NIH Awards for ESI with preliminary data verboten!

The NIH has launched a new FOA called the Stephen I. Katz Early Stage Investigator Research Project Grant (Open Mike blog post). PAR-21-038 is the one for pre-clinical, PAR-21-039 is the one for clinical work. These are for Early Stage Investigators only and have special receipt dates (e.g. January 26, 2021; May 26, 2021; September 28, 2021). Details appear to be a normal R01- up to 5 years and any budget you want to try (of course over $500k per year requires permission).

The novelty here appears to be entirely this:

For this FOA, applications including preliminary data will be considered noncompliant with the FOA instructions and will be withdrawn. Preliminary data are defined as data not yet published. Existence of preliminary data is an indication that the proposed project has advanced beyond the scope defined by this program and makes the application unsuitable for this funding opportunity. Publication in the proposed new research direction is an indication that the proposed work may not be in a new research direction for the ESI.

This will be fascinating. A little bit more specification that the scientific justification has to rest on published (or pre-printed) work only:

The logical basis and premise for the proposed work should be supported by published data or data from preprints that have a Digital Object Identifier (DOI). These data must be labeled and cited adjacent to each occurrence within the application and must be presented unmodified from the original published format. Figures and tables containing data must include citation(s) within the legend. The data should be unambiguously identified as published through citation that includes the DOI (see Section IV.2). References and data that do not have an associated DOI are not allowed in any section of the application. Prospective applicants are reminded that NIH instructions do not allow URLs or hyperlinks to websites or documents that contain data in any part of the application

So how is this going to work in practice for the intrepid ESI looking to apply for this?

First, there is no reason you have to put the preliminary data you have available in the application. One very hot comment over at the Open Mike blog post about the proposals being unsupported and therefore the projects will be doomed to failure is totally missing this point. PIs are not stupid. They aren’t going to throw up stupid ideas, they are going to propose their good ideas that can be portrayed as being unsupported by preliminary data.

Twill be interesting to see how this is interpreted vis a vis meeting presentations, seminars and (hello!) job talks. What is a reviewer expected to do if they see an application without any preliminary data per the FOA, but have just seen a relevant presentation from the applicant which shows that Aim 1 is already completed? Will they wave a flag? See above, the FOA says the “existence” of preliminary data, not the “inclusion” of preliminary data will make the app non-compliant.

But there is an aspect of normal NIH grant review that is not supposed to depend on “secret” knowledge, i.e., that available only to the reviewer, not published. So it is frowned upon for a reviewer to say “well the applicant gave a seminar last month at our department and showed that this thing will work”. It’s special knowledge only available to that particular reviewer on the panel. Unverifiable.

This would be similar, no?

Or is this more like individual knowledge that the PI had faked data? In such cases the reviewers are obligated to report that to the SRO in private but not to bring it up during the review.

If they ARE going to enforce the “existence” of relevant preliminary data, how will it be possible to make this fair? It will be entirely unfair. Some applicants will be unlucky enough to have knowledgeable whistle blowers on the panel and some will evade that fate by chance. Reviewers being what they are, will only variably respond to this charge to enforce the preliminary data thing, even if semi-obligated. After all, what is the threshold for the data being specifically supportive of the proposal at hand?

Strategy-wise, of course I endorse ESI taking advantage of this. The FOAs list almost all of the ICs with relevant funding authority if I counted correctly (fewer for the human-subjects one, of course). There is an offset receipt date, so it keeps the regular submission dates clear. You can put one in, keep working on it and if the prelim data look good, put a revised version in for a regular FOA next time. Or, if you can’t work on it or the data aren’t going well, you can resubmit “For Resubmissions, the committee will evaluate the application as now presented, taking into consideration the responses to comments from the previous scientific review group and changes made to the project.” Win-win.

Second strategy thing. This is a PAR and the intent is to convene panels for this mechanism. This means that your relative ESI advantage at the point of review disappears. You are competing only against other ESI. Now, how each IC chooses to prioritize these is unknown. But once you get a score, you are presumably just within whatever ESI policy a given IC has set for itself.

I’m confused by the comments over at Open Mike. They seem sort of negative about this whole thing. It’s just another FOA, folks. It doesn’t remove opportunities like the R15. No it doesn’t magically fix every woe related to review. It is an interesting attempt to fix what I see as a major flaw in the evolved culture of NIH grant review and award. Personally I’d like to see this expanded to all applicants but this is a good place to start.