Fucken Dumshitte Redneck Imbeciles Are The Stupidest Motherfuckers On Earth »« Loonie Pit Bull Enthusiasts Indistinguishable From AR15-Loving Gun Nuts

Revising And Resubmitting NIH Grants In The Only-One-Resubmission-Allowed Regime

We have been living in the “no-A2″ regime for a number of years now, where NIH grant applicants are only allowed a single opportunity to respond to the reviews of their initial grant submissions with a revised version. If that single revised version does not get funded, it is impermissible to revise it by responding to the reviews and resubmitting a second revised version. At that point you can only write a “new” grant, which has to be substantially different from the unfunded one.

I have had a lot of experience in this regime dealing with my own grants as well consulting on those of colleagues and reviewing them in study section. Based on this experience I have arrived at the following two rules that should only in exceedingly rare circumstances be violated:

(1) If the signifiance criterion scores from the assigned reviewers on the initial grant submission were the *best* scores, and you were hammered on approach, then you have a chance, and it is worth revising and resubmitting. If, however, the significance scores were the *worst* scores, and this was driven by fundamental concerns about the impact of the planned research, then revision and resubmission is unwise. The better course is to rethink the presentation of the overall thrust of the planned research and submit as a new application. This is because it is generally impossible to convince reviewers that they were flat-out wrong about their conclusions that your proposed research is fundamentally uninteresting.

(2) Regardless of what is driving the scoring, if in order to successfully respond to reviewer concerns, you find yourself completely reorienting the thrust of two or more of the specific aims–for example, by discarding experimental approaches wholesale and employing different ones, or by replacing entire hypotheses with other ones–you should submit as a new grant. This is because you don’t want to waste an opportunity to revise and resubmit multiple brand new aims at least once with an explicit response to reviewer concerns regarding those particular specific aims. If you submit these brand new aims in a resubmission, then you lose that opportunity.

(3) You need to get detailed case-specific advice about this shitte from people who have a lot of experience with the new no-A2 regime, not from olde geezers who have been renewing the same single R01 for decades.

Comments

  1. Grumble says

    “This is because you don’t want to waste an opportunity to revise and resubmit multiple brand new aims at least once with an explicit response to reviewer concerns regarding those particular specific aims.”

    Makes sense. Except that despite the fact that the no-A2 rule was supposed to eliminate the “wait in line” attitude of the study section, I get the feeling it’s still there, at least a little bit: A0s get funded more rarely than A1s. On top of that, if you revise based on what the reviewers say, and think you’ve done a pretty good job of addressing their concerns, you want the same reviewers to review the revision. Of course that’s never guaranteed, but at least there’s a greater chance that at least one of the previous reviewers will also be reviewing the revision.

  2. Yoshimi says

    Great advice! I’m a new investigator and recently decided to resubmit a poorly scored A0 w/ the highest scores in significance. In my case it worked out — the grant moved from barely scored to 7th percentile. I did have to change the aims significantly (e.g., we published the original aim 1, so I added a whole new aim 3) and was torn about resubmission vs. a new application for exactly the reason you mention. I ended up risking resubmission for Grumble’s reasons, and my summary statements seem to provide support for the choice. Clearly a risky move though — a fact that grew clearer and clearer as I waited for my score and planned my next steps if it did poorly!

  3. says

    There is also merit to the perspective that if one is triaged not to go for the A1. The idea being that it is too deep of a hole to dig yourself out of, and that you can count on the fact that it will be new to the study section when if it does get a hearing. Given that there are other A1’s ahead of you in the queue, that its worthwhile reworking in a significant way for a new submission.

  4. DrugMonkey says

    I really want to hope you are wrong Potnia but I can’t recall having gone from triage to fundable in one round before.

  5. DrugMonkey says

    Except that despite the fact that the no-A2 rule was supposed to eliminate the “wait in line” attitude of the study section, I get the feeling it’s still there, at least a little bit: A0s get funded more rarely than A1s.

    that ratio is going to vary but the real question is how many A0s are genuine newly developed apps and how many are warmovers of a prior A1 that didn’t fund? NIH is highly motivated to obscure this so they can pretend the A2 ban worked really well to shorten the “time to funding”. What they need is some good anonymous surveying of the lucky PIs to find out when they first submitted the plan that led to the funded application.

  6. says

    Don’t forget that if all else fails, wait 37 months and it will be “new” again. Of course, by then science has probably moved on–so hopefully you and your aims have too. But it might be useful sometimes–three years seems like a long time but then when A0–>A1–>funding decision cycle times could take as long as a total of 18-24 months to shake out, three years doesn’t seem as long anymore.

  7. Grumble says

    One thing I wonder about is whether the low success rate of A1-after-triage is due to psychological effects (reviewers are biased against grants that did poorly in the first round) or to the fact that many triaged grants truly are crappe.

    Study sections are tending nowadays to triage more than just the worst 50%. They discuss only the top third or so – maybe even less. So, the proportion of triaged grants that are truly crappe is now lower than it used to be. Hopefully reviewers take this into account. (Personally, when I review grants, I don’t really care whether it was triaged before. Even if I triaged it. But then again, I am 100% uncorruptible and unsusceptible to bias.)

  8. says

    @DM I may be wrong, has happened before. But…. I’ve never seen it in all my years on study section, including lusher times than now. Despite kicking out the A2’s, there is still a queue. And triaged R01’s are behind the ones that got 20’s. People remember the 30’s (cause they talked about them), let alone 20’s. Part of the problem of triage is that the nit-pickers didn’t get a chance to pick nits, and All Else Equal, tend to be less favorable. I wish CSR would release some specific numbers about longitudinal voting in study sections.

  9. Grumble says

    “Part of the problem of triage is that the nit-pickers didn’t get a chance to pick nits”

    Typically the nits are picked by the reviewers who actually read the grant, not the rest of the study section, which hasn’t.

    That’s not to say that there isn’t a biased against grants triaged in A0, but I don’t think the reason has to do with nitpicking opportunity. I’m guessing it’s more subtle – something like, “grrrroan, this was clearly considered crappe last year, what are the chances it’s magically improved this year? This is going the be awful and I’m reading it last.”

    It’s hard to fight against first impressions, especially when your only weapon is pages and pages of dense type.

Trackbacks

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>