Diagnosing diagnoses. Receiver operating characteristic methods and psychiatry.

With every issue of theArchives, the reader is inundated with psychiatric research: new treatments, new biological abnormalities, new diagnostic methods, new diagnoses—the list seems endless. It is difficult to conceive of anyone, except, of course, our dedicated Editor, who bothers to read every article in every issue. Indeed, it would be a rare individual who had the time to read every abstract. In the current issue of theArchives, Mossman and Somoza 1 reexamine the dexamethasone suppression test (DST) using a statistical method that is new to psychiatry, namely, receiver operating characteristic (ROC). Using this method, they are able to demonstrate that there is significant variability in the performance of the DST among different psychiatric centers. As with any statistical method, ROC involves more than a little mathematics, and the natural question as one leafs through the See also p 653 .Archivesis "Why should I bother with this?"