In his fall 2006 "desk-to-desk letter," Making It Work for Emerging Scientists, National Institutes of Health (NIH) Director Elias Zerhouni emphasizes the importance of getting scientists started on independent research careers while they're still young. This is hardly the first time Zerhouni has sounded this theme. But there was, I think, something new--and potentially important--in his latest remarks. Until recently, Zerhouni has emphasized scientific "stars"--but in this latest desk-to-desk letter, he focused on the lower tier of emerging biomedical scientists, beyond the elite group whose success is inevitable.
"The best and brightest," Zerhouni wrote, "are reaching their full potential in solid research careers. ... For others, the queue is backing up. For them, it is taking longer to complete the Ph.D. The years under postdocs as these individuals seek permanent positions have stretched. We do not want them to become discouraged. We need new-investigator programs aimed at helping people move from training to independent research careers that match the needs of today’s science."
Once these folks have found independent jobs, getting funding for their research is at least as hard. NIH got rid of its R29 "new investigator" grants long ago; because they were small, they were never really that effective anyway. Today, young investigators must utilize the usual R01 machinery to get funded--and that means that NIH has to make sure that the standard review process works well for this second tier of young investigators. I spoke to Antonio Scarpa, director of the Center for Scientific Review (CSR), to see what NIH is doing to make sure they get a fair shake from the review process.
I'll start, however, by defining the problem--as I understand it. According to NIH insiders I've spoken to, proposals written by new investigators generally are inferior to those written by more experienced scientists. This is hardly news; it's hard not to learn something when you've been through the process a few times.
Furthermore, new proposals--including new-investigator applications--are reviewed alongside competitive renewals. Because their authors have had years to refine their language and their scientific approach, competitive renewal proposals tend to be superior to all new proposals--not just proposals from new investigators. So in a typical study section, there tends to be an overlapping, trimodal distribution: At the top of the stack are the competitive renewals, followed by new proposals from experienced investigators. Toward the bottom come proposals from new investigators.
This is the problem NIH faces, in a nutshell: The proposals that are easiest to score well--those with the best preliminary data and the most carefully considered methods and well-crafted scientific objectives--come from the most experienced investigators and, indeed, from long-established projects. Newer ideas tend to be less well conceived, and the youngest, freshest thinkers have the least experience. "Innovation" is one of NIH's five key review criteria, but less experienced grant writers are generally not as good at convincing reviewers that their work is innovative, even if it is. The qualities that make a proposal stand out from the crowd come with age and experience. It's no one's fault--least of all the study section members, who tend to be highly sympathetic to proposals from new investigators.
So what is to be done? Science depends on meticulous, careful work; how can reviewers and panelists be persuaded to favor radical, half-assed (though perhaps brilliant) ideas presented (relatively) poorly? And should they be? I asked Scarpa how he saw the role of CSR in trying to advance the interests of new investigators. Not surprisingly, CSR's focus, Scarpa says, is on implementing existing policy as efficiently and effectively as possible. CSR is working, Scarpa says, "to establish best practices [and] consistency between study sections. In the past, different people were doing the same things in different ways." He cites, for example, standardizing (and shortening) the time NIH takes to post critiques (now 5 days after the meeting for new investigators; 1 month for all others). CSR is also, Scarpa says, standardizing the way "résumés"--those summaries of the discussion of a particular proposal--are written. "Though the process is human--there will still be some inconsistency--there should be more consistency" now, Scarpa says.
Another important goal, Scarpa says, is spreading the word about new investigators and what reviewers and panelists ought to be doing to help them. NIH's leadership has long been committed to funding new investigators, he says. But today the message is getting down into the trenches. New investigators are always indicated by an asterisk, but "just to be sure, the chair will remind people that this is a new investigator" each time a new-investigator proposal is introduced. As awareness grows deep in the organization, Scarpa says, new investigators are likely to get a better deal.
Once they've been made aware that a proposal comes from a new investigator, what are panelists expected to do? You may have heard stories about the "new-investigator discount"--a fraction-of-a-point adjustment in the score for new-investigator proposals in the fundable range--but that's not NIH policy. "There is not a numerical edge factor; that is not the way we operate," Scarpa says. He acknowledges, however, that some reviewers may do this on their own, and that's okay, he says. But CSR's focus is on providing a qualitative--not quantitative--edge. "There is some forgiveness" of the grant-writing sins of new investigators, he says.
I asked about the apparently common practice of "spreading the scores." "People tend to want a grant funded, so they try to imagine what the funding rate" will be, says Scarpa. Then "they tend to compress the score" toward that point. The solution is to encourage reviewers to assign scores over a wider range within a particular mechanism--R01s, say, or R15s--to defeat the tendency to clump the scores near the presumed pay line. "It's not an order but a guideline," Scarpa says.
Does this practice hurt new investigators? Possibly. Because new-investigator proposals tend to be weaker than the other proposals within the mechanism, "spreading the scores" pushes new-investigator scores downward more often than it pulls them up. And the further down a proposal's score is pushed, the harder it is for an institute to fund that proposal, it seems to me. But Scarpa--who should know--insists it doesn't matter; all that matters, he says, is the relative ranking of the proposals, and that is unaffected by the practice of spreading the scores.
In any case, NIH's constituent institutes are free to base their funding decisions on their own scientific and training priorities--a very important fact and one that, in my view, the institutes should take every opportunity to exploit in order to promote new investigators. As Zerhouni noted in his desk-to-desk letter, the National Heart, Lung, and Blood Institute (NHLBI) is already doing that. NHLBI has established a separate pay line for new investigators; new-investigator proposals considered by NHLBI are no longer competing head-to-head with proposals from more experienced investigators. Other institutes should emulate the NHLBI approach.
A full solution to the problem of young investigators, however, is beyond the reach of NIH. To fund new biomedical scientists at a level that meets the future needs of the country requires healthy budgets from Congress. Institutions that train and mentor scientists must also do their part; although no amount of training can make up for a lack of experience, becoming a better grant writer doesn't require a separate Ph.D.--just lots of time, good advice, and meticulous attention to detail. Universities and their constituent departments need to become more involved in helping new investigators submit the best proposals possible. Many new investigators run their proposals by colleagues before submitting them. But, as one panelist I spoke to put it, "Most of the proposals that I have seen from new investigators could really benefit from discussing ideas with experts in the field." There may be some risk that your ideas will be stolen, but as long as you're cautious--don't show your lab's obvious competitors--it's probably worth the risk.
And when your proposal is rejected, read the reviews carefully. Don't get angry or defensive. Call the program officer for a chat.
Meanwhile, there's an opportunity here for early-career scientists willing to do their homework. If you study up on grantsmanship, get an early start, and write (and rewrite) a better-than-average new-investigator proposal, there's an excellent chance that you will take home a prize.
The GrantDoctor is a pseudonym.
Comments, suggestions? Please send your feedback to our editor.