observational

Implications of Interactions in Treatment Comparisons

This article explains how the generalizability of randomized trial findings depends primarily on whether and how patient characteristics modify (interact with) the treatment effect. For an observational study this will be related to overlap in the propensity to receive treatment.