Whenever I come across government announcements like this, I always wonder if discussing them in depth is a bit too much “inside baseball” to be of interest to SBM readers. For no topic is this more true than when I start discussing the nitty gritty of how the National Institutes of Health evaluates grant applications in its peer review study sections to decide which ones will be funded. The cited RFI (Request for Information NOT-OD-26-088) proposes to alter this peer review by altering how scoring results for NIH grant applications are reported to applicants in a manner that belies the ridiculous claims of this administration to be the “most transparent in history.” (I’ll get into the details shortly.) On the one hand, the details of how panels of scientists evaluate and score grant applications and then how those scores lead to ranking the grant applications by percentile, (low percentiles are higher scored grants), which then, with fairly uncommon exceptions, determines which grants are funded tend to be technical and boring to the non-scientist (and even to those with science-training who have never dealt with the NIH). On the other hand, as I’ve said many times before, arguably the most important characteristic of the NIH that has led it to being arguably the world’s foremost engine of biomedical discovery over the last eight decades has been the incredible efforts made to make grant review as fair and based on scientific merit as possible.
Is the NIH perfect at this? Of course not! However, say what you will about the NIH and its grant funding process, it can’t be denied that incredible effort has been made over the last several decades to continually tweak the peer review process to emphasize scientific merit, while also balancing policy goals that are good for science policy in the US and that contribute to a stronger scientific workforce, such as giving newly minted independent investigators a chance to be NIH funded by carving out funds for them or giving them a leg up when competing with long-established investigators and trying to ensure that previously underrepresented groups receive funding. (Of course, it’s the latter efforts that have sparked a backlash and this administration’s war on “DEI.”) Ever since the Trump administration took power over a year and a half ago, it has been doing its damnedest to dismantle the science-based, reasonably objective peer review process used by the NIH (and other federal science agencies) to decide which external grant applications merit being funded. Before I discuss the NIH RFI, it’s useful to look at what has occurred thus far on this front.
Lysenko 2.0 arrives
When last I wrote about peer review at the NIH, I noted that a recent proposed change in federal rules by OMB Director Russell Vought was, in essence, a big step forward in the Lysenko-ization of the NIH and Lysenkoism 2.0 in action. As most SBM readers know, Trofim Lysenko was a Soviet-era agricultural scientist whose ideas—which included the denial of Mendelian genetics and Darwinian evolution by natural selection in favor of a version of Lamarckism—Stalin loved not because of how scientific they were, but rather because they aligned with his perception of Soviet ideology. Unfortunately, when put into practice, Lysenko’s ideas were utterly disastrous for Soviet agriculture for decades (and, in one instance, Chinese agriculture), resulting in crop yields plummeting, turbocharging periodic famines from the 1930s through the early 1960s, when he was removed from power after his influence had slowly waned in the years after Stalin’s death. Yet, because his ideas were ideologically in sync with the regime, particularly under Josef Stalin, Lysenko yielded ruthless power over Soviet agricultural science for decades, and scientists who dissented in favor of good science risked being ostracized, fired, imprisoned, and in some cases even executed. This history is why the term “Lysenkoism” is now commonly used to refer to the domination of science by a pseudoscientific ideology, rather than evidence, experimentation, and data. History also shows that it took Soviet agricultural science decades to recover after Lysenko had fairly been deposed.
The Lysenkoism in Vought’s proposed rule changes for federal grantmaking was easy to see. I wrote in detail about it here and here, but I’ll briefly recount the CliffsNotes version, because it’s important to your understanding of this NIH RFI, and I don’t want to force you to have to click on links if you don’t want to. In brief, the proposed rule changes included:
- De-emphasizing peer review as “advisory” only. In the past, the priority scores of grant applications was the primary way that grant applications were ranked for funding, with only occasional interventions by Institute directors to advance specific scientific priorities. Under the new rules, political appointees will be able to ignore peer review in order to accomplish the next aim.
- Grants must be consistent with and advance “administration priorities.” In other words, if these rules are implemented, political appointees will be empowered to overrule peer review to reject important projects based in rigorous science in favor of crap if they think those projects “advance administration priorities.”
- Grants must not “fund, promote, encourage, subsidize, or facilitate” the usual cast of right wing bogeymen, including “racial preferences or other forms of racial discrimination by the grant recipient”; “denial by the grant recipient of the sex binary in humans or the notion that sex is a chosen or mutable characteristic”; “illegal immigration”; or “any other initiatives that compromise public safety or promote anti-American values,” “anti-American values,” apparently, being defined by the Trump administration.
