NIH Cut Off Comments on a Plan to Hide Grant Scores

Scientists who tried to file comments on a National Institutes of Health plan to hide grant scores found the portal closed hours before the posted October 13 deadline. An archived copy of the form listed 11:59:59 p.m. EST that night. Joshua Weitz, a biology professor at the University of Maryland, said he tried to reach the site before midnight and could not get in.

NIH has not issued a public explanation for the early cutoff. The missed hour is not the whole fight. The notice those comments were meant to answer would take the ranking number away from applicants and from the staff who still build award lists.

A Midnight Deadline That Did Not Hold

NIH opened a 60-day comment window with a request for information posted August 14, notice NOT-OD-26-088. Comments had to arrive by October 13. The agency said it would compile the replies, run a content analysis, and share that analysis after the window closed. It also said the request was for planning only and should not be read as a finished policy.

Several scientists said they could not submit in the hours before the listed time. The Society for Neuroscience had already told members the changes “could weaken scientific judgment in funding decisions and limit transparency for applicants.” Those late filings never landed. NIH did not say how many comments it had already stored, and it did not say whether it would reopen the form.

THE COMMENT WINDOW

1. November 21, 2025: NIH tells institutes to drop paylines and weigh peer review beside priorities, budgets, and the wider portfolio.
2. August 14, 2026: NIH posts NOT-OD-26-088 and opens the 60-day comment period.
3. September 8, 2026: A separate notice keeps thinner study-section discussions in place through the January 2027 council.
4. October 2, 2026: The Federation of American Societies for Experimental Biology files comments asking NIH to keep scores and percentiles.
5. October 13, 2026: The listed deadline passes; scientists say the portal was already down.

A docket that closes early is not a random clerical bruise. NIH said the comments would inform any later recommendation, including how the change would be rolled out and explained. Whoever got through before the form died now owns a larger share of that record.

Three Buckets Would Replace the Numeric Score

Study sections would still work much as they do now. Assigned reviewers would still write critiques and enter individual overall impact scores on the 1-9 scale. After discussion, all voting members would still score. NIH would still average those votes into one final overall impact score, then use that number only to drop the application into a band.

The bands would be “Most competitive” for the top 25% of those final scores, “Competitive” for 26-50%, and “Not discussed” for scores worse than 50% or for applications never brought up. The category would go to the principal investigator, the applicant institution, NIH program staff, institute leadership, and the relevant advisory council. The number behind the label would not.

WHAT WOULD STILL BE SHOWN

NIH said the score is an estimate with imperfect power to tell proposals apart, and that funding already rests on qualitative review plus other program facts. It also said the change would line NIH up with the National Science Foundation, the Department of Energy, and the U.S. Department of Agriculture. Those arguments are about a noisy number. They do not explain why the people who currently see that number inside NIH would lose it too.

What Program Officers Gain When Scores Vanish

Program officers would face a wider field of scored applications with no public rank inside each band. NIH said withholding the final number would let staff use portfolio analysis and judgment, and that the buckets were sized to avoid pay lines. The same notice would hide that number from the staff, institute leaders, and advisory councils who now see it when they build a pay plan.

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.

National Institutes of Health, NOT-OD-26-088

On August 18, NIH’s extramural news shop put the same point in plainer operational language. The size of the categories, it said, was designed to empower program staff to read peer review as a whole, and the bands were made large enough to avoid artificial pay lines and to broaden the set of applications under funding consideration. A top-quarter band that wide, with no rank inside it, treats a large set of scored proposals as interchangeable. That is a quiet form of lottery among the “most competitive,” and NIH designed the width on purpose.

In November 2025 the agency had already told Institutes, Centers, and Offices to stop leaning on paylines, the cutoffs about half of them used to turn a score into a yes or no. Directors kept the power to decide what their institutes fund. Peer review was supposed to be read in full, beside health priorities, scientific opportunity, the workforce, available money, and the rest of the portfolio. Hiding the score is how that 2025 turn gets teeth. Staff can no longer be second-guessed against a number they no longer receive.

Most January Applications Will Carry No Score at All

The comment fight is about a proposal. A related change is already live. On September 8, NIH said peer review modifications still in force after the 2025 funding lapse would continue through the January 2027 Advisory Council, notice NOT-OD-26-114. That notice had no comment period.

