Estimating the health effects of interventions: the target trial approach

Abstract

Results from randomized trials are often considered the gold standard for medical evidence. However, large randomized trials can be costly, infeasible, or untimely. Even when randomized trials are available, trial participants may not be representative of the target population. In this talk, I will present causal inference approaches to estimate the health effects of interventions in the target population. Specifically, I will describe a novel framework to emulate a target trial using observational data, which helps researchers identify and avoid unnecessary biases in observational studies. When randomized trials are available, but the trial participants are not representative of the target population, we provide a framework to extend inferences from trials to populations of substantive interest. Finally, I will describe a common and unappreciated bias that affects generalizability and transportability analyses.

線上視訊請點選連結

 

附件下載