NIH to Overhaul Peer Review of Grants

The U.S. National Institutes of Health (NIH) today released a widely anticipated plan to improve its system for peer reviewing grant proposals. The plan generally follows the recommendations of two advisory committees, including shortening the grant application. But NIH rejected a more radical suggestion aimed at eliminating an apparent bias toward researchers who resubmit their grant applications. Instead, NIH will try other ways to fund the best ideas quickly.

One year ago, NIH Director Elias Zerhouni asked external and internal advisory panels for advice on how to cope with a record number of applications, a flat NIH budget, and a shortage of quality reviewers. The two panels issued recommendations this winter (Science, 29 February, p. 1169). NIH's response was presented today to the Advisory Committee to the Director by National Institute of Dental and Craniofacial Research Director Lawrence Tabak.

NIH agreed with the panels on the need to shorten the application--by more than half, from 25 pages to 12--and to emphasize the anticipated impact of the research over methods and other details. Applicants will also be given more explicit feedback on their proposals. To attract more reviewers, NIH will allow them to serve over 6 years rather than 4 years, test out online reviews to reduce travel (review committees currently meet in person), and give those who attend at least 18 meetings a grant supplement of up to $250,000, about as much as 1 year of an average grant. NIH also plans to have reviewers segregate the applications of young investigators from the rest of the pool and assign a different NIH-wide cutoff point for funding them so that at least 1500 a year are funded.

However, NIH officials nixed the panels' recommendation to jettison a system in which unsuccessful applicants can resubmit their proposal two more times. Reviewers tend to favor these amended applications (A1s and A2s) over first-time awards (A0s) because the applicant responded to reviewers' comments or out of sympathy, the advisory committee found. Since the doubling of NIH's budget ended in 2003, the percentage of first-time applications funded has shrunk from 60% of the total pool to about 30%.

To level the playing field, the advisory committee recommended that all proposals be considered "new." The committee also urged that the weakest proposals be marked "not recommended for resubmission." Its goal was fewer resubmissions and a lighter workload on reviewers.

These two proposals didn't go over well with the community. "There was a huge outcry about this. People feel like they need a second chance, a third chance," Zerhouni told Science. Rather than throw out the current system, the agency plans to "carefully rebalance success rates among" the three types of submissions so as to fund a larger portion of high-scoring grants on the first round, according to Tabak's slides. Tabak said this will be done by an institute's advisory council, which makes the final decisions about which grants to fund.

NIH's selective adoption of the recommendations is a relief to many scientists, says Howard Garrison, public affairs director of the Federation of American Societies for Experimental Biology in Bethesda, Maryland. "We're not comfortable with changing the system radically to reduce the number of resubmissions," he says. Keith Yamamoto, the University of California, San Francisco, cell biologist who co-chaired the external peer-review advisory committee with Tabak, says he's "disappointed" that NIH didn't follow all his group's suggestions. But Yamamoto says "I'm basically happy with" the report. NIH says it will take 18 months to implement the changes.

Related site