Things NIH Program Officers probably shouldn’t say to applicants
May 27, 2011
I have blogged, now and again, about the ~four-tiered nature of the selection process for NIH grant awards. These consist of initial peer-review, Program Officer expression of programmatic interests, the National Advisory Council (of peer scientists) for each Institute or Center of the NIH and the Director of each IC. We tend to clump the latter three into “Program”, since they really all do express the interests of a particular IC whereas the initial peer review process focuses so strongly on the quality of the science and the strength of the overall proposal.
Unlike the apparent position of many fellow scientists, I have no problem with this multi-tiered selection process nor, more pointedly, do I have a problem when Program interests override or overturn the strict priority order / priority score that comes out of initial peer review.
I have even mentioned now and again that I think a little more transparency from Program on how they select applications outside of the review order would go a long way to damping down the whinging.
A bit of Program Officer transparency has emerged, however, that gives me pause.
Commenter Rose at the writedit blog expressed disappointment that her PO decided to pass on a K23 revision that received a 20 priority score. That’s pretty decent in my view and it means the panel voted an average of 2 out of the 1-9 possible score range. The interesting part is the reason offered as to why program was giving it a pass:
On the phone she was a bit harsher a few weeks ago and basically told me she didn’t agree with the reviewers’ score and comments and thought they were “just being nice” to me. When I asked why they would just be nice, she said “because it is a resubmission and you are working with underserved populations.”
Hmm, sounds like the PO making a decision not just about the science but the peer reviewers’ opinion of the scientific plan, no?
We don’t take the decision on not funding an application lightly. We really don’t like to fund projects without a suitable control group, which is the major reason for not funding your application.
oh boy. Smells like one of those tricky “Approach” issues that already can become a matter of extensive discussion in a study section, right? But then….the panel bought the applicant’s argument in revision.
The reviewers supported my reasons the first time for not having a control group (clustering issues for an RCT and inability to match on ethnicity and language for a case-control study)
Okay, I’m leaning the applicant’s way on this already, without any further info. This is a K23, remember? $20K-$50K in research funding, the rest is for the obligatory 75%time+ salary support. The project almost inevitably is tagged on to a more substantial research project funding the data collection, subject poor or some sort of infrastructure, if you ask me. Or maybe it is just a small potatoes human subjects study. Either way, not a lot of room here for adding in extra groups willy nilly.
And the study section bought the arguments.
Here’s the PO making a decision to override. Not for what I would describe as legitimate programmatic interests in the topic domain, but as a single person (or small handful within the Branch/Division) overturning on the basis of scientific structural issues in the approach.
That sort of interference by Program is something that I find to be a bit of a problem.
NIGMS blog on the role of the Advisory Council