There were other disturbing points in the proposed rule changes, but these are the key three that stood out as designed to fuel Lysenko 2.0. It doesn’t take a scientist to see how such rules would, in essence, end scientific independence of the NIH, allowing it to become, as its Director, Podcast Jay Bhattacharya once promised, the “research arm of MAHA” (the Make America Healthy Again movement). The other way that I like to put it is that the Trump administration tends to view science funding not as a public good that should be protected and nurtured to fund the best science out there, but rather as patronage not unlike something out of The Godfather, yet another bin of cash that can be doled out to loyalists and withheld from the “disloyal.”
With that background, it will, I hope, be easy for you to understand how the proposed changes in the reporting of peer review results will facilitate the Trump administration’s clear desire to move the center of gravity in the peer review carried out at NIH study sections away from science-based grantmaking towards a more ideology-based grantmaking.
Proposed Changes to Reporting Outcomes from NIH Peer Review: Facilitating Lysenko 2.0
Now we’re ready to discuss this RFI (NOT-OD-26-088, issued on August 14, 2026) and entitled Request for Information (RFI) on Proposed Changes to Reporting Outcomes from NIH Peer Review. First, the background:
In November 2025, NIH announced the Unified Funding Strategy, based in the long-standing two-stage system of review NIH uses in making funding decisions. The framework calls for NIH Institutes, Centers, and Offices (ICOs) to consider peer review outcomes in the context of their and NIH’s priorities, strategic plans, and budgets, instead of relying on funding paylines to develop pay plans, which over-emphasize numerical scores. Core tenets of this funding strategy include that funding policies must “prioritize scientific merit; ICOs should consider peer review information in its entirety.”
As a reminder, the Unified Funding Strategy purports to want all Institutes, Centers, and Offices (ICOs) to follow these principles in peer review and deciding which grants to be funded:
- Align with the NIH’s mission
- Prioritize scientific merit; ICOs should consider peer review information in its entirety
- Integrate a breadth of topics and approaches relevant to the ICO’s priorities
- Consider investigator career stage and promote sustainability of the biomedical research workforce
- Promote broad distribution and geographic balance of funding, considering the total amount and type of NIH funding already available to each investigator
- Align with the availability of ICO funds
None of this sounds, at least on the surface, objectionable. In fact, it’s not all that different from what the NIH has been doing for a while. Interestingly, the document cited in the quote above does not mention that ICOs should emphasize funding administration priorities, although Podcast Jay’s precursor document, Advancing NIH’s Mission Through a Unified Strategy, sure enough does, as I wrote at the time. One can even make an argument that current NIH funding decisions might indeed overemphasize numerical scores, but here’s the rub. Current funding levels (or, as we colloquially call them, paylines) for most institutes are well under the 10th percentile, meaning that less than 10% of grant applications per cycle are funded. Some are much lower. When the paylines are that tight, it almost forces the NIH to rely on numerical scores, because to do otherwise interjects a lot of subjectivity into the decision-making process.
On the other hand, NIH priority scores and percentiles are a rather blunt instrument, as I’ve discussed before while analyzing certain questionable proposals for revamping how the NIH does peer review. For example, is a grant that scored at the 3rd percentile substantially better than one that scored at the 5th percentile? Probably not detectably so, and both are clearly fantastic grants that greatly impressed the peer reviewers and the study sections that evaluated them. Yet, if the payline is at the 4th percentile, one will be funded, and the other will not be. Moreover, one reviewer who really hates the grant (and assigns it unjustifiably low priority scores) can affect the overall averaged aggregate priority score just enough to move a fundable grant into the unfindable category. On the study sections where I’ve served on an ad hoc basis, I’ve tended to observe, though, that the discussion tends to cause reviewers issuing scores far outside range that includes other reviewers to move their scores closer to the mean of those reviewers.
For all its flaws and imperfect objectivity, one aspect of the current system is that it is fairly transparent. Sure, the actual discussions at the study section are not revealed, but the peer review reports of the primary and secondary reviewers, along with the priority scores issued by each, are returned to the applicants. In the case of grant applications that are among the best one-half or so, applicants also receive back a lengthy summary statement with the individual priority scores for each criteria rated (Significance, Investigators, Innovation, Approach, and Environment), the percentile ranking of the grant, and a synthesis of the discussion of the grant application at the study section meeting.
