EDITOR—Is it possible that the evidence based medicine lobby is so busy reviewing the literature that it has lost touch with the rather disorderly world of real medical practice? Certainly Straus and Sackett provide convincing evidence of that in their article telling us how to use research findings in practice.1 Having decided that the most important of several questions that a casualty officer could ask when faced with an alcoholic, confused man with cirrhosis who is bleeding is “Does treatment with somatostatin reduce the risk of death?”, they conclude that the answer is unknown. The correct course is therefore to form a therapeutic alliance with the patient, discuss the potential …