INFORMATION ACCESS:
NIH Public Access Policy
Elias A. Zerhouni*
A new National Institutes of Health public access
draft policy is raising a tremendous amount of interest in the
scientific, patient, and publishing communities. I would like to
clarify what the proposed policy is, describe its rationale, and
explain why the NIH thinks this is a reasonable, balanced policy that
will serve all interests.
As recently outlined (1,
2)
the draft policy requests, but does not require or mandate, that
NIH-funded investigators submit electronically to the NIH the final,
peer-reviewed author's copy of their manuscript (unless the publisher
agrees to replace it with the final published copy). The author's copy
will be embargoed from release by NIH for 6 months after the
publisher's date of publication, providing at minimum a 6-month delay
between final peer review and public availability in the National
Library of Medicine's (NLM) PubMed Central (PMC) (3).
Some are concerned that grantees or smaller or not-for-profit
publishers will be harmed. This is why NIH elected to leave the
decision to submit the author's copy to PMC in the hands of the
investigators and their publishers. We believe that this aspect,
combined with the 6-month window, will preserve the critical role of
journals and publishers in peer review, editing, and scientific quality
control.
We believe a stable, permanent archive of peer-reviewed, NIH-funded
research publications will help NIH better meet its mission and will
augment the ability of scientists to exchange information more
effectively. This archive, searchable with modern information
technology tools, will enable NIH to manage more efficiently and to
understand better its research portfolio, to monitor its scientific
productivity, and to help set research priorities. It will also help us
to create an end-to-end, paperless grants-management process.
NLM and its predecessors have been responsible for building and
maintaining medical literature archives for more than 150 years. NLM
pioneered electronic database retrieval in the 1960s. Congress assigned
to the NLM the responsibility to acquire, organize, disseminate, and
preserve biomedical information for the benefit of public health.
However, the proposed policy does not mandate that PMC will be the sole
repository of NIH-funded published research. In fact, NIH welcomes
multiple archiving approaches.
The Internet is used increasingly to search for health-related
information. For example, about 93 million Americans searched for at
least 1 of 16 health topics online within the past year (4).
In a 2003 survey, 58% of Internet users said they brought information
obtained from the Internet to their doctor's office (5).
Now, research information is largely available only to scientists,
clinicians, patients, and educators through personal subscriptions or
at academic and hospital libraries. It is important for NIH to provide
the public access to an electronic archive of the findings resulting
from publicly funded research.
NIH supports the current publishing process by encouraging
publication
of NIH-supported original research in scientific journals. NIH already
provides an estimated $30 million annually in direct costs for page
charges and other publication costs (6).
In addition, NIH provides funds, through indirect costs, to grantee
institutions for library journal subscriptions and electronic site
licenses.
NIH highly values traditional routes of research information
dissemination through peer-reviewed journals. Peer review is a hallmark
of quality and is vital for validating accuracy and interpretation.
Publication in peer-reviewed journals is a major factor in determining
professional standing and in making hiring, promotion, and tenure
decisions. We also value the communities of research created by
scientific organizations and the journals they publish.
NIH support is involved in approximately 65,000 articles per year.
Using 2003 data, NLM estimates that publications resulting annually
from NIH-funded research represent about 10% of the articles in the
nearly 5000 journals indexed by PubMed. NIH-funded articles account for
more than half of the total published articles for only 1% of these
journals (7).
It is unlikely that scientists and libraries would use the proposed
public access policy to gain access to the scientific literature in
lieu of their journal subscriptions, because if they did, they would be
able to access only a fraction of a journal's content. In addition,
there are many other components of journals, such as science news,
industry information, literature reviews, job announcements, and other
products that bring value to the reader. These will not be a part of
the NIH archive.
The proposed NIH policy will not affect the ability of NIH-funded
investigators and journals to copyright works. Investigators may assign
these rights to journals in accordance with current practice, and
copyright holders may continue to enforce their rights as before. A
member of the public viewing or downloading a copyrighted document from
PMC is subject to the same rights and restrictions as when copying an
article from the library. In addition, PMC includes a notice alerting
the public to the rights of copyright holders and will continue to post
this notice as part of any NIH public access system.
Some have expressed concern that archiving NIH-funded manuscripts in
PMC will incur huge costs. In fact, by building on an existing
information technology infrastructure housed at the NLM, the NIH public
access policy can be exceptionally cost-effective. Our estimates of $2
to $4 million per year (7)
reflect incremental costs from creating a Web site for the manuscripts
and for XML tagging them into PubMed Central's archival format.
Implementation of the proposed policy may result in certain
efficiencies for our grantees, the majority of whom currently submit
paper copies of their published articles for end-of-year reporting and
for renewal competition.
NIH has received over 6000 comments on its proposed policy. We plan
to
make these available in a public reading room. Currently, the NIH Web
site (www.nih.gov/about/publicaccess/index.htm)
contains answers to frequently asked questions about the public access
draft policy. Because our proposed policy is timely and flexible, NIH
is confident that it will not only advance science, but also be
advantageous to scientists, to publishers, and to the public.
References and Notes
- "Enhanced public access to NIH research
information" (Notice NOT-OD-04-064, NIH Guide to Grants and Contracts,
3 September 2004).
- Fed. Regist. 69 (180), 56074 (17
September 2004).
- See www.pubmedcentral.nih.gov/.
- Internet Health Resources, Pew Internet and American Life
Project, Washington, DC, 2003; available at www.pewinternet.org/pdfs/PIP_Health_
Report_July_2003.pdf
- Cybercitizen Health 3.0 Survey, Table 10 (Manhattan Research, New
York, 2003); available for purchase at www.manhattanresearch.com.
- The estimated $30 million is a conservative
figure based on amounts spent on page charges and other publication
costs on a sample of R01 grant application budgets, scaled up to
provide an estimate of direct costs paid on all research grants.
- Details are in the supporting online material.
10.1126/science.1106929
Supporting Online Material
www.sciencemag.org/cgi/content/full/306/5703/1895
The
author is director, National Institutes of Health, Bethesda, MD 20892,
USA. ![]()
