Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 244 - Recent Advances in Causal Inference with Applications for the Public Good
Type: Invited
Date/Time: Tuesday, August 4, 2020 : 1:00 PM to 2:50 PM
Sponsor: Section on Statistics in Epidemiology
Abstract #308036
Title: Assessing Surrogate Paradox Risk in the Meta-Analytic Causal Association Framework
Author(s): Michael R Elliott and Jeremy Taylor and Yun Li* and Anna Conlon and Nico Kaciroti
Companies: University of Michigan and University of Michigan and University of Pennsylvania Perelman School of Medicine & The Children’s Hospital of Philadelphia and University of Michigan and University of Michigan
Keywords: causal inference ; surrogate marker; meta-analysis; clinical trial
Abstract:

Because of the time and expense required to obtain clinical outcomes of interest, clinical trials often focus on the effects of treatment on more easily obtained “surrogate markers”. Definitions of surrogate markers have been developed in the causal inference framework, in either the “causal effect” or “causal association” settings. In the causal association setting, high-quality surrogate markers have large treatment effects on the outcome when there are large treatment effects on the marker, and vice-versa. A particularly important feature of a surrogate marker is that the direction of a treatment effect be the same for both the marker and the outcome. Settings in which the marker and outcome are positively associated but the marker and outcome have treatment effects in the opposite direction have been referred to as “surrogate paradoxes”. We propose assessing the risk of the surrogate paradox using the meta-analytic causal association framework, by estimating probability that a new treatment will yield treatment effects in different directions between the marker and the outcome, either overall or as a function of the size of a beneficial effect of the treatment on the marker.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2020 program