Causally-interpretable random-effects meta-analysis

Recent work has made important contributions to the development of causally-interpretable meta-analysis. These methods transport treatment effects estimated in a collection of randomized trials to a target population of interest. Ideally, estimates targeted toward a specific population are more interpretable and relevant to policy-makers and clinicians. However, between-study heterogeneity that does not arise from differences in the distribution of treatment effect modifiers can create difficulties in synthesizing estimates across trials. We propose a conceptual framework and estimation procedures that attempt to account for such heterogeneity, and develop inferential techniques that aim to capture the accompanying excess variability in causal estimates. This framework also clarifies the kinds of treatment effects that are amenable to the techniques of generalizability and transportability.