In fact, let me quote the RFI regarding the current system:
NIH uses a two-stage system of review in developing extramural funding recommendations. In the first stage, scientific experts (peer reviewers) evaluate the scientific and technical merit of grant applications. Each proposal is assigned to a peer review group and is assigned to a minimum of three peer reviewers from that group to prepare written critiques. Each written critique addresses the likely overall impact of the proposed work, as well as strengths and weaknesses guided by the review criteria in Section V of the relevant Notice of Funding Opportunity. The assigned reviewers enter their individual scores from 1-9 reflecting their preliminary assessment of the overall impact of the proposal and criterion/factor scores for the review criteria. Generally, applications with average preliminary overall impact scores in the top 50% of applications assigned to the review group are discussed. After discussion, all reviewers–assigned and not–enter their final individual overall impact scores, which gets averaged into one final overall impact score.
The review outcome is currently reported as a summary statement which includes a resume of discussion for discussed applications, the written critiques of the assigned reviewers, criterion/factor scores from assigned reviewers, and a final overall impact score averaged from all voting members of the panel. The principal investigator(s), applicant institution, NIH program staff, and the members of the funding ICO’s advisory council have access to these review outcomes.
NIH program staff examine proposals and consider the final overall impact scores given during the peer review process, percentile rankings (if applicable), and the summary statements, in consideration of the ICO’s priorities. Program staff then provide a proposed funding plan to ICO leadership for award decisions.
This is more or less how funding decisions are made now, with final funding decisions being made at the Advisory Council level. In brief, the Advisory Council or Board for each Institute will examine the applications ranked on the basis of percentile scores, apply tests on criteria such as mission relevance and maintaining a balanced research portfolio before finalizing funding decisions. In practice, for the most part with relatively few exceptions, percentile ranking determines which grant applications are funded, with most Institutes funding the lowest percentile application first and then working upwards through the applications until funds are exhausted.
Now here’s where the RFI gets most “un-transparent,” although the justification sounds benign enough:
In late 2025, a working group of NIH leaders was charged with recommending changes in how NIH reports out peer review evaluations of scientific merit, to facilitate the Unified Funding Strategy. Considerations for their work included the following: (1) funding decisions rely on qualitative peer review assessments of merit, along with other important programmatic considerations; (2) peer review scores are estimates with imperfect discriminative ability; and (3) peer review outcomes are multifaceted and contain valuable information beyond the overall impact score. The group reviewed literature on the predictive value of peer review scores for scientific outcomes, considered reporting approaches utilized by other funding entities, and examined NIH peer review data.
Yes, this is all true (although in this administration I can’t help but read “programmatic considerations” as “administration priorities), but, as you will see, the changes proposed do not address these flaws in NIH peer review:
The first-level peer review process is proposed to be modified to place an application in one of three overall impact categories:
- “Most competitive” (based on the top 25% of final overall impact scores),
- “Competitive” (based on the 26-50% of final overall impact scores), or
- “Not discussed” (scores >50%, or ND)
With the exception of the reporting of final overall impact scores and percentiles, all other initial review group procedures would be retained. As happens currently, all voting members of a peer review group would continue to enter individual overall impact scores for each application discussed. A final overall impact score would still be calculated for each discussed application for purposes of assigning the application to the appropriate category. The category assigned would be communicated to the principal investigator(s), applicant institution, NIH program staff, ICO leadership, and the relevant NIH Advisory Council. However, the final overall impact score or percentile on which the categorization was based would not be released to any of these parties.
Other reporting procedures would be unchanged. Discussed applications would still receive a resume of discussion and all applications would still receive written critiques from the assigned reviewers, to include individual criterion/factor scores.
In other words, priority scores would still be issued, and percentile rankings would still be generated, but investigators submitting grants would not receive any of them. They’ll just get the comments from the primary reviewers, the summary statement if they aren’t in the “not discussed” group, and the percentile group they are in.
Such a system is basically useless for grant applicants. Let me explain. Let’s say that you submit a grant, and it’s not funded. You learn, however, that it fell into the “most competitive” group, meaning somewhere in the top 25%. In deciding whether to resubmit, think of this: When paylines are under the 10th percentile, there is a huge difference between a grant application whose mean overall priority score leads it to be ranked as 25th percentile versus one ranked as, say, 11th percentile. The latter grant is in striking distance of being funded on a resubmission, the first grant much less so. Moreover, what is being proposed is a lot less transparency, not more transparency. The “raw data”—i.e., the raw priority scores and the percentile ranking—that led to an application being grouped in one of the above three categories will, contrary to past practices, be withheld from the grant applicant.
