NHLBI explains their thinking on killing the R21...sort of.

Jul 12 2011 Published by under NIH Budgets and Economics, NIH funding

Well, well, well. Hard on the heels of my interpretation of OER head Sally Rockey's comment about R21s having lower success rates as "Kill the R21", what should appear but a note from NHLBI (thanks to PiT for making me aware).

The R21 grant mechanism is intended to encourage exploratory and developmental research by providing two years of support for the early and conceptual stages of research project development. The NIH has standardized the Exploratory/Developmental Grant (R21) application characteristics, requirements, preparation, and review procedures in order to accommodate investigator-initiated (unsolicited) grant applications.

The R21 grant is meant to facilitate the award of larger regular research project grant (R01) applications by providing investigators with limited funds and time to pursue initial studies to obtain preliminary data for larger, more fully developed research projects. However, the NHLBI has determined that this grant mechanism is not having the effect or the impact on its research grant portfolio in the manner for which it was originally intended.

Consequently, the NHLBI will no longer accept investigator-initiated R21 applications in response to the NIH’s Parent Funding Opportunity Announcement (FOA), NIH Exploratory Developmental Research Grant Program (Parent R21). Investigator-initiated R21 applications will not be accepted for review and funding beyond Fiscal Year 2012. Therefore, the last receipt dates for NHLBI investigator-initiated R21 grant applications are October 16, 2011 for new applications and November 16, 2011 for resubmission (amended) applications.

No additional word as to why they think it is not having the intended effect.

Nor, of course, what they might have considered as alternative strategies to get the allegedly intended effect. 'cause you see where I'm going here, right?

If you have a desired goal, and the thing you thought would accomplish that goal isn't getting the job done, you might want to fix or replace it. Otherwise someone might conclude you aren't really interested in the goal in the first place. Or that you are perhaps no longer interested in a goal that you previously had established.

As you know, Dear Reader, I am of the opinion that there is nothing structurally wrong with the idea of the R21, in fact I think it is a great idea. The problem is that study sections are not, and perhaps never were, able to get their collective heads into the game.

This is not entirely the fault of study section members. If there was one single meta-review issue (having to do with the process of review rather than the actual review of proposals) that ate up the most time in my collective experiences on sections to date it is the intent of the R21 mechanism. Reviewers are capable of altering their behavior and many, if not most, are trying hard to do the job they think they have been set. The trouble is, the NIH is very loathe to address certain (not all, which is what makes it odd) issues. Such as what in the hell the R21 is for. Exploratory? Developmental? High risk/High reward? JuniorMint Starter Grant? Are they serious about the no-preliminary-data thing? how can they be when the "best" grants all have supporting evidence...? Etc.

It is my belief that the NIH should have stepped in a long time ago and fixed the review of R21s. Perhaps by putting them in their own sections. Perhaps by eliminating Preliminary Data. Perhaps by a series of reviewer instructions. Or, maybe, by prioritizing any R01 that arose from a successful R21 interval? Or maybe, just maybe, merely by pointing out all along the way how R21 scores were shaking out, telling study sections they weren't (obviously) doing the intended job of treating it like a different mechanism from the R01 and urging them to do better. It is also my prediction that the NIH could fix a lot of the bad study section behavior just by shining a bright light upon it.

Throwing up their hands and closing out the mechanism is no solution.

Nor is it a solution, I would venture, to replace the R21 with some other mechanism without fixing the root problems. Take note of the long failed history of the R29/FIRST, NI checkbox, ESI focus, etc attempts to help out the junior investigators. It has quite obviously been an issue with the NIH since at least the 80s, going by their attempts at structural fixes. None of that kept study sections from beating up the junior faculty to an unfair degree. Why didn't they try to fix the behavior of study sections first?

It's funny. I was just being regaled by a Program type in a somewhat different context about how great the ARRA experience was. Two years, up and out! Try it out and cut it off if it isn't working! Let's not jump into these 5 year plans that slog along fruitlessly in the second half. I'm paraphrasing but this was the essence. This kind of thinking should shift Program my way....towards favoring even more R21 awards.

