ABSTRACT
After history and physical examination, the clinician is often left with a short differential diagnosis that must be further narrowed before appropriate treatment can begin. This almost always means that additional radiographic and/or laboratory tests must be performed in order to confirm or discard a suspected diagnosis. In fact, even the questions asked of patients during the history and physical exam are all informal diagnostic tests in themselves. How likely is the patient to have a mass if you don’t feel one? How likely is the patient with recurrent biliary colic and normal transaminases to have a common bile duct stone, and should they have an ERCP before their cholecystectomy? How confident can you be that vour patient with a normal barium enema does not have polyps or cancer? What if they had had a negative sigmoidoscopy as well? The interpretation of the literature studying these tests and the translation of that information into a form that is easily applicable to a clinical scenario is thus critical. In this review, we will introduce several statistical concepts relating to diagnostic tests and provide examples demonstrating how these can be used in day to day practice.
