Field Note / 2026-07-27

A Volunteer to Exemplify the Problem

Fauci’s pandemic diary shows the distinction between approval and exposure becoming visible inside the room.

Senator Rand Paul has released Anthony Fauci's pandemic diary, and Dr. Fauci has volunteered himself to provide the example my last paper was missing.

Open Published Paper Read Draft 2 in HTML Read the Foundational Draft

The paper is called Approval Is Not Exposure. Its argument is simple. FDA approval and accumulated human exposure are both evidence, but they are not evidence of the same thing.

Approval tells us that a particular product completed a particular regulatory process for a particular use. Exposure tells us how many opportunities a substance has had to surprise us after entering actual human bodies.

Those categories overlap. They do not become identical merely because people are accustomed to saying "FDA-approved" when they mean "safe."

The mistake is everywhere. It is useful to drug companies, regulators, hospital systems, insurers, lawyers, reporters, and anyone else who needs a complicated medical decision compressed into two words. It is also a problem for informed consent. A person cannot weigh what is known and unknown when the institutional category used to describe the treatment hides the difference.

The examples in my paper explain the mistake. Fauci's diary shows it happening.

What Fauci knew on March 18

On March 18, 2020, the White House was considering making hydroxychloroquine and remdesivir broadly available. Fauci objected to treating the two drugs alike.

He followed the meeting with an email. Hydroxychloroquine, he wrote, was an approved drug with adverse effects but "tons of experience with it." Remdesivir was different. It was unapproved and investigational. Fauci wrote that he had "more concern about the potential toxicities" of remdesivir and noted that it had shown no clear benefit in the Ebola experience. His proposed course was equally plain: "We should consider leading with HC in a flexible distribution protocol" while accelerating the remdesivir trial.

That is Approval Is Not Exposure in Fauci's own handwriting.

He understood that the two candidates occupied different evidentiary positions. Hydroxychloroquine had extensive prior human exposure. Remdesivir had less. Neither had convincing randomized evidence for COVID-19 at that moment, but uncertainty about efficacy did not erase the difference in accumulated safety knowledge.

The next day's entry is where the problem appears.

The administration, Fauci wrote, was going to make the drugs broadly available through compassionate use. "I convinced them not to." Three days later, an online system intended to track the drugs and make them available through physicians was, in his words, a "Bad idea," acceptable to him only if it did not interfere with randomized trials.

This does not prove that hydroxychloroquine worked. It does not need to.

The question is why an emergency response facing a new lethal disease would contract the number of interventions available for monitored use instead of expanding the number available for rapid, disciplined comparison.

The indication nobody could already possess

At the beginning of COVID-19, no established medicine could already be FDA-approved for COVID-19. The disease had not previously existed.

"Not approved for COVID-19" was therefore the necessary starting condition of every known medicine. It was not evidence that all of those medicines had failed comparative evaluation. It was not evidence that their ordinary toxicities were unknown. It did not distinguish a failed candidate from an unstudied one, or an unsafe substance from a familiar substance without a sponsor willing to finance a new indication.

Yet that unavoidable absence became a public warning label.

Old medicines were discussed through the approval package they did not have. Novel proprietary products were discussed through the approval package their sponsors could produce. The regulatory outcome then appeared to describe the biology.

It did not.

It also described who owned the product, who could finance trials, who could prepare an application, who could recover the expense, and which candidate the government had organized itself to move through the channel.

Then the category error changed sides

The mRNA vaccines arrived with COVID-specific trial evidence and almost no accumulated human exposure. Their intended mechanism was to deliver messenger RNA instructing human cells to manufacture a modified SARS-CoV-2 spike antigen.

Spike protein has demonstrated cardiