It can be hard to distract scientists from their lab work, but the US National Institutes of Health (NIH) is hoping a threat to their funding could do the trick. On 16 November, the NIH announced that, starting as soon as this spring, any investigator who receives a research grant from the NIH and publishes the results in a scientific journal must submit an electronic version of the final peer-reviewed manuscript to the government's PubMed Central (PMC) repository to continue receiving federal funds.

“We spent money on the research, so they're telling us what we paid for,” says Neil Thakur, special assistant to the deputy director of extramural research at the NIH and the point person for the agency's public access policy.

The NIH established its public access policy on a voluntary basis in 2005, asking grantees to submit a finished manuscript to PMC. Responding to low compliance rates, the Bethesda, Maryland–based agency then made the rule mandatory in 2008, requiring manuscript submission within a year of publication. Even so, only around 75% of papers stemming from NIH-funded research are submitted to PMC today, according to a 2012 report from the US President's National Science and Technology Council. This suboptimal compliance rate has prompted the NIH to try the new enforcement strategy.

“I expect it to increase our numbers quite a bit,” says Thakur. How quickly those numbers rise, however, will probably depend on how rapidly the 25% of papers that do not already comply with the policy start doing so.

Ultimately, the submission of papers to PMC is a simple step. “It's a clerical task,” notes Peter Suber, director of the Harvard Open Access Project at the Berkman Center for Internet and Society in Cambridge, Massachusetts. Still, according to Suber, there is considerable confusion over who must complete that task because of the four ways in which manuscripts can end up in PMC.

Public-access central: The National Library of Medicine, home to PubMed and PubMed Central. Credit: Sage Ross

Which method applies for any given paper depends on the journal in which it's published: the journal publisher can take care of the submission process, with varying degrees of author involvement, or the authors might have to do everything themselves on the NIH Manuscript Submission website. (Like all Nature Publishing Group journals, Nature Medicine submits applicable papers to PMC at the request of the author.)

According to Thakur, researchers must determine their own publishing journals' policies so they're aware of their obligations although the NIH does provide some basic resources to help: the agency's website has a tool to help researchers find specific journals' policies, as well as a bibliography tool that can tell researchers whether any past articles listed in the government's reference database still have to be submitted to PMC.

Excuses, excuses

Academic institutions have started to alert their staff of the NIH's planned crackdown on noncompliance to the policy. For example, after the November announcement librarians at the University of Maryland–Baltimore sent an email to research staff outlining the stricter enforcement rules scheduled to go into effect. “I would have thought that the reason things weren't deposited was to [intentionally] not adhere to the policy,” says M.J. Tooey, executive director of the university's Health Sciences and Human Services Library. Instead, she was surprised as replies to the email rolled in.

The most often cited excuse: many researchers simply didn't understand the difference between PubMed Central, where the government provides full-text content of the scientific literature, and PubMed, which serves only as a database of references and abstracts. (Both are run by the NIH's National Library of Medicine.) A survey conducted by the University of California–San Francisco Library also found that researchers commonly struggle with the administrative burden of complying with the policy (J. Med. Libr. Assoc. 98, 256–259, 2010).

To simplify the submission process for scientists, some libraries, such as the Taubman Health Sciences Library at the University of Michigan–Ann Arbor, file manuscripts on behalf of their staff. Others, including the University of California–San Diego Biomedical Library, offer classes about the compliance process. Most libraries at large research institutions also now have their own websites that summarize the policy and have librarians on staff who are specially trained in its implementation.

Beyond the central role that libraries can have, Scott Lapinski, a digital resources librarian at the Harvard Medical School's Countway Library in Boston, notes that administration offices that handle extramural research funding will have to take some responsibility to ensure better compliance—for example, by maintaining a list of all of NIH grants and all resulting papers. “It has to be a sustained and ongoing outreach campaign,” Lapinski says.

Whatever their level of preparedness, he notes, schools will have to learn quickly. “If you accept the NIH dollars,” Lapinski says, “this is a term and condition that needs to be met.”