Statistical methods for clinical research

临床研究的统计方法

• 批准号: 250051-2011
• 负责人: Sinha, Sanjoy
• 金额: $ 1.17万
• 依托单位: Carleton University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2015
• 资助国家: 加拿大
• 起止时间: 2015-01-01 至 2016-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=570808
• 关键词: Statistical; methods; clinical; research

The objective of the proposed research program is the development of novel statistical methods for analyzing longitudinal or clustered correlated data commonly encountered in clinical experiments. An important feature of longitudinal data is that the repeated responses from an individual are often correlated. For example, in AIDS clinical trials, patients may be measured repeatedly at a number of assessment times, and the longitudinal responses from the same patient may follow a certain correlation structure. Furthermore, if genetic data are collected from siblings in a family, then the data may also be correlated by the fact that the siblings carry their parental genes as well as live in a similar environment. The goal of this work is to develop and explore suitable methods to describe the correlation structure and to investigate the association of a response variable (for example, disease status) with some other explanatory variables (for example, candidate genes, treatments, etc.). The analysis of longitudinal data is often complicated by the fact that the outcome variable is not always observed at all assessment times. This missingness often depends on the unobserved value of the outcome at that time. This work will examine various missing data patterns to better understand the significance of a missingness mechanism when analyzing incomplete longitudinal data by incorporating the dependent structures inherent in the data. The proposed research will contribute significantly to the advancement of biostatistical analysis in the areas of clinical study, population genetics, and genomic and proteomic experiments.

拟议的研究计划的目的是开发新的统计方法，用于分析临床实验中常见的纵向或聚类相关数据。纵向数据的一个重要特征是，来自个体的重复响应通常是相关的。例如，在艾滋病临床试验中，可能会在多个评估时间重复测量患者，并且来自同一患者的纵向响应可能遵循某种相关结构。此外，如果遗传数据是从一个家庭的兄弟姐妹中收集的，那么这些数据也可能与兄弟姐妹携带父母基因以及生活在类似环境中的事实相关。这项工作的目标是开发和探索合适的方法来描述相关性结构，并调查响应变量（例如，疾病状态）与其他解释变量（例如，候选基因，治疗等）的关联。 纵向数据的分析往往是复杂的事实，结果变量并不总是在所有评估时间观察。这种缺失往往取决于当时未观察到的结果价值。这项工作将检查各种缺失的数据模式，以更好地理解缺失机制的意义时，分析不完整的纵向数据，将数据中固有的依赖结构。 拟议的研究将大大有助于在临床研究，群体遗传学，基因组学和蛋白质组学实验领域的生物统计分析的进步。