Wednesday, June 30, 1999

British Journal and Stanford Plan an 'E-Print' Server for Biomedical Research By GOLDIE BLUMENSTYK While the National Institutes of Health continues to debate the idea of establishing a World-Wide Web site for biomedical research, a British medical journal and the Stanford University Libraries say they will open a similar site this fall. "It makes sense for us to work to find the eprint's right place in the new digital environment," say editors of the BMJ. The site will be a searchable electronic archive devoted to clinical medicine and health research. Authors will be able to post articles that have not been accepted for publication or subjected to peer review. Officials of the journal, the BMJ, say the site is designed to break the hammerlock that journal publishers have on the dissemination of scholarly information. And by allowing authors to publish electronic versions of their journal preprints -- articles that have not been formally published -- the new "e-print" server could also help get important new research into the public domain more quickly, the editors say. Right now, an article can take months to get into print -- even longer "if it is bumping its way through the hierarchy of journals," says Tony Delamothe, deputy editor of the BMJ. The journal is published by the British Medical Association. The free archive, similar in approach to an e-print server that has operated in the field of high-energy physics since 1991, is scheduled to go into operation in September or October. The site will not be designed to accommodate on-line peer review, but will include features allowing readers to record their comments about posted articles. Backers of the site say they hope it will help reduce the cost of publicizing biomedical research, as well as overcome the supposed "positive bias" in scientific publishing that makes it harder for papers to be published if they show that a technique or treatment had no effect. "The eprint looks like the first substantially new form of scientific communication since the peer-reviewed article," says Mr. Delamothe in an editorial he wrote with Richard Smith, the journal's editor. "And as we're in the business of transmitting scientific information, it makes sense for us to work to find the eprint's right place in the new digital environment." Although it is intended as an experiment, the BMJ editors are hoping the e-print server becomes a vehicle for changing practices in scholarly communication. "In future, authors who go public on their data without posting it on an open forum where others can append their comments will raise questions about their credibility and motives," they contend in their editorial. The e-print venture is being announced as scientists and publishers continue to debate the merits of a new Web site for biomedical research proposed by the director of the N.I.H., Harold E. Varmus. That site would be called E-biomed. Intended to help researchers locate medical information more easily and inexpensively, E-biomed would post published articles as well as other "legitimate" research that had not been accepted for publication or undergone formal peer review, as long as it had the okay of two approved reviewers. "Varmus's proposal has produced a cacophony of complaints from publishers," the BMJ editors note in their editorial. "Our reaction is that it is not radical enough to produce a long term solution yet may be too radical to achieve implementation." To avoid confusion with regular, peer-reviewed BMJ publications, the new e-print server's Web site will be separate from that of the BMJ Publishing Group. Articles -- and the site itself -- will carry disclaimers to warn readers that the material is intended mainly for the benefit of researchers and that "casual readers should not act on its findings and be wary of reporting them." Organizers have not yet chosen a name for the site but say that one of the names they've registered -- dontquote.com -- is a leading contender. "It makes the point rather beautifully," says Mr. Smith. Cheekiness aside, the editors and their Stanford partners say they recognize the nettlesome issues involved in creating a health-oriented e-print server that does not involve peer review. The only vetting the articles will receive will be a cursory review by BMJ personnel checking for such offenses as libel, obscenity, breach of patient confidentiality, and "blatant self-advertisement." Mr. Smith and Mr. Delamothe say they think that will be adequate. The site is aimed more at medical researchers than doctors, they note, and will carry adequate warnings. They also dispute the suggestion that patients might be endangered by the posting of articles that have not been carefully peer-reviewed. Already, unreviewed research findings make their way into the public arena "via authors and conferences -- and press conferences," they note in the editorial. "The appearance of a full account on an eprint server would be far preferable to what happens now, when patients and doctors may have read a garbled account in a newspaper and have no way of accessing a full study that can be appraised." But not everyone believes that the safeguards for the proposed e-print site are adequate. Jerome P. Kassirer, editor of The New England Journal of Medicine, says the proposed warnings remind him of the warnings on cigarette packages -- "which people ignore," he says. Dr. Kassirer also questions whether the information on the site would be truly useful. Unless the research has undergone some careful peer review, "the validity of the information is questionable," he says. He also worries that the site might become a haven for articles that have been rejected by established journals because of defects in the authors' methods or interpretation. "I am not convinced that putting information on a Web site that is incomplete or unpublished would have any benefit at all," he says. A recent editorial by the journal's editor emeritus, Arnold S. Relman, criticized the E-biomed proposal on the same grounds. The New England Journal does not publish articles that have appeared elsewhere, and that policy would extend to work posted on the new e-print server, Dr. Kassirer says. A number of other journals have similar policies, which might discourage authors from submitting their articles to the e-print server. But Stanford officials predict that those policies will change as more and more scholarly publishing moves onto the Internet. HighWire Press, the electronic-publishing arm of the Stanford libraries, recently surveyed officials of the 130 journals it works with and found that the majority of the 53 respondents were rethinking rules that had until now prohibited their publishing articles that had appeared in some form on the Internet. The new e-print server will include information about the publishing policies of journals, in hopes of assuring that authors understand the consequences of posting to it. "You don't want people to make a mistake and suddenly find themselves cut out of a journal," says John Sack, the director of HighWire. Michael Keller, the director of the Stanford libraries, says he expects that even The New England Journal might soon bend. "As soon as they see themselves losing articles they want to have, they'll change their position," he says. Meanwhile, Stanford is pursuing plans to create additional e-print servers in other fields, such as cellular and molecular biology. Mr. Keller says he is talking with major journal publishers about such an offering now, and hopes to announce it as soon as at least one of them commits to the project. Background stories from The Chronicle: Medical Journal Criticizes NIH Proposal for On-Line Scholarly Publication (6/7/99) NIH May Use the Internet to Distribute Research Findings (4/30/99) Journals Differ on Whether to Publish Articles That Have Appeared on the Web (7/17/98)