Going beyond randomised controlled trials to assess treatment effect heterogeneity across target populations