I was going to blog about this a couple of days ago, but the Scientific Activist beat me to it, leading to a heads-up from PZ Myers. Never let it be said that a little thing like that stopped me from putting my two cents in. Besides, I think I bring a certain perspective that hasn’t been addressed thus far about this subject, namely the declining success rates for applications for R01 grants from the NIH. For one thing, I have an R01. I’m about a year and a half into a five year grant, which means that I have about two and a half years or so before I have to go for renewal. Consequently, I can bring the perspective of a first time grant holder. For another thing, I’m also a clinician, which means that I have to devote a significant chunk of my time to the care of patients, which decreases the amount of time I can spend on science compared to basic scientists.
First, I should explain just what an R01 grant is. It’s the gold standard of NIH grants, an unsolicited investigator-initiated grant that can be about whatever the applicant wants it to be about. In other words, it’s not a grant program that calls for research proposals on specific topics. As the NIH explains:
The NIH awards R01 grants to organizations of all types (universities, colleges, small businesses, for-profit, foreign and domestic, faith-based, etc.) and the R01 mechanism allows an investigator to define the scientific focus or objective of the research based on a particular area of interest and competence. Although the Principal Investigator writes the grant application and is responsible for conducting the research, the applicant is the research organization.
The R01 grant mechanism has been an incredibly successful program, because it allows any scientist with an idea and enough preliminary data to apply for funding. Robert Weinberg, at his address at the AACR meeting last spring, emphasized just how much the R01 grant mechanism has contributed to the success of the U.S. biomedical research effort, because traditionally, it has been individual researchers pursuing new ideas that generate the basis for new treatments. If the idea is judged worthy by the reviewers in the study section that covers the topic of the grant, an R01 can provide for up to five years of uninterrupted funding, including salary support and indirect costs to the institution. A typical R01 grant these days funds approximately $125,000 to $250,000 per year in direct costs. It is difficult to overemphasize the importance of this stable source of funding to biomedical scientists. Other grant mechanisms don’t last as long or provide as much money, forcing scientists to reapply every couple of years, rather than every five years. Indeed, the R01 is the “gold standard” grant mechanism for individual investigators. In the basic biomedical sciences, an investigator is unlikely to achieve tenure without having obtained an R01. In fact, to earn an R01 is a key milestone in an academic scientist’s career, a key sign that a young scientist has “arrived,” and renewing that R01 for the first time is a milestone that indicates that the researcher has become established. Consequently, the chances for investigators to obtain a new R01 for a new idea and to keep an R01 once they’ve gotten it are of great concern.
In last week’s Science, H. George Mendel and Elliot Veseli published a letter in which they described the findings of their analysis of the success rates of R01 applications. The results, as anyone who is struggling to get or keep an R01, are not encouraging:
Nick and PZ are dwelling on how much the success rates have dropped so much during the Bush administration. True, there is plenty of blame there, particularly for the last two fiscal years, in which the NIH budget has gone from flat to decreasing in real numbers. However, the story is a bit more complicated than their simple Bush-bashing would lead you to believe, as richly deserving as the Bush administration may be of bashing because of its science policies. Notice that the steepest decline actually began in fiscal year 2002 and accelerated in 2003. I would point out that fiscal years 2002 and 2003 were the last two years of the much vaunted doubling of the NIH budget over five years that began during the Clinton administration and concluded during the Bush administration. In fact, the success rate of new R01 applications actually fell from 20.1% to 15.4% during four years when the NIH budget was climbing by some 15-20% per year and resubmissions fell from 55.1% to 45% during the very five years that the NIH budget was doubling. In fact, it’s very telling that, during the exact years the NIH budget was in the process of doubling (fiscal year 1999 to 2003), the funding devoted to R01 grants only increased between 10-15% at best. One has to wonder where all that extra money went, as most of it clearly didn’t go for funding R01s. Then, when the doubling of the NIH budget ended in 2004 to be replaced by flat or declining budgets, R01 success rates went into a free fall. This is not that surprising, given that, thanks to the flat to declining NIH budget and commitments from grants funded during the period of the budget doubling (which will continue through fiscal year 2008), the 2007 NIH budget will have 11% less purchasing power adjusted for inflation than it did in 2003.