So how does Podcast Jay, who must have signed off on this RFI, justify this proposal? Here’s how:
Reporting categorization as “most competitive,” “competitive,” and “not discussed” in lieu of final overall impact score aligns with the Unified Funding Strategy which encourages ICOs to actively balance many competing and dynamic factors, including peer review, health priorities, scientific opportunities, the workforce, availability of funds, and the wider research portfolio when determining the most meritorious research ideas to support. Categorizing proposals in this way conveys valuable information about the assessed scientific merit from peer review of without providing over-emphasis on overall impact scores. Omitting a final overall impact score from what is reported will empower program officials to utilize their substantial expertise, portfolio analyses, and good judgment to make award recommendations that better incorporate ICO strategic plans and NIH priorities. The removal of final overall impact scores also supports a focus on substantive comments from the assigned reviewers and the panel, elevating the role of scientific peer review. Finally, this change would align NIH practices with those followed by other major funders such as the National Science Foundation, Department of Energy, and U.S. Department of Agriculture.
It is, of course, true that the NSF and Department of Energy for instance, do use a more qualitative system that doesn’t include numeric scores, which does inherently allow for more judgment and subjectivity on the part of program managers/officers regarding funding decisions. it is also true that one huge difference between the NIH and these other agencies is that the peer review system used by the NIH is far larger and more formalized, with hundreds of standing study sections, which each have members who serve for years plus some ad hoc reviewers, all run through the Center for Scientific Review. In contrast, the NSF and DoE, for instance, lack that infrastructure, with their program managers or officers tending to send requests for peer review to ad hoc panels of chosen experts. It should be noted, also, that the NIH disburses close to $40 billion a year in extramural funding (projects done outside of the NIH campus), compared to one fifth as much or less by the NSF and the other agencies. It therefore makes sense that the NIH has a more formalized, rigorous process.
One can readily see that the NSF, DoE, and USDA don’t need to be targeted this way because the existing systems already allow for a lot of leeway in funding decisions. The NIH, on the other hand, has a formal numerical system that, for whatever flaws it might have, would make it a lot more difficult to hide the hand of political appointees in boosting grant applications with lower scores that “advance administration priorities” (or are to scientists aligned with the administration) and not funding high-scoring applications that contain the dreaded buzzwords like “DEI” or “disparities” and/or would go to scientists who are critical of the administration. Indeed, perusing the NIH subreddit, I didn’t even have to look for people who suspect exactly the same thing that I do, namely that this proposed new scoring reporting system is a not-so-subtle way to make it easier for political appointees to influence grant funding decisions.
For example, let’s quote some scientists who compete for NIH funding or work for the NIH tell us what it’s all about:
And an excellent addition to the quote above:
And:
And:
And:
Exactly. Under another administration without the history of the Trump administration’s assaults on the NIH, I might have been less cynical and a bit more open-minded about—or at least less hostile to—this RFI. I might have even been open to the possibility that such a system might be worth trying out, albeit with more categories than the three listed in the RFI. However, given the recently proposed OMB rule changes designed to de-emphasize scientific peer review and empower political appointees to have final say over which grant applications are funded, it’s impossible for me not to take a much more alarmist view of this proposed change. There is little doubt in my mind that the primary intent of this proposed change in how NIH reports its scoring to applicants is to make it very difficult for applicants and outside analysts to identify very high-scoring grants that were rejected for not aligning with administration priorities or coming from institutions in disfavor, as well as grants that might not have scored as well but are the sorts of MAHA-aligned “gold standard science” that the current administration favors.
There is a 60 day comment period. If you are an scientist or academic, particularly if you have or have had NIH funding in the past, there’s still time to comment. The rule changes proposed by the OMB generated a tsunami of comments, nearly all negative. Whether that will matter in the end, I don’t know, but even this administration might find it difficult to ignore such unpopularity. Perhaps the same can happen here.
Finally, remember this. Sometimes, “inside baseball” considerations matter—and matter a lot. Changing technical procedures known to relatively few (e.g., scientists who submit grants to the NIH), such as how NIH peer review functions, is but one strategy by which this administration is ushering in Lysenko 2.0 in a manner that the general public doesn’t see. Sadly, it seems to be noticed only by academics like me who have knowledge of the issues involved. Worse, compared to Donald Trump, Robert F. Kennedy Jr., and Russell Vought, Lysenko was a rank amateur. After all, he only turned Soviet agricultural science and genetics into ideologically driven sciences. This administration is trying to do for all science in four years what Lysenko did to just two over decades.