News

NIH's plans for online publishing could threaten journals

Tony Delamothe , BMJ

The National Institutes of Health (NIH) may set up a single electronic repository for peer reviewed, biomedical research papers in what would mark the most radical shift in scientific publishing since the first peer reviewed journals appeared in 1665. Such a web based service would free up millions of dollars spent by libraries on journal subscriptions, while leaving the business plans of many scientific publishers in shreds.

It would be a logical extension of the National Library of Medicine's PubMed service, similarly funded by the US taxpayer but freely available worldwide. Whereas PubMed provides electronic access to Medline's nine million citations, the new service would provide the full text of articles as wellall available from one place.

Pat Brown, a genetics researcher at Stanford University and one of the driving forces behind the new proposals, complains that journals currently "just balkanise the literature and then charge a toll for access."

The main inspiration has been the success of the high energy eprint server set up by Paul Ginsparg in 1991. Physicists now routinely submit their completed articles to this fully automated electronic archive, which is freely accessible over the world wide web. About 2500 articles are submitted each month, and the service has supplanted traditional journals as the means of first publication in many areas of physics. Most of these articles are subsequently submitted to traditional peer reviewed journals.

What is different about the NIH's plan is that it will add some form of peer reviewthus making peer reviewed journals redundant. David Lipman, the director of the US National Council for Biotechnology Information, has spoken of trying to develop a "third way"not the traditional journal or an eprint server, "but a completely different model with a different philosophical basis."

Details of what this might be are scarce, and the NIH's ideas for peer review are apparently undergoing rapid evolution. Last week Science suggested that articles might be posted alongside the comments of two peer reviewers. Another recent version had traditional journals retaining their functions as guarantors of quality and stamping their approval on articles they deemed worthy of it.

The idea was first aired publicly by Harold Varmus, the institute's director, last week. He plans to publish a fuller discussion within the month, acknowledging the need for more input and discussion before proceeding. Despite the funds at NIH's disposal Science quotes Varmus as saying that this "doesn't mean a thing if the scientific community doesn't want to play."

The announcement has set off waves of excitement and fear among academic circles and scientific publishers. Many learned societies pay for their activities from their publishing profits.