[论文解读] Risk ratio, odds ratio, risk difference... Which causal measure is easier to generalize?

Bénédicte Colnet, Julie Josse+2 more|arXiv (Cornell University)|Mar 28, 2023

Advanced Causal Inference Techniques被引用 9

一句话总结

本文分析不同因果效应测度（RD、RR、SR、NNT、OR）如何在不同人群间外推/泛化，方法是从非参数结果模型出发，澄清可折叠性与异质性，并展示泛化对结果类型和泛化方法的依赖。

ABSTRACT

There are many measures to report so-called treatment or causal effects: absolute difference, ratio, odds ratio, number needed to treat, and so on. The choice of a measure, e.g. absolute versus relative, is often debated because it leads to different impressions of the benefit or risk of a treatment. Besides, different causal measures may lead to various treatment effect heterogeneity: some input variables may have an influence on some causal measures and no effect at all on others. In addition some measures -- but not all -- have appealing properties such as collapsibility, matching the intuition of a population summary. In this paper, we first review common causal measures and their pros and cons typically brought forward. Doing so, we clarify the notions of collapsibility and treatment effect heterogeneity, unifying existing definitions. Then, we show that for any causal measures there exists a discriminative model such that the conditional average treatment effect (CATE) captures the treatment effect. However, only the risk difference has its CATE and ATE (average treatment effect) disentangled from the baseline, regardless of the outcome type (continuous or binary). As our primary goal is the generalization of causal measures, we show that different sets of covariates are needed to generalize an effect to a target population depending on (i) the causal measure of interest, and (ii) the identification method chosen, that is generalizing either conditional outcome or local effects.

研究动机与目标

• 澄清可折叠性和异质性如何关联到因果测度的可泛化性。
• 在以协变量为中心的视角下统一治疗效应异质性的定义。
• 提出一种非参数生成模型方法以理解每种测度捕捉的内容。
• 表征在何种情况下不同测度更容易泛化以及需要哪些协变量。
• 用模拟和临床示例来说明发现。

提出的方法

• 评审并比较常见的因果测度（RD、RR、SR、OR、NNT）及其性质。
• 定义并将可折叠性与效应的可泛化性（可运输性）联系起来。
• 通过对结果进行非参数建模来颠覆传统的“先测度后推断”的思路，观察每种测度捕捉的内容。
• 分析泛化需要根据目标人群和方法调整不同的协变量集合。
• 使用模拟演示在各种情景下异质性和可泛化性如何表现。

实验结果

研究问题

• RQ1 哪些因果测度在可折叠性方面更强或更弱，以及这如何影响它们向新人群的可泛化性？
• RQ2 从非参数生成功模型出发，如何改变我们对共变量上治疗效应异质性的理解？
• RQ3 对于每种测度和泛化方法，哪些协变量（治疗效应修饰因子 vs 预测因素）是泛化效应所必需的？
• RQ4 结果的性质（连续 vs 二元）以及泛化方法（条件结果的标准化 vs 局部效应）如何影响泛化性？
• RQ5 在试验和运输分析中选择测度时会带来哪些实际影响？

主要发现

• 不同的测度意味着不同的尺度和解释（绝对值与相对值），对人群变动的响应也不同。
• 可折叠性与可泛化性相关：只有可折叠的测度才允许通过重新加权来直接从局部效应得到总体效应。
• 非参数结果模型揭示某些协变量作为治疗效应修饰因子，而另一些影响基线风险；这一区分指导泛化。
• 对于某些测度，泛化效应可能只需针对治疗效应修饰协变量进行调整，而对其他测度则需要调整更广泛的预后协变量。
• 该框架统一了现有的治疗效应异质性定义，并阐明异质性如何与测度选择与目标群体相关。