Medical Examiner

Open Access

Should scientific articles be available online and free to the public?

If you’re like most Americans, chances are at some point you’ve searched online for medical or health-related information. In fact, of Americans with Internet access, more have used the Web to answer health-related queries (66 percent) than to check sports scores (43 percent), hunt for a job (42 percent), or find a date (9 percent), according to the Pew Internet and American Life Project. There is clearly no shortage of medical and pseudomedical information available online. But some physicians and members of Congress have pointed out that much of the best literature, published in prestigious, peer-reviewed journals such as the Journal of the American Medical Association and the New England Journal of Medicine, is not immediately available for free to the public—and they have begun to argue that it should be.

To the consternation of journal publishers—many of them not-for-profit associations that rely heavily on journal subscription fees for their revenue—the National Institutes of Health has thrown its considerable weight behind the notion of free access to biomedical research. In September, NIH Director Elias Zerhouni proposed that all NIH-funded scientists provide full-text copies of their research papers to PubMed Central, an online government archive, once these papers have been accepted (and thus vetted) by peer-reviewed journals; PubMed Central would then post the work, citing both the author and the journal where it first appeared, six months after publication. (The original plan called for immediate access but was mitigated in response to publishers’ outcry.) The initiative is expected to cost roughly $2 million to $4 million per year, a small fraction of the NIH’s total budget. As the NIH considers final guidelines, a vitriolic debate has erupted in the scientific community: How best to balance the needs of journal publishers against those of scientists, students, and members of the general public who would benefit from unfettered access?

Increased access is a worthwhile enterprise. Highly accessible, electronic archives are simply too valuable to scientists—and, by extension, to the public who may eventually benefit from their work—not to establish and support as a matter of public policy. Scientists at large research universities have access to many journals, but increasingly there are gaps in library collections, and not all publications are easily searchable. The problem is of course even worse outside of major institutions and in countries other than the United States. A more robust public database would go a long way toward speeding and enhancing scientific work. (There is, in fact, precedent for this approach to biological data. Click here to read more.)

The public would also benefit directly from greater access. True, not every hair-splitting research paper in, say, cell biology or biochemistry will be of interest to the populace at large. But for patients who have rare diseases or are considering controversial therapies, access to cutting-edge medical research may be vital—especially since new work is accruing so quickly that even the most assiduous general practitioners may have a hard time keeping abreast of it all.

In addition, it doesn’t make sense for journals to charge fees for access to primary research articles when that research is supported largely by public funding. Academic scientists, some backed by hundreds of thousands of dollars of public moneys, provide their findings and articles to most journals for free. Peer-reviewers, who perform the crucial quality-control work, also offer comments on research papers for no charge. In other words, journals receive what are in effect large subsidies from the rest of the scientific community, which in turn receives significant funding from taxpayers, through organizations such as the NIH. As a matter of basic fairness, members of the public who wish to read about this research should be able to do so without paying further charges.

The movement for free online access to science research dates back as far as 1991, before most of us had even heard of the Internet. (Click here to read more.) In recent years, the project has gained considerable momentum. In 2000, Harold Varmus, then director the NIH, established PubMed Central; later that year, Varmus, along with biologists Patrick Brown and Michael Eisen, also founded the Public Library of Science, or PLoS, an advocacy organization dedicated to promoting open access.

In an early effort to sway existing journals, Varmus and other key players tried to convince scientists not to publish in—or even to subscribe to—journals that did not provide articles to PubMed Central; over 30,000 scientists signed a letter in which they made this pledge. Not surprisingly, few adhered to it: With so much of a scientist’s career resting on his or her publication in prestigious forums, few could afford to displease, let alone categorically abandon, the highly respected traditional journals. So, in an attempt to establish a competing venue—one that would attract the best scientists and demonstrate that open access could be compatible with very high quality—PLoS members decided to launch their own peer-reviewed journals. PLoS Biology appeared in October 2003, followed by PLoS Medicine in October 2004.

PLoS Biology and PLoS Medicine exemplify the philosophy of open access and thus differ from traditional forums such as JAMA, NEJM, and Science in several ways: 1) In addition to providing its content for immediate unrestricted use, PLoS allows authors to retain their copyrights—most magazines control copyrights and profit from any reprints—but requires them to sign a broad licensing agreement that allows their work to be reproduced as long as they are given proper credit through citation. (Scientists do not receive payments or royalties from article publications in either old-style or open-access journals.) 2) PLoS provides all content to PubMed Central. 3) PLoS relies on an alternate business model—instead of earning money through subscriptions and online access fees (a “subscriber-pays” model), PLoS charges authors or their funding agencies a publication fee of $1,500 (an “author-pays” model). Varmus argues that such payments can be folded into laboratory research budgets, and several major funding institutions, such as the Howard Hughes Medical Institute, have agreed to supplement publication costs for their researchers. PLoS has also promised to waive fees for authors who are unable to pay. In theory at least, “author-pays” seems a viable economic model, according to a recent analysis by the Wellcome Trust.

How have the PloS publications fared thus far? It’s still hard to say, since the PLoS journals have had so little time to establish a track record of important or highly cited articles. Some researchers may doubt that work offered to the public for free will command the same respect as content people pay for. But if any journal is well-positioned for success, the PLoS vehicles are: Both have received plenty of media buzz, and Nobel Prize winners and other rock star scientists populate the PLoS board of directors.

There are a few other publishers and journals that have adopted some elements of open access. But most traditional journals have resisted calls to post articles in PubMed Central, even with delays of six months. Only around 160 journals currently offer full-text articles through PubMed Central, out of the roughly 4,000 that are indexed annually by the National Library of Medicine. It would thus be virtually impossible for a scientist to build a first-rate career by publishing entirely in these forums, however sympathetic he or she might be to the goals of open access. (Some of the biggest names, such as JAMA, NEJM, and Science, have begun to make some content available for free on their own Web sites.)

If the NIH decides to implement its proposed policy—and an announcement is expected sometime this month—the open-access movement will receive a crucial shot in the arm. Not only will many more articles be available through PubMed Central, but the call by advocates for faster free access (without the six-month delay) will likely grow louder. Some journal publishers will no doubt rethink their business strategies, in order to compensate for actual or expected revenue losses.

Of course, it is hard not to sympathize with traditional journal publishers, who for years have been careful stewards of the scientific literature, editing articles and selecting scientists to participate in peer review. But the change would not necessarily mean their demise. Many scientists (and science writers) would likely still subscribe to publications like, say, Science and Nature, largely for the valuable extras they offer, such as news blurbs and research summaries published each week. Journals could continue to charge fees for these features, even while adopting open access policies for new research.

This is a story, then, about adapting to the digital world: How should we harness its enormous possibilities while trying to reckon with vested interests that stand to lose in the transition? Ultimately, the potential benefits of open access, for both scientists and the public, are too significant not to pursue aggressively. At a time when technical material is accumulating more quickly than ever before, extensive online databases can help scientists do the most informed work possible. At a time, too, when patients are asked to participate much more actively in health care decision-making, better access to information is crucial. The NIH’s proposed compromise, in which articles are posted to PubMed Central six months after publication, is a reasonable one. But it is only a first step toward a larger goal—namely, true open access, immediately upon publication, for as much scientific work as possible.