Not towards killing the mechanism off.

RIP, R21, RIP.

16 responses so far

  • anon says:

    That NHLBI timeline to the R21 death makes no sense. If someone submits a new application for the next cycle (Oct 16th), they can't resubmit if the score isn't high enough (last resubmit is Nov 16th)? WTF? Of course, a lot of what the NIH does makes no sense.

  • Dr Becca says:


    OK, 2 questions.

    1. Hypothetically speaking, is it even worth sending in an A0 for the upcoming Oct deadline, now that there's no chance for an A1?
    2. Will the money previously designated for R21s now go to R01s? Some other mech? Will this extra money be a change people can actually feel, i.e. raised paylines?

  • whimple says:

    This is not entirely the fault of study section members.

    Yes, it pretty much is. Sort of a, "what part of 'exploratory' do you not understand?" thing. The NCI killed the R21 parent announcement years ago.

  • drugmonkey says:

    Dr. Becca, this will be Institute by Institute so check your relevant ones carefully. As to NHLBI applicants, is it worth it? well, if you have a great idea that is a perfect fit for an R21....maybe? Somebody has to get the last one....

  • I also hear through the grapevine that NCI is planning R15- and R21-only study sections in the next round. Don't know if this is only for one particular IRG or not.

  • Arlenna says:

    NCI has been doing all of their R21s as special RFAs with their own study sections for at least the last 5 or so years as far as I know. The ones I've participated in and sent grants to in the last two years have been run really well, with the exploratory/innovative aspects at the forefront and SRO/Chair intervention whenever preliminary data are pushed. I agree with DM--why are so many institutes weenie-ing out on this, rather than just getting their shiznit together to organize these reviews more effectively? This is a great mechanism for starting something new and higher risk that needs to generate preliminary data before getting a longer-term commitment.

  • It must be the move to an R15-only study section that I heard about then.

  • drugmonkey says:

    NCI is very large. Perhaps the smaller ICs are pleading that there are not enough R21s in close enough topic domains to get proper review. Biasing for scientific specificity over mechanism specificity?

  • Arlenna says:

    Perfect opportunity for IAR meetings! Especially since the discussion record is there for all to see, so any "blah blah too risky need prelim data" will jump out to SROs, chairs and other reviewers.

  • crystaldoc says:

    No not all NCI R21s have been reviewed in study sections dedicated to them. Last cycle I had an R21 reviewed in one of the regular sections that reviews R01s. My colleague that just ad hoc'd on another one of the regular cancer study sections said they reviewed 29 R21s. He said they were reviewed with essentially the same criteria as R01s, and not surprisingly scored terribly by comparison.

  • Dr. O says:

    NIAID reviews [at least some of] their R21s separately as well. From what I've heard it works fairly well for exploratory/developmental projects, but there is still a pretty strong push for at least some bit of preliminary data.

  • Arlenna says:

    Okay I revise my comment to say "many" are reviewed in their own study sections. They don't participate in the generic R21 RFA, so they are all specific RFAs.

  • They just shut off R21 because they didn't had any money. Of course, all of the things that you are saying may be true as well.

  • drugmonkey says:

    R21s are half the price of R01s, broadly speaking. Fewer apps too. They could make their overall R-mech success rates look better by putting more $$ into the R21s and funding fewer R01s....

  • Transitional one says:

    The recent trend is that reviewer looks for a mini RO1 in R21. Even I know few people who were asked for more supporting data for their R21 application. Then why R21 is explained as the 'exploratory/developmental/innovative/high risk' bla bla. If somebody has that much data to prove that his R21 will work, then why don't he apply and RO1? Is not it a good idea to provide more R21 to a new investigator, who will be able to start developing themselves? The more established researcer and the new investigators those who have a more complete story with evidences should get the RO1. Anyway, may be the BIG HEADs are better deciding this.

  • fotografia says:

    Hi, I just found your blog and thought I'd say thank you, and thats because I finally found information that i was looking for.

Leave a Reply