Clearly, though, something else has been going on here, in addition to a flat or declining budget. Nick’s linking the decline to 2001, the year Bush took office, is far too simplistic. What could it be? For one thing, as suggested by this Inside Higher Ed article, the doubling of the NIH budget encouraged universities to build and hire new investigators, leading to more applications and driving the success rate down. It is thus possible that smaller budget increases sustained over a longer period of time could have avoided the boom and bust cycle, the “bust” part of which we are presently suffering through until the grants funded during the doubling period work their way through the system and political pressure builds to increase NIH funding again. Another complaint comes from the aforementioned Robert Weinberg, who, while lamenting the fact that the average age of investigators getting their first R01 has increased from 34 to 42 over the last generation, pointed out that another part of the problem is that the NIH has shifted its priorities to emphasize “big science” projects involving many investigators and collaborators over grants to individual investigators:
As a consequence, American biomedical research is increasingly reverting to models of researchorganization that have held back scientific progress in many other parts of the world. In these models, researchers acquire their scientific independence only when in their 40s and even 50s, long after the peak of their scientific creativity has passed.
The failure to recognize and halt this trend is compounded by another problem. As time goes on, ever larger proportions of NIH funds are diverted to funding research collaboratives of various sizes to the detriment of small, investigator-initiated projects. Perhaps those in power have been influenced by the obvious successes of the Human Genome Sequencing Project and the bounty of useful information that it has yielded. Those who control the scientific purse strings seem to have lost sight of the fact that this undoubted success does not provide a useful template for how most discovery research is conducted. In the case of the National Cancer Institute, this vision of grand projects and their utility has caused this particular Institute to invest large amounts of funds in proteomics, nanotechnology, and a massive software development program that aspires to make the data systems of American research hospitals intercompatible. Implied in the launching of these large-scale projects is the notion that if small scale projects yield relatively small advances, much larger projects will yield proportionately more.
As a consequence of these trends, small-scale discovery research is under siege, yet it is precisely such small-scale science that attracts the best and the brightest of our young people. Many of those who are training for careers in research do not look forward to working as members of large research consortia, in which they will only serve as small cogs in very large wheels.
I agree with most of this, except for one part. Given that I’m in my 40′s right now, I take a bit of offense at the suggestion that the height of my scientific creativity has passed, but maybe I’m deluding myself. In any case, Mandel and Vesell appear to echo Weinberg here, although they do not put it in quite so stark terms:
Particularly surprising and regrettable is the continuing erosion in the allocation for total R01 annual funding of new unamended applications. This decreased from $510 million in FY 2002 for type-1 grants to $351 million in FY 2005 (see table). These dollar figures represent less than 1% of the entire NIH budget. Of similar concern is the 38% decrease in total number of unamended R01 applications awarded during this period for new applicants (type-1), even though submissions increased 24%. Major reductions are also evident in renewal applications for competing ongoing investigations (type-2).
What they left out is that the emphasis of what the NIH is looking to fund has also shifted to emphasize more bioterrorism defense projects, further exacerbating the problem. They did, however, point to a more pernicious effect when paylines decrease to only 9%:
Peer review cannot discriminate among and accurately select only 1 of 11 meritorious applications. FY 2006 data are not yet available, but because the total NIH allocation for that period has been less than the biomedical inflation index, a trend toward further diminished support of R01 applications is evident.
