Abstract: While randomized clinical trials (RCT) are the gold standard for estimating treatment effects, their results can be misleading if there is treatment heterogeneity. The effect estimated in a trial population may differ from the effect in some different population. In this presentation we combine methodology from machine learning and causal inference to translate the results from a RCT to a target EHR population.
Learning Objective 1: Understand how effects estimated in randomized clinical trials may not be applicable in real-world clinical setting. If there is treatment heterogeneity, and the demographic distribution differs in different settings, the estimated treatment effects may not hold.
Learning Objective 2 (Optional): Learn how to translate a result from a clinical trial to a target population. We will show how to apply the random forests algorithm to a causal inference framework to estimate individual treatment effects.
Benjamin Goldstein (Presenter)
Matthew Phelan, Duke Clinical Research Institute
Neha Pagidipati, Duke University