Researchers who receive grant money from the National Institutes of Health will be "asked" to submit their results to a public Web site within a year after they are published in a scientific journal, under a new and controversial NIH policy announced yesterday.
The highly anticipated "public access" policy -- which aims to make it easier for Americans to see the results of research they paid for with their tax dollars -- represents a compromise between competing forces that had lobbied the agency intensely during the past year.
On one side were the publishers of highly profitable scientific journals who feared that free access -- even months after paper publication -- would undermine their subscription base. They were joined by some not-for-profit scientific societies that count on revenue from their print journals to support their research and training programs.
On the other side were patient advocacy groups and others who argued that taxpayers should not have to pay subscription or per-article fees to see the results of federally supported medical research.
They argue that journals would not be significantly harmed because the policy applies to only the 10 percent or so of published biomedical articles that result from NIH-funded research. Individuals and libraries would still subscribe, they say, to read the other 90 percent of the contents.
Both sides expressed irritation with the decision, which NIH Director Elias A. Zerhouni said at a news briefing yesterday would take effect May 2.
Proponents of quick access complained that the policy marks a retreat from an earlier version, floated by NIH in September, which had called for public access within six months. Even the 12-month deadline is voluntary, they noted.
The policy "falls short of the bright light of transparency that Dr. Zerhouni promised," said Rick Johnson, director of the Scholarly Publishing and Academic Resources Coalition and a member of the Alliance for Taxpayer Access, a coalition of groups that support changes in scientific publishing.
Former NIH director Harold Varmus, now president of Memorial Sloan-Kettering Cancer Center in New York and a longtime proponent of public access, called the new policy "a significant move" but echoed others' disappointment that it does not use stronger language. Instead of requesting that scientists submit their results, the policy could have said scientists are "expected" to do so, he said.
But the publishing industry's campaign to oppose NIH's efforts -- spearheaded by former House member Patricia Schroeder (D-Colo.), who is now president of the Association of American Publishers -- also fell short of its goals. The AAP yesterday said it is "concerned" about the outcome. Not-for-profits also complained.
"It's wasteful and duplicative of what we're already doing," said Martin Frank, chairman of the DC Principles Coalition for Free Access to Science and executive director of the American Physiological Society, one of several not-for-profit science organizations that Frank said already make their articles available to the public relatively quickly after publication. "The $2.5 million to $4 million that the NIH is going to spend on this could be better spent on biomedical research," he said.
Zerhouni has said there are advantages to having the articles all in one federally managed database, including easier cross-comparisons of data among different articles.
Both sides had at least one complaint in common: The policy leaves it up to scientists to decide when to make their articles public. That puts scientists in an awkward position of wanting to release them quickly to please the NIH -- their funding source -- and slowly to please their paper publishers -- upon whom they are equally dependent for professional prestige.
"For many authors, the dilemma will be painful and career-jeopardizing," said Johnson, noting that life would have been much easier for scientists had the NIH simply demanded public access as a condition of receiving grant money.
Details of the new policy can be seen at: www.nih.gov/about/public access/publicaccess_imp.pdf