Bivariate survival analysis with complete or censored data

使用完整或审查数据进行双变量生存分析

• 批准号: 356635-2008
• 负责人: Beaudoin, David
• 金额: $ 0.87万
• 依托单位: Université Laval
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2008
• 资助国家: 加拿大
• 起止时间: 2008-01-01 至 2009-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=403032
• 关键词: Bivariate; survival; analysis; complete; censored

Measuring the level of dependence between two variables is of great importance in practice. For example, suppose data are collected on patients receiving heart transplantation. It may be of interest to determine if there is a relationship between age at transplantation and the survival time following the operation. Such measures of association are not easy to estimate in presence of incomplete data. The two most common phenomenons causing incomplete data are censoring and truncation. For instance, an observation is said to be "right-censored" if the only information available to us is that its true value is greater than some number c. A variable X is said to be "left-truncated" when only individuals whose value of X is greater than some number c are included in the sample (other individuals are completely missed by the data collection process). There exist several other schemes of incomplete data. I intend to develop methods of estimating the dependence between two variables under such settings.

测量两个变量之间的依赖程度在实践中非常重要。例如，假设收集接受心脏移植的患者的数据。确定移植年龄与术后生存时间之间是否存在关系可能很有意义。在存在不完整数据的情况下，这种关联度量不容易估计。导致数据不完整的两种最常见的现象是审查和截断。例如，如果我们唯一可用的信息是某个观察值的真实值大于某个数字 c，则该观察值被称为“右删失”。当样本中仅包含 X 值大于某个数字 c 的个体（数据收集过程完全遗漏其他个体）时，变量 X 被称为“左截断”。存在其他几种不完整数据方案。我打算开发在这种设置下估计两个变量之间的依赖性的方法。