Science Policy

Policy Corner: Policy change for peer review of extramural grant proposals

Priya S. Shah, University of California Davis

ASV Newsletter, August 2026

On August 14, 2026, the National Institutes of Health (NIH) released NOT-OD-26-088, which proposes a policy change to peer review of extramural grant proposals. A request for information (RFI) comment period is open until October 13, 2026. We summarize the proposed changes and their possible impact on ASV members. We encourage them to submit comments to NIH.

In this notice, NIH proposes making reviewer feedback more qualitative so program officers and institute directors can make funding decisions that balance scientific merit and programmatic considerations. One of the primary reasons for the proposed change is that “peer review scores are estimates with imperfect discriminative ability; and peer review outcomes are multifaceted and contain valuable information beyond the overall impact score”.

While most veteran NIH-funded researchers will not argue with the imperfect nature of the NIH peer review process, understanding how NIH peer review has been conducted for decades is critical to placing these proposed changes in context. Compared to other federal funding agencies, NIH has been the most quantitative in its peer review and funding process. The current process is as follows:

Three reviewers read and critique a grant proposal. They assign a preliminary impact score between 1 and 9 based on three factors: significance and innovation, rigor and approach, and investigator and environment. For each study section, these preliminary scores are used to rank the proposals. The top half to one third of proposals usually get selected for further discussion with the entire panel. After discussion, the entire panel provides an overall impact score between 1 and 9 based on the same three factors noted earlier, while also accounting for nuances emerging from discussion. This score gets averaged, percentile-ranked, and is provided to the investigator, the NIH program officer, and the NIH advisory council. Until 2025, this number typically dictated whether a proposal was funded. Most NIH institutes were payline-driven, and it was rare for a grant with a lower score to be funded before one with a higher score.

In November 2025, NIH released a new Unified Funding Strategy. In this new strategy, they eliminated strict paylines across all institutes. Several institutes, including the National Institute of General Medical Sciences (NIGMS), already used this payline-agnostic approach. This change means the National Institute of Allergy and Infectious Diseases (NIAID, a major funder for ASV members), can fund proposals “out of order” and in line with programmatic considerations.

The proposed change in NOT-OD-26-088 is the next step in this effort of overhauling NIH peer review. If enacted, the proposed changes would hide the final overall impact score from the investigator, program officer, and advisory council. Instead, the proposals would be categorized into three bins:

  1. Most competitive (top 25% of scored proposals)
  2. Competitive (between the top 25% and 50% of top scores)
  3. Not discussed (truly ND or the bottom 50% of scores)

The overall resume of discussion and three separate critiques would still be public.

The main arguments for the advantages of this change are that it will allow for funding decisions more in line with the Unified Funding Strategy, which encourages institutes to make funding decisions that balance competing and dynamic factors, including peer review, health priorities, scientific opportunities, workforce, availability of funding, and breadth of research portfolio.

In fact, the proposed changes are not too different from what other federal funding agencies already do, including the National Science Foundation (NSF), Department of Energy (DOE), and the US Department of Agriculture (USDA). At face value, this change will allow for more flexibility in funding decisions and acknowledge that a purely quantitative funding strategy may be misguided at times – is a 4th percentile proposal truly different from a 9th percentile proposal, given the stochasticity of peer review and score compression? They are certainly both great proposals with minor flaws at most.

However, the proposed changes likely fall short of solving this problem, and potentially create new ones related to investigator feedback, tracking general trends, and transparency of administrative priorities. Let’s tackle the easy ones first. The changes would obscure important information that investigators can use to evaluate what they should do. Even NSF allows for more nuance in final funding categories – they have four instead of three:

  1. Highly competitive/High priority
  2. Competitive
  3. Low competitive
  4. Not competitive/Not recommended

While this seems like a minor difference, in the world of 5th-percentile (theoretical) paylines, there isn't enough money to fund all the “most competitive” proposals. Being at the 23rd percentile is very different from being at the 8th percentile of deciding a path forward for resubmission. Many, if unsure about the viability of a proposal, may view it as safer always to resubmit. Yet, with a new maximum of six PI submissions per year, this is probably not a good allocation of scientist effort. Why dispense with this information if it is already being collected? As scientists, shouldn’t we review all the information available to us in making decisions?

Taking this point to its conclusion, why throw away all that information and deprive us people making these important funding decisions? Information that can detect score compression or other anomalies in specific study sections. Information that can be used to uncover reviewer bias (and ideally create policies to remedy them). Information that, in the worst possible scenario, can be used to intentionally obscure how funding decisions are made, and prioritize funding for proposals with less scientific merit but in line with administration political goals. The lack of transparency is a concern given the administration’s track record to date. The lack of transparency could be especially impactful for ASV members whose research topics have become lightning rods for this administration – fundamental coronavirus biology, mechanisms of viral evolution, vaccine development and efficacy, and disparities in health from viral diseases. While the proposed changes may reflect the best intentions of the advisory panel for the Unified Funding Strategy, ASV members should consider some nuances carefully.

You can submit comments about this proposed change to the NIH until October 13, 2026. You can submit comments here.

Disclosure: Priya S. Shah is a standing member of an NIH study section.