Testing for treatment effect twice using internal and external controls in clinical trials | II-Commons