Testing for Treatment Effect Twice Using Internal and External Controls in Clinical Trials