Generalizing results of randomized trials to clinical practice: reliability and cautions

Abstract
Background Well designed randomized controlled trials provide reliable evidence of treatment effects, but there is no consensus on how best to apply these results to clinical practice. The main concerns are that populations enrolled in trials are more selected than those treated in a clinical setting, and whether the treatment effects observed in trials will also be observed in clinical practice.