Extension of Causal Inference - FIRCA

描述(申请人提供)：这项研究将主要在北京大学的中国和张军妮博士合作完成，作为美国国立卫生研究院#R01HL62567号拨款的延伸。在父母资助中，我们建议为随机激励设计研究或不符合规定的随机临床试验的因果处理分析开发一种通用的统计方法。在不符合规定的临床试验中，我们没有在母方拨款中考虑的另一个问题是“因死亡而截断”。解决这一问题的传统方法忽略了这样一个事实，即截尾后的结果既不是“经审查”的，也不是“缺失”的，而应被视为在扩展的样本空间上定义的。在母方拨款的这一扩展中，我们将讨论一种新的公式来解决这个问题，我们将把结果称为“被死亡截断”，因为截断事件掩盖了结果变量的隐藏值。我们将进一步将框架扩展到可能存在缺失数据的观察性研究。

DESCRIPTION (provided by applicant): This research will be done primarily in Beijing China at Peking University in collaboration with Junni Zhang, Ph.D. as an extension of NIH grant # R01 HL62567. In the parent grant, we propose to develop a general statistical methodology for causal treatment analysis of the randomized encouragement design studies or randomized clinical trials with non-compliance. One additional issue in clinical trials with non-compliance we have not considered in the parent grant is 'truncation by death". Traditional approaches addressing this issue ignore the fact that the outcome after the truncation is neither "censored" nor "missing", but should be treated as being defined on an extended sample space. In this extension of the parent grant, we will discuss a new formulation for tackling this issue, where we call the outcome "truncated by death" because there is no hidden value of the outcome variable masked by the truncating event. We will further extend the framework to observational studies with possibly missing- data.