Statistical Methods for Longitudinal Studies

描述（由申请人提供）：由于测量和数据库技术的不断进步，医学纵向研究面临着新的分析挑战。具体而言，分子测定、医学成像和心理评估的创新产生了许多新的疾病进展的推定标志物。此外，电子数据记录的进步现在允许纵向调查收集测量结果变化和治疗变化的观察数据，其中动态治疗协变量由当前临床指南、患者健康特征或其他因素驱动。本提案的总体目标是开发用于分析现代纵向生物医学数据的统计方法和软件工具。具体的重点领域是：1.重复测量和随时间变化的精度。生物标志物是表征患者健康状态的特定方面的测量。这一目标将开发半参数和非参数统计方法，以估计预后评分或标志物准确预测事件时间的能力，其特征在于敏感性和特异性的时间依赖性测量。2.观测纵向数据和随时间变化的照射。纵向研究现在常规收集患者健康信息和变化的协变量（治疗，暴露）数据。这一目标将开发可用于估计暴露的因果效应的统计方法，并评估标记值指导治疗选择或时机的能力。 公共卫生相关性：在本提案中，我们将开发用于分析纵向数据的新统计方法。特别是，我们将开发可以评估生物标志物的时间依赖性灵敏度和特异性的方法，用于预测未来的事件时间，如疾病发作或死亡。此外，我们将评估和开发用于分析电子病历中通常记录的观察性纵向数据的方法，其中健康状况和治疗措施都随时间变化。我们将重点研究随着时间的推移而改变的治疗的因果效应的估计，以及生物标志物的能力的估计，用于指导选择哪些受试者可能从特定的治疗方案中获得最大的好处。

DESCRIPTION (provided by applicant): Longitudinal studies in medicine are faced with new analysis challenges due to continually advancing measurement and database technologies. Specifically, innovations in molecular assays, medical imaging, and psychological assessment have generated numerous new putative markers of disease progression. Also, advances in electronic data recording now allow longitudinal investigations to collect observational data measuring changes in outcomes and changes in treatments where dynamic treatment covariates are driven by current clinical guidelines, by unfolding patient health characteristics, or other factors. The overall goals of this proposal are to develop statistical methodology and software tools for analyzing modern longitudinal biomedical data. The specific areas of emphasis are: 1. Repeated measures and time-dependent accuracy. Biomarkers are measurements that characterize specific aspects of patient health status. This aim will develop semi-parametric and non-parametric statistical methods to estimate the ability of prognostic scores or markers to accurately predict event times as characterized by time-dependent measures of sensitivity and specificity. 2. Observational longitudinal data and time-dependent exposure. Longitudinal studies now routinely collect both patient health information and changing covariate (treatment, exposure) data. This aim will develop statistical methods that can be used to estimate causal effects of exposure, and to evaluate the ability of marker values to guide the choice or timing of treatment. PUBLIC HEALTH RELEVANCE: In this proposal, we will develop new statistical methods for the analysis of longitudinal data. In particular, we will develop methods that can evaluate the time-dependent sensitivity and specificity of a biomarker for the prediction of future event times such as disease onset or death. In addition, we will evaluate and develop methods for the analysis of observational longitudinal data commonly recorded in electronic medical records where both measures of health status and measures of treatment change over time. We will focus research on estimation of the causal effect of treatments that are modified over time, and on the estimation of the ability of biomarkers to be used to guide the choice of which subjects are likely to obtain the largest benefit from specific treatment options.