Research and Funding
Inside NIA: A Blog for Researchers

Why support R21 awards?

Posted by Paolo on Sep 11, 2013 - 2:50 pm
All this is very nice theory. In practice, reviewers want to see strong preliminary data for R21 applications as they do for R01 applications - although in principle preliminary data are not required for R21's. There is still much conservativism among reviewers, and SRO's don't seem to be able to correct this attitude. A young colleague I supervised years ago had to do so many experiments to supplement his preliminary data that in the end he had to remove one of the two specific aims from the revised application that was funded. That is, he was requested to do half of the work he proposed before the reviewers were convinced...

Posted by Curious George on Sep 11, 2013 - 6:46 pm
Theoretically, it's a great idea. Practically, it does not work. The reviewers are demanding strong preliminary results. On my applications, I was told that my idea is too innovative, and I should have "strong preliminary" results to back it up. My senior colleagues are telling me not to waste my time with R21s, since they had similar issues. BTW, I have received very similar comments even for R03s. My favorite one was when a reviewer asked for preliminary results.

Posted by Dorry Segev on Sep 13, 2013 - 1:55 pm
Thank you for posting this blog entry and thank you for your support of the R21s. I am frustrated that other institutes have pulled away from the R21s, which I think are a critical part of the NIH pie, and an even more critical one in this funding environment. >>> Having had 3 R21s funded as a junior investigator, and having reviewed R21s on study section, I can attest to the fact that (in whatever experience I have had) these are different entities from the R01 and seen in a different light. The tendency towards only funding super-conservative, incremental work where the investigator has basically done half the project just to show feasibility/etc is much less with the R21s, and I think that's a wonderful thing. The application is only 6 pages, so there isn't even room for much preliminary data at all, which is also a good thing. >>> I'll give a couple personal examples, for what they're worth. >>> Sometimes a project just isn't big enough for an R01, and doesn't "fit" as an aim in an R01, but is exciting and novel and will move the field forward. One of my R21s was exactly this -- a project that we could do in 2 years, that didn't require "R01-level" resources. We were funded on the first submission and each aim resulted in at least one paper in the highest impact journal in our specialty. >>> Sometimes you need evidence that your entire line of thinking for an R01 is sound. You might find this evidence from less expensive epidemiologic studies. For example, a retrospective study might support the hypothesis underlying a proposed RCT or prospective study. So the R21 is not specifically to produce preliminary data (like feasibility for the RCT, etc) but is to defend a hypothesis that, until the R21 research is done, might not be considered "conservative" enough for the bigger, more expensive study to get done. Along these lines, one of my R21s (funded on the first submission) produced the epidemiologic framework that we then used to support an R01 (funded on resubmission). >>> Anyway, those are my thoughts. I've had several great experiences with this mechanism. Actually the only bad experience I ever had was getting a near-perfect score on an R21 submission and then being told that the institute was funding almost no R21s and was not really supporting that mechanism in the future! So do speak with NIH staff before submitting -- although from this blog entry at least we know how NIA feels!

Posted by A different experience on Sep 13, 2013 - 2:32 pm
I recently received an R21 to do what is essentially a feasibility study with some tangible deliverables that can be used by other researchers. There was no preliminary data in my application, but reviewers were enthusiastic about the data that would result from the proposed study. Maybe the need for preliminary data varies by study section, field, and topic. But now you know of at least one proposal funded with no preliminary data.