NIH's plan to hide peer-review scores swaps a contestable number for a 25-point band

The National Institutes of Health posted a Request for Information on August 14 titled "Request for Information (RFI) on Proposed Changes to Reporting Outcomes from NIH Peer Review," proposing that grant applicants no longer be shown the numerical scores their applications receive in scientific peer review ^[1]. Instead of a composite score, an applicant would see one of three categories: "Most competitive," drawn from the top 25% of final overall impact scores, "Competitive," covering the 26th to 50th percentile, or "Not discussed," meaning scores above the 50th percentile ^[2].

That matters because the composite score has been the operative number in the system. Applications have traditionally been reviewed by a committee of external scientific peers, each assigning a numerical score, with the composite in most cases determining whether the application is funded when funding is available ^[3]. Applicants could then compare their score to published paylines for a given Institute, Center or Office, which told them whether they were close and, if not, how far off, and therefore whether to revise and resubmit or abandon the approach ^[4]. According to the Forbes account, external review meant outside NIH and outside the federal government, structured that way both to keep politicians and government insiders from exerting undue influence over awards and to ensure reviewers had relevant expertise ^[5]. Reviewer lists have been made available to the applicant and the public ^[6].

The loss here is not the score itself but the gap. A band spanning 25 percentiles cannot be compared with a payline expressed as a threshold, so an application sitting just outside the funding line and one sitting well outside it become indistinguishable to the principal investigator ^[7]. That is the same information a research office uses to decide where to put bridge funding, which resubmissions to back, and which programs to wind down. If your internal rules key off proximity to a payline, they will need a new trigger, because the input will not arrive.

The second consequence is discretionary room. Forbes columnist Bruce Y. Lee frames the central question as how far the change, if enacted, moves NIH funding decisions away from science and scientists and closer to the U.S. President and his political appointees ^[8]. He notes the announcement went up late on a Friday, before the weekend, when few people are watching ^[9], and attributes the proposal to NIH under the Trump administration ^[10]. That is a single-sourced characterisation, but the structural point stands on the mechanics: a number is contestable in a way a category is not, and an applicant told only that they were "Not discussed" has nothing specific to argue with.

Comments are open until October 13, roughly 60 days from the posting ^[1][11]. Two things are worth tracking in whatever emerges. First, whether NIH says anything about the future of published paylines and public reviewer lists, since bands without paylines remove the last external benchmark ^[4][6]. Second, whether institutions with large NIH portfolios file comments on the record, because the comment file is the only formal leverage available before the policy is set ^[1]. In the meantime, treat internal scientific review as your own scoring system and keep the numbers, since it may soon be the only quantified read your investigators get.