[Insight-users] IMAGINE... Papers from NIH funded research to be freely available.

[Luis Ibanez]

Science, Vol 305, Issue 5685, 764 , 6 August 2004

http://www.sciencemag.org/cgi/content/full/305/5685/764b

SCIENTIFIC PUBLISHING:
Seeking Advice on 'Open Access,' NIH Gets an Earful
Jocelyn Kaiser


The National Institutes of Health is forging ahead with plans to require
that papers from NIH-funded research be made freely available.


Last week, in a hastily called meeting, NIH director Elias Zerhouni told
journal publishers he is not happy with the "status quo" and is under
pressure from the public to expand access to research results. He got an
earful from scientific societies worried that any mandatory plan will
drive their journals under.

The discussion was sparked by a July report from the House
Appropriations Committee instructing NIH to consider requiring its
grantees to deposit manuscripts in PubMed Central, its full-text
Internet archive, when they are accepted by a journal. PubMed Central
would post them 6 months after the journal published them, or
immediately after publication if the author's NIH grant pays for any
publication charges (Science, 23 July, p. 458).

In response, Zerhouni held an invitation-only meeting on 28 July with 44
participants, many from scientific societies, as well as commercial and
open-access journals. "There really is a strong advocacy for this" from
scientists and universities as well as patients, explains NIH Office of
Science Policy Director Lana Skirboll. Zerhouni also thinks an archive
of NIH-funded research would help the agency manage its grants portfolio.

Many journals already make content freely available within a year or 6
months, but imposing a time limit could doom some journals, participants
warned. Martin Frank, executive director of the American Physiological
Society, noted after the meeting that publishers are already tinkering
with having the author pay publication costs in exchange for immediate
open access, and he argues that a single policy mandated by NIH "doesn't
take into account the broad diversity of publishing." Says Frank: "Let
me do the experiment."

Another concern is that posting manuscripts could be confusing: Would
the PubMed Central version or the published paper be the document of
record? Some publishers suggested that instead, MEDLINE, the NIH
abstracts database, could include links to full-text papers on journals'
sites. Zerhouni, however, said he's concerned that some journal archives
won't remain stable over the long term.

Critics also question whether NIH should divert funds from research to
expand PubMed Central, which now costs $2.5 million a year and contains
papers from about 150 journals. Frank estimates that it would cost $50
million to post full-text articles for all 4500 journals in MEDLINE.

Skirboll says NIH expects to hold at least one more meeting, this time
with patient groups, then post a proposal for comment in the NIH grants
guide, probably by December. Even when the plan is final, it can be
modified if it causes harm, she adds. "Policies are not laws. ...
Anything NIH puts in place, we will evaluate."