Recognizing controlled trials