Peer review and the process of publishing of adverse drug event reports.

Most doctors finish their careers without ever having published a scientific paper. The small percent that do mostly end up writing "case reports". But, unless the report describes some truly extraordinary event, very little ever comes of it, and their publication may do more harm than good. The problem is particularly acute in cases of alleged drug toxicity. Case reports are incomplete, uncontrolled, retrospective, lack operational criteria for identifying when an adverse event has actually occurred, and resemble nothing so much as hearsay evidence, a type of evidence that is prohibited in all courts in all of industrialized societies. When a journal, even a highly respected one, decides to publish hearsay, readers are utterly reliant on the integrity of the journal, the author, and the peer review process; such reliance may not always be warranted. Some recent examples of process failure are provided, as well as some suggestions about possible remedies, including the use of pharmaco-vigilance algorithms.