Michael Rawlins is right. In his Harveian Oration at the Royal College of Physicians last month he clearly signalled the medical profession’s overuse of randomised controlled trials to guide clinical decision making.1 Randomised controlled trials have had a long and influential relationship with the medical profession. Milestones in this relationship include James Lind and his controlled trial of lime juice for scurvy in 1747,2 and Bradford Hill’s trial of streptomycin for pulmonary tuberculosis in 19483; both trials championed this experimental approach to answering clinical questions. Perhaps Michael Rawlins will be similarly influential, but in the opposite direction—by emphasising the dangers of relying too much on randomised controlled trials almost to the exclusion of other types of useful information.

Healthcare information, or evidence, varies in its source, quality, and usefulness. Information derived from research may be …