Reports of randomised control trials should begin and conclude with up-to-date systematic reviews of other relevant trials: a 25-year audit of the quality of trial reports.