NIH Web publishing plan raises scientific journals' ire
In the hallowed halls of science, a publishing feud is brewing.
By Patricia Daukantas
GCN Staff
In the hallowed halls of science, a publishing feud is brewing.
The director of the National Institutes of Health has proposed a comprehensive online publication, E-biomed, that has drawn a quick and negative response from the scientific journal community.
Critics say E-biomed would encourage scientists to publish research on the Web instead of in traditional print journals, threatening those publications' existence.
Representatives of three journals recently debated E-biomed with NIH's director, Dr. Harold Varmus, at the National Library of Medicine in Bethesda, Md.
Varmus said E-biomed is roughly based on the electronic print archive of physics research run since 1991 by theoretical physicist Paul Ginsparg at Los Alamos National Laboratory's Web site, www.lanl.gov. Physicists routinely post copies of articles, called preprints, before journal publication.
Through NLM's Web site, at www.nlm.nih.gov, NIH already provides public access to many databases, including PubMed, a search engine for bibliographic listings and abstracts of medical literature. But to download the full text of articles, users must obtain online access to the databases of the journals that originally published the articles. Many journals restrict access to paying subscribers or charge fees for the service.
Varmus characterized the way scientists have traditionally mined paper journals as 'a pretty hit-or-miss operation.' In contrast, he said, E-biomed would be a free, seamless, continuously available tool for accessing not just journal text but also multimedia and large data sets.
Under Varmus' plan, an article would follow one of two tracks before landing in E-biomed. One route would take the submitted manuscript through the peer review process, in which other scientists rigorously and often anonymously examine the research. 'Peer review will be an essential, perhaps the most important, part of the whole system,' Varmus said.
The other route, he said, would be a minimal review process through which scientists could make available research results not published in journals. They might post data from failed clinical trials or experiments that produced indefinite results. E-biomed would inspect the manuscripts only for gross violations of site policy before putting them online.
E-biomed would clearly label the minimally reviewed reports, Varmus said. He suggested that the categories initially should be restricted to discourage abuse and dubious information.
''Although Varmus insisted that E-biomed will not be part of the U.S. government, and although his proposal calls for an advisory board to make policy decisions, he did not address questions of funding or infrastructure.
M. Michele Hogan, executive director of the American Association of Immunology, said E-biomed would put private publishers in the 'untenable position of competing directly against NIH.' It could make NIH the sole supplier of scientific content, she said.
The current peer review process gives NIH's internal reviews of biomedical research some external validation, Hogan said. Under E-biomed, peer review independent of the government would be lost. She said the proposal would discourage competition for good manuscripts, diversity and technological innovation in electronic publishing. She predicted that E-biomed would draw congressional intervention.
Mary Waltham, president of Nature America Inc. of New York, said the E-biomed proposal needs much more study. She said her company, which publishes the interdisciplinary weekly Nature and six monthly life-science journals, adds value by supervising peer review and editing manuscripts.
But Karen Hunter, senior vice president of Elsevier Science of New York, said 8,900 academic journals now provide text on the Web and, by some estimates, get 10 to 100 times more use than paper editions.
Elsevier, which publishes about 150 life-science journals, started making journals available on the Internet in 1991, Hunter said. The company now has 3.5 million users and close to 750,000 articles in its database, she said.
'Just because it's on the Web doesn't mean you can get to it,' Varmus said. Many of the articles referenced in PubMed cannot be accessed in full text without payment or registration, he said.
Varmus' proposal is posted online at www.nih.gov/welcome/director/ebiomed/ebiomed.htm.
NEXT STORY: Hill frees IRS' Prime funding