How do clinical researchers’ and patients’ preferences influence study hypotheses and reported outcome results for clinical randomised controlled trials? A critical appraisal