Generalizing causal inferences from randomized trials: counterfactual and graphical identification

When engagement with a randomized trial is driven by factors that affect the or when trial engagement directly affects the outcome independent of, the average treatment effect among trial participants is unlikely to to a target population. In this paper, we use counterfactual and causal models to examine under what conditions we can generalize inferences from a randomized trial to the target population of-eligible individuals. We offer an interpretation of generalizability using the notion of a hypothetical intervention to "scale-up" trial to the target population. We consider the interpretation of analyses when trial engagement does or does not directly the outcome, highlight connections with censoring in longitudinal, and discuss identification of the distribution of counterfactual via g-formula computation and inverse probability weighting. Last, we how the methods can be extended to address time-varying treatments,-adherence, and censoring.