Abstract:

In longitudinal studies where outcomes may be truncated by death, standard causal estimands often fail to capture meaningful treatment effects, particularly when survival is affected by treatment. Traditional survivor average causal effects (SACEs), which condition on post-treatment survival, are challenging to interpret and identify without strong assumptions, and their direct extension to longitudinal settings poses additional difficulties. We propose a flexible class of marginal causal estimands that aggregate potential outcomes over time among individuals who would survive under both treatment and control. These estimands support a range of clinically relevant summaries, such as cumulative or last-observed outcomes, and can be tailored using weighting schemes to align with different decision-making goals. We illustrate these ideas through a reanalysis of a prostate cancer clinical trial, highlighting how different estimands may lead to different treatment conclusions.

Bio:

Linbo Wang is Canada Research Chair in Causal Machine Learning, and an associate professor in the Department of Statistical Sciences and the Department of Computer and Mathematical Sciences, University of Toronto. He is also a faculty affiliate at the Vector Institute and holds affiliate positions in the Department of Statistics at the University of Washington and the Department of Computer Science at the University of Toronto. His research focuses on causality and its interaction with statistics and machine learning.