MANY scientists and journal editors espouse the principle that a piece of scientific work or criticism is supposed to be judged solely on its scientific merit. In recent years, however, many biomedical journals, JAMA among them,1 have adopted new policies that require authors to disclose potential conflicts of interest. These disclosures are intended to provide information that is supposed to aid editors and readers in their evaluations of the work: in the words of the former editor of The New England Journal of Medicine, such disclosure "alerts readers to information they may wish to have as they assess the published report."2 Examples of scientific misconduct ranging from sloppiness to fraud have been widely publicized; disclosure policies are supposed to reduce these problems by allowing readers to make a more informed and therefore a better interpretation of published work.
Better interpretation of published work is a lofty goal that