To publish or not to publish.
The arcane workings of the peer review system were the subject of the First International Congress on Peer Review in Biomedical Publication, sponsored by the American Medical Association, and held May 10-12 in Chicago. Nearly 300 biomedical and scientific editors, reviewers, researchers, clinicians, and journalists from twenty-one countries attended, including the chief editors of the world's leading general biomedical journals, the British Medical journal, The Journal of the Americann Medical Association The Lancet, The New England Journal of Medicine, and Science. Several presentations illustrated that there are no uniform standards by which reviewers review papers and by which editors rely on reviewers to select papers for publication. For example, two reports presented conflicting evidence on whether re-review of manuscripts would alter the decisions made by editors to publish or not to publish specific papers. Another study indicated that neither authors nor academic institutions follow any set of uniform standards for manuscript preparation, submission, and publication. As a result, actions considered taboo by most editors, such as duplicate submission (submitting the same manuscript to more than one journal) and duplicate publication ("salami slicing," or multiple publications reporting various aspects of the same clinical trial) may occur.
Bias in peer review received much attention. While no one denied its existence and effects, debate centered on whether bias was inappropriate in all circumstances and how to minimize the biases that do cause problems. For instance, a reviewer with an obvious conflict of interest may provide the most useful review, while a reviewer free of conflict of interest may not be critical enough of a dubious manuscript. Much discussion also arose over whether double-blind review (blinding both the reviewers to the authors and institutions, and the authors to the reviewers) can reduce bias, can be equally assured for all papers and authors, or has any effect on the final decision to publish a paper.
Whether peer review is designed to catch inconsistency, error, or fraud, and whose responsibility it is to regulate and correct such work fueled considerable debate among the representative "gatekeepers" of the leading biomedical journals, academic institutions, and government agencies. Discussants generally agreed that the purpose of peer review is not to catch fraud but to help improve those papers that are publishable (that is, with a thorough statistical review) and to weed out those papers considered not publishable. However, many attendees agreed that most papers are eventually published somewhere. The final presentation posed two of the most intriguing questions concerning peer review: How many published papers actually achieve the goal of changing the behavior of those who read them? And what happens to that rare, innovative paper that might meet this idealistic goal but is denied publication again and again because editors and reviewers fail to recognize its potential importance? In this light, is the peer review system a giant sieve with gaping holes? And are readers, those who have no interaction whatsoever with peer review, the ones who are really blinded to the process?
The Congress on Peer Review has prompted careful scrutiny of the answers to these and many other questions. More important, the research presented and the future studies stimulated by the conference provide a database, long and varied enough to contribute to the beginning of a new science-the science of editorial peer review.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||scientific peer review|
|Publication:||The Hastings Center Report|
|Date:||Sep 1, 1989|
|Previous Article:||A healing-killing conflict in military research?|
|Next Article:||Italy: abortion revisited.|