THE JANUARY 2027 REVIEW PILE

• The load: NIH said it already had more than 35,000 applications under review for the January 2027 council, before small-business deadlines, a pile it expects to beat the May 2026 record of 33,640.
• The discussion cut: Most meetings will discuss 30-35% of applications, down from the usual share of about 50%.
• The middle third: Applications voted into the middle third are labeled “competitive but not discussed” and still sit in the funding pool.
• The score: Only discussed applications receive an overall impact score on the summary statement.

For the middle and bottom of that record pile, there is already no overall impact score to withhold, because none was assigned. Summary statements are thinner too. Narrative paragraphs that used to recap the panel talk are gone, replaced by a sentence on how together the votes were and bullets on what drove the score. The three assigned critiques remain. The two notices are not the same document, and mixing them is a common error, but they lean the same way. One is already how January applications will be handled. The other would make the missing number the rule for everyone.

FASEB Says Keep the Percentile and the Audit Trail

FASEB filed on October 2 and agrees with NIH on one narrow point: small gaps in impact scores may not mean much on their own. Its answer is more context, not less information. The letter offered seven recommendations across four areas and asked NIH to keep reporting scores and percentiles both to applicants and to the people inside NIH who rank a council round.

HOW LABS USE THE NUMBER

• Resubmission: The score is the only calibrated signal of how a proposal sat against the rest of the round, and the 26-50% “Competitive” band would not separate a near miss from a rewrite.
• Early careers: FASEB said early-career researchers have the most to lose, and that the hit would fall hardest at less well-resourced institutions.
• Campus money: Institutions use scores for bridge funding, startup extensions, protected time, and reappointment.
• Staff workload: Program officers now compare applications with scores and percentiles; without them they would rebuild rankings from critiques that were not written for that job, during a stretch of reduced NIH staffing.
• Oversight: Scores and percentiles now let NIH leadership, Congress, and watchdogs trace how merit review feeds awards; FASEB wants an auditable file that still holds the numbers, even if categories are added.

FASEB also asked NIH to say whether category cutoffs would be set by study section, by cycle, or agency-wide, and how Early-Stage Investigator rules and some K and F awards would be handled. It wants full guidance out before any switch, and at least one full application cycle between that guidance and the start date. As a fallback, it said NIH could report the bands beside the scores, with a note that tiny gaps may not be meaningful, rather than burying the number.

Written critiques would still arrive. They do not travel. A 19th percentile in one year can be compared with a 19th percentile in another, and with a colleague’s score in a different study section. A label that covers 26-50% cannot. Labs planning a resubmit, and chairs deciding whether to extend a startup clock, would be guessing inside a bucket.

December 11 Reopens the Political-Appointee Fight

Scientists who call the scoring plan a back door for politics are not inventing the surrounding calendar. In May, the Office of Management and Budget proposed a rewrite of government-wide grant rules that would put senior political appointees into the award process after technical review. Researchers at the University of North Carolina at Chapel Hill, reading 52,322 comments posted in full, found about 95% opposed that plan and about 1% supported it. The docket held 496,769 comments.

Congress then blocked OMB from finishing that rule in a stopgap spending bill. The current funding law runs through December 11. In September, a separate idea for an executive order that would have given a board of political appointees, including OMB Director Russell Vought and NIH Director Jay Bhattacharya, a veto over NIH grants drew a public warning from Sen. Susan Collins, the Maine Republican who chairs the Senate Appropriations Committee. Collins wrote that research should be funded “based on scientific merit and value rather than political ideology.” People familiar with the talks said the White House backed off the order. The White House did not put that retreat on paper as a final decision.

WHAT WE KNOW

• The listed close: The comment form, as archived, named 11:59:59 p.m. EST on October 13 as the deadline.
• The lockout: Weitz and other scientists said they could not submit in the hours before that time.
• The proposal’s reach: The final overall impact score would be withheld from applicants, institutions, program staff, institute leaders, and advisory councils.
• The live practice: For the January 2027 council, most meetings will discuss 30-35% of applications, and only that discussed slice gets an overall impact score.

WHAT IS UNCONFIRMED

• The cause: NIH has not said why the portal went down early, or whether the outage was a glitch, a clock set to the wrong zone, or a choice.
• The docket: The agency has not said how many comments it accepted, or whether it will reopen the form.
• The end state: NOT-OD-26-088 is still a planning request, not an adopted rule, and NIH has not named an effective date.

NIH said it will compile what it received and share a content analysis. That analysis will rest on whoever got through before the portal died. The number those comments were about is the same number program staff still use, when they have it, to sort a pile already larger than 33,640. After December 11, the louder fight over who signs the awards returns to the calendar. The quieter one, over who is allowed to see the score, is already on the page.