BMJ Public Health retracted a May 2024 paper by Saskia Mostert and colleagues on excess mortality and COVID-19 vaccination on August 11, 2026 — roughly two years after the journal first placed an expression of concern on the article in June 2024. The paper had circulated well beyond the usual readership of a public-health journal: it was cited in a June 2026 U.S. Senate subcommittee hearing on alleged harms of COVID-19 vaccination, arriving in a legislative venue while the underlying integrity question was still formally unresolved.
For CASRAI’s audience — research-administration offices, research-integrity officers, and editorial and publishing staff — the case is less interesting for its politics than for its mechanics: a roughly two-year gap between the moment a journal itself said it had open questions and the moment it acted on them, with a legislative citation landing squarely inside that gap.
What was published, and what the retraction says
The article, led by Saskia Mostert of Princess Máxima Center for Pediatric Oncology in the Netherlands with co-authors Minke Huibers, Gertjan Kaspers, and Marcel Hoogland, was published in BMJ Public Health on May 6, 2024. According to BMJ Group’s retraction notice and Retraction Watch’s reporting, the journal’s stated grounds for retracting the paper were that it had given rise to widespread misreporting and misunderstanding
— specifically, claims reading the paper as establishing a direct causal link between COVID-19 vaccination and mortality — and that the article did not adequately convey the limited nature of its own underlying data and methods. BMJ Public Health first placed a formal expression of concern on the paper about a month after publication, in June 2024, after outside criticism (including, by later reporting, disputes over how the paper characterized vaccine data) raised concerns severe enough that the journal described some of the paper’s framing as containing inappropriate insinuation.
Mostert has publicly disagreed with the retraction itself while saying she requested it jointly with her institution, citing ongoing concerns about how the paper was being misinterpreted, and has said proposed revisions she offered were not acted on. That the retraction was framed partly around downstream misreporting, rather than a finding that the underlying data or analysis was fabricated or falsified, is itself notable: it places the case closer to the harder editorial judgment calls CASRAI’s retraction statement and retraction vs. correction pages describe than to the more clear-cut misconduct cases NISO’s CREC recommended practice was built around.
The Senate hearing in the middle of the timeline
Between the June 2024 expression of concern and the August 2026 retraction, the paper was cited in a U.S. Senate Homeland Security and Governmental Affairs Committee subcommittee hearing in June 2026 examining alleged mechanisms of harm from COVID-19 vaccination and, separately, alleged pressure on researchers publishing findings critical of the vaccines. At the time of that hearing, the paper carried a published expression of concern but had not been retracted — meaning a document the journal had already flagged as under active question was cited in a legislative proceeding as though its findings stood, with no requirement that the hearing record reflect the outstanding editorial concern.
This is not a claim that the hearing or its participants acted in bad faith; expressions of concern are, by design, not verdicts, and CASRAI’s own expression of concern entry is explicit that the notice signals open questions rather than a finding of error. The point relevant to research-integrity practice is structural: once a paper draws policy or legislative attention, the editorial-review clock and the political-attention clock run independently, and there is no standard mechanism that pauses citation or use of a paper in the second while the first is still running.
Timeline
| Date | Event |
|---|---|
| May 6, 2024 | BMJ Public Health publishes Mostert et al.’s paper on excess mortality and COVID-19 vaccination. |
| June 2024 | BMJ Public Health places a formal expression of concern on the article, roughly a month after publication. |
| June 2026 | The paper is cited in a U.S. Senate Homeland Security and Governmental Affairs Committee subcommittee hearing on alleged COVID-19 vaccine harms, while the expression of concern remains in place and unresolved. |
| August 11, 2026 | BMJ Public Health formally retracts the paper, citing widespread misreporting/misunderstanding of its findings as implying vaccination causes mortality, and inadequate description of the study’s limitations. |
Why the two-year gap matters for research-integrity practice
A roughly two-year span between an expression of concern and a retraction is not, on its own, unusual — investigations that involve outside data requests, author responses, institutional review, and legal caution routinely run long, and COPE guidance explicitly counsels against rushing an investigation’s conclusion for the sake of speed. What this case adds is a concrete illustration of the cost of that duration when the paper in question has already left the scholarly literature’s normal audience: for roughly two years, a paper the journal itself had already flagged as containing unresolved problems remained citable, and was in fact cited, in a forum with no obligation to track or disclose its editorial status.
For editorial offices and research-integrity officers, the practical takeaways are narrow but real:
- An expression of concern is a live signal, not a closed matter — and its persistence for an extended period does not mean the underlying question has gone away or been resolved in the author’s favor.
- Downstream users of a paper (journalists, policy staff, expert witnesses, legislative committees) have no standard way to be alerted that a cited paper carries an active editorial notice unless they check the publisher’s site directly at the time of use; this is exactly the gap that CrossMark status indicators and clear Retraction Watch-style tracking exist to narrow, though neither guarantees a citer will look.
- When a journal’s stated retraction rationale centers on misinterpretation and inadequate caveats rather than fabricated data, that is a different, and in some ways harder, editorial judgment than a fraud finding — and it is worth research-integrity offices tracking as its own category, since it turns partly on how a paper was written and framed, not only on what it found.
What is, and isn’t, established
- Established: BMJ Public Health published the Mostert et al. paper on May 6, 2024, placed an expression of concern on it roughly a month later, and issued a formal retraction on August 11, 2026, citing widespread misreporting/misunderstanding and inadequate description of the study’s limitations. This is confirmed via BMJ Group’s own retraction notice and Retraction Watch’s reporting.
- Established: the paper was cited in a June 2026 U.S. Senate Homeland Security and Governmental Affairs Committee subcommittee hearing on alleged COVID-19 vaccine harms.
- Reported, not independently verified by CASRAI beyond secondary reporting: the precise content of Mostert’s disagreement with the retraction and her account of proposed revisions being ignored, which we have only through reporting rather than a primary statement we independently obtained.
Frequently asked questions
What did BMJ Public Health give as its reason for retracting the paper?
The journal cited widespread misreporting and misunderstanding of the paper — particularly its use to claim a direct causal link between COVID-19 vaccination and mortality — along with inadequate description of the study’s limitations, rather than a finding of fabricated or falsified data.
How long was the paper under an expression of concern before it was retracted?
Roughly two years: BMJ Public Health placed the expression of concern about a month after the May 6, 2024 publication, and the retraction followed on August 11, 2026.
Was the paper actually used in U.S. policy discussions before it was retracted?
Yes. It was cited in a June 2026 U.S. Senate Homeland Security and Governmental Affairs Committee subcommittee hearing examining alleged COVID-19 vaccine harms, at a point when the paper carried an active, unresolved expression of concern.
Does an expression of concern mean a paper is wrong?
No. It signals that the journal has open questions it has not yet resolved, not a finding of error or misconduct. See CASRAI’s expression of concern entry for how the notice differs from a retraction.
Is this the same case as the HHS letter to Sage Open Medicine over a different vaccine-related expression of concern?
No. That case, covered separately by CASRAI, involves a 2020 Sage Open Medicine paper by different authors, a different journal, and a different U.S. official pressing for more disclosure about the notice itself. This BMJ Public Health case involves an already-completed retraction of a different paper, with the Senate connection arising from a citation in testimony rather than a direct inquiry to the journal.
