期刊
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
卷 118, 期 542, 页码 846-857出版社
TAYLOR & FRANCIS INC
DOI: 10.1080/01621459.2021.1955691
关键词
Continuous exposures; Longitudinal data; Modified treatment policies; Sequential double robustness; Targeted minimum loss-based estimation
Most causal inference methods focus on fixed values of exposure under interventions, which may not be practical for continuous or multi-valued treatments. Longitudinal modified treatment policies (LMTPs) provide an alternative that yields immediate and practically relevant effects, with an interpretation in terms of meaningful interventions such as changing the exposure by a specific amount. LMTPs also have the advantage of satisfying the positivity assumption required for causal inference.
Most causal inference methods consider counterfactual variables under interventions that set the exposure to a fixed value. With continuous or multi-valued treatments or exposures, such counterfactuals may be of little practical interest because no feasible intervention can be implemented that would bring them about. Longitudinal modified treatment policies (LMTPs) are a recently developed nonparametric alternative that yield effects of immediate practical relevance with an interpretation in terms of meaningful interventions such as reducing or increasing the exposure by a given amount. LMTPs also have the advantage that they can be designed to satisfy the positivity assumption required for causal inference. We present a novel sequential regression formula that identifies the LMTP causal effect, study properties of the LMTP statistical estimand such as the efficient influence function and the efficiency bound, and propose four different estimators. Two of our estimators are efficient, and one is sequentially doubly robust in the sense that it is consistent if, for each time point, either an outcome regression or a treatment mechanism is consistently estimated. We perform numerical studies of the estimators, and present the results of our motivating study on hypoxemia and mortality in intubated Intensive Care Unit (ICU) patients. Software implementing our methods is provided in the form of the open source R package lmtp freely available on GitHub (https://github.com/nt-williams/lmtp) and CRAN.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据