by James Lyons-Weiler, PhD, Popular Rationalism, ©2025
(May 21, 2025) — “The FDA is now saying: Here are restricted groups we will likely approve your products for, now show it is effective and safe in those groups.”
When the FDA released its new framework for COVID-19 vaccine policy via a “Sounding Board” article in the New England Journal of Medicine, authored by Dr. Vinay Prasad and Dr. Marty Makary, it marked a sharp pivot away from the all-ages-for-all-times universal booster doctrine that has dominated U.S. policy for the last three years.
Predictably, the move was met with confusion and criticism from establishment voices, as reflected in Kristina Fiore’s recent article in MedPage Today. Critics such as Dr. Paul Offit and Dr. Kathryn Edwards expressed concern about ethical trial design, presumed efficacy, and procedural irregularities. Yet their critiques miss key elements of the FDA’s position—and ignore the deeper structural problems in how we’ve previously evaluated vaccine efficacy and safety.
Let’s unpack what Prasad and Makary are really doing here, and why some of the loudest complaints are aimed more at preserving a fragile narrative than at protecting public health.
The Shift to Risk-Based Policy: Long Overdue
For years, critics of the “vaccinate-everyone-every-six-months” model have argued that the evidence simply doesn’t support universal boosting—especially in young, healthy individuals. European regulators acknowledged this early on, shifting to risk-stratified strategies. The U.S., however, held the line, often citing population-level benefits despite rapidly waning immunity, immune imprinting, and evidence of negative efficacy post-infection in some cohorts.
Now, under the new FDA framework, COVID-19 vaccine approvals for low-risk individuals will require randomized controlled trials (RCTs) with inert placebos and long-term outcomes—something long demanded by independent scientists. Meanwhile, high-risk groups (e.g., the elderly, immunocompromised) will be eligible under immunogenicity-based approvals, with post-marketing safety studies still required.
This may seem like a compromise, but it’s far more radical than the critics let on.
“For all healthy persons — those with no risk factors for severe Covid-19 — between the ages of 6 months and 64 years, the FDA anticipates the need for randomized, controlled trial data evaluating clinical outcomes before Biologics License Applications can be granted. Insofar as possible, when approving a Covid-19 vaccine for high-risk groups, the FDA will encourage manufacturers to conduct randomized, controlled trials in the population of healthy adults as part of their postmarketing commitment.”
Can you imagine FDA saying that under Marks? I cannot.
Immunogenicity ≠ Effectiveness — So Why Approve Based on It?
Here, Prasad and Makary walk a line between pragmatic regulation and institutional continuity. Immunobridging is not the gold standard—it’s a regulatory workaround. They know this. Prasad, in particular, has been critical of surrogate endpoints in oncology and COVID alike.
But by pairing immunogenicity-based approvals with mandatory postmarketing studies and demanding RCTs in the 50–64-year-old group, they’ve made it clear that the burden of proof is shifting—finally—back toward outcomes.
This matters. It puts industry on notice: Show us the data, not just the antibody titers.
The Lyons-Weiler/Fenton Effect: Still Ignored, Still Relevant
Strikingly absent from both the FDA’s new framework and the MedPage Today coverage is any recognition of a core methodological failure that underlies perhaps every prior estimate of COVID-19 vaccine efficacy: the Lyons-Weiler/Fenton Effect, aka, the Case Counting Window Bias.
This bias occurs when vaccine trials or observational studies exclude cases in the immediate post-vaccination window—often 7 to 14 days—while assigning those same cases to the unvaccinated comparison group. The result? Grossly inflated efficacy, underestimated risk, and fundamentally flawed ethics calculations.
Prasad knows this. He discussed it publicly when Peter Doshi and colleagues published on it. Yet in his new role at the FDA, he hasn’t addressed it directly. That silence is troubling—but it doesn’t invalidate the broader direction of the new policy.
What it does mean is that none of the prior data on which vaccine mandates and universal recommendations were based should be taken at face value. If we ignore the counting window bias, we’re building on sand.
There must be a good strategic reason why the fundamental estimates of efficacy are being taken at face value; perhaps it’s easier to force new studies under new scrutiny. New rulers, new rules. People like Edwards and Offit find that frustation.
That has to be a good sign.
Read the rest here.