmers to overcome challenges and implement methods that Core C develops inside the secure computing environment at The Dartmouth Institute, so that these can be used to analyze the patient identifiable data relevant to each project. Motivated by the recent surge of interest in health and health care for Alzheimer’s disease and related dementias (ADRD), a growing concern for health disparities and their exacerbation by algorithmic bias, and advances in statistical methods and network analysis, Core C will advance methods that will make all four projects more effective and impactful. Core C will focus on: (1) statistical models for complex data structures that expand understanding of disparities and intersectional disparities, (2) development of networks that better capture the structure of relationships linking physicians of multiple specialties through the interdependencies in their medical practice, and (3) construction of novel quality of care measures and models of network data. These methods will be applied to patients with claims- based diagnoses of ADRD or who are at risk for developing ADRD. A common new theme in this renewal PPG is the use of comparative-effectiveness research, as it overcomes bias and enhances statistical power to quantify disparities and variations by modeling hierarchically structured and other forms of dependent data and by forming inferences that are causally-defendable. Core C key personnel have made substantial progress in establishing methods that use directional relationship information from patient-physician encounters in Medicare claims to form innovative physician networks with directed edges. They will enhance this work by using specialty and services-performed data to establish even more informative networks and network-based measures of access to care and other important aspects of health care. Statistical models will also be devised that use these networks to explain additional variation in outcomes and disparities, while longitudinal models of the network itself will provide insights into trends in and the dynamic association of factors affecting ADRD diagnosis and care. These innovations will allow the PPG-supported projects to delve deeper than would be possible otherwise. Results and products (e.g., software) will be widely disseminated to research and stakeholder communities. Core C will have a major supportive and innovative impact on the research of the four projects and on the wider research communities working in ADRD, health disparities, and variations in care with a goal of improving the U.S. health system.

摘要-核心C 该PPG的统计和网络方法核心(核心C)将积极与每个 通过帮助项目在他们的研究中使用统计、网络和计算方法；通过推进 统计和网络方法，以扩大他们的成就范围；并通过帮助他们解决任何 出现意想不到的问题。核心C人员将与项目负责人就研究设计进行合作，并将 监督Core B程序员克服挑战并实施Core C内部开发的方法 达特茅斯研究所的安全计算环境，以便这些可以用来分析患者 与每个项目相关的可识别数据。受最近对健康和医疗保健的兴趣激增的推动 阿尔茨海默病和相关痴呆(ADRD)，人们越来越关注健康差距和他们的 由于算法偏差以及统计方法和网络分析的进步，核心C将 先进的方法，将使所有四个项目更有效和更有影响力。核心C将专注于：(1) 用于复杂数据结构的统计模型，可扩展对差异和交叉性的理解 差异，(2)发展网络，更好地捕捉联系医生的关系结构 通过医疗实践的相互依存实现多个专科；(3)构建新型 网络数据的护理质量衡量标准和模型。这些方法将适用于有索赔要求的患者- 基于对ADRD的诊断或有发展ADRD风险的人。更新PPG的一个共同的新主题 是比较有效性研究的使用，因为它克服了偏见并增强了统计能力 通过对分层结构和其他形式的相关数据进行建模来量化差异和差异 通过形成可以在因果关系上辩护的推断。核心C关键人员在以下方面取得了实质性进展 建立使用来自患者-医生接触的方向关系信息的方法 联邦医疗保险声称形成了具有定向优势的创新医生网络。他们将通过以下方式加强这项工作 使用专业和服务执行数据来建立更具信息性的网络和基于网络的 获得护理的措施和卫生保健的其他重要方面。还将设计统计模型 使用这些网络来解释结果和差异的额外差异，而纵向模型 网络本身将提供对影响ADRD的因素的趋势和动态关联的洞察 诊断和护理。这些创新将使PPG支持的项目能够比预期的更深入 也有可能不是这样。结果和产品(例如，软件)将广泛传播到研究和 利益相关者社区。核心C将对全球经济一体化的研究产生重大支持和创新影响 四个项目，以及在ADRD工作的更广泛的研究社区，健康差距，以及 医疗保健，目标是改善美国的医疗体系。