Assessing the applicability of trial evidence to a target sample in the presence of heterogeneity of treatment effect.