Is a randomized controlled design sufficient for a trial to be valuable?