When paylines become this tight, politics, personal biases, and conservatism become the order of the day in the decisions over who gets funded because, while it is relatively easy to distinguish a proposal that is in the top 20-30% from one that is in the top 10%, it is very difficult to differentiate and rank proposals that are in the top 5-10%. They’re all excellent. In such a climate, reviewers become much more conservative and much less likely to fund riskier research projects and more likely to let their personal biases dictate what is funded. After all, each R01 funds between about $750,000 to $1.5 million when you count indirect costs, and reviewers will be reluctant to take a chance when so few projects can be funded and you also want to make sure that science you like gets one of them. The science supported by the NIH will likely become more and more conservative and risk-averse, just as it did in the early 1990′s, the last time pay lines fell this low. Anecdotal reports from colleagues on study sections already suggest that this is already happening.
I can also tell you from personal experience that it’s even worse than that. In an attempt to keep the number of R01s from plummeting even more precipitously than it is, the NIH has been cutting the funding of new grants. For example, the budget of my R01 was cut a cool 22% right off the top before I ever saw a cent. I can barely afford to pay my postdoc and technician now. Worse, their salaries are mandated to increase by 4% a year, but my grant doesn’t a dime. Each year, I have less and less to pay for supplies. I’ve calculated that, if I do not somehow obtain more funding by the fourth year of my grant, I could easily have no money for supplies, putting me in the strange position of having an R01 but being desperately resource-poor, so much so that I would have to dip into the remnants of my laboratory startup funds, which would only get me through maybe another year, or let someone go. On top of that, the NIH has been cutting noncompeting renewals (the second, third, fourth, and fifth years) of R01s. My second year was cut by 3%, and the NIH cut a month from the end of my last year. I have no idea what they will do for my third year. It would be so much easier if I just gave up the lab and became a clinician, but I won’t do that without a fight. My basic science colleagues do not have that option. If they lose their funding, they lose their labs and, if not tenured, their jobs.
So what to do? Throwing more money at the NIH won’t necessarily correct this problem if present NIH priorities remain the same, although increasing the NIH budget so that it at least keeps up with the rate of biomedical inflation would certainly help. In that, the Bush Administrating is being penny wise and pound foolish, saving an insignificant amount of money over the last two years (in the $2.3 trillion federal budget, the total NIH budget is around $28.6 billion, or around 1.2%) at the potential cost of a degradation of our biomedical research enterprise that will be difficult and expensive to reverse. This is exacerbated by the shifting priorities of the NIH away from funding R01 grants. True, the NIH is funding more small grants (R21s, for example) for one or two years that are designed to fund riskier pilot projects, and this is a good thing, but these cannot make up for the decrease in R01s, which provide stable funding. What is more critical is to halt or reverse the decrease in the funds devoted to funding investigator-initiated R01 research grants. Big science can yield dramatic results, as the Human Genome Project has shown, but it is individual projects where riskier ideas are likely to be first explored. I’m not sure I’m as pessimistic as Robert Weinberg, but I will leave you with his warning:
Compounding this is the current deplorable state of funding investigator-iinitiated research: pay lines in which only 10% of submitted grant applications are funded constitute profound disincentives for researchers. Why should a young person invest in the laborious task of writing an NIH grant application that has only a minimal chance of being funded?
These factors, when taken together, have made careers in biomedical research increasingly unattractive for many young people. Imagine the prospects of predoctoral students starting out in their early 20s, who confront a wait of two decades until they can procure their first R01 grant, become scientifically independent, and flex their scientific muscles for the first time. Increasingly, these factors dictate that the best and the brightest are not entering our ranks. As a consequence, those of us who conduct discovery research are confronting the prospect of a lost generation, a wide gap in our ranks, as bright young people look elsewhere to discover their career paths. The marvelous engine of American biomedical research that was constructed during the last half of the 20th century is being taken apart, piece by piece. We will all pay for this destruction for decades to come.
When Dr. Weinberg finished his address the AACR, which served as the basis of his editorial in Cell which I have quoted, he received a prolonged, standing ovation from a packed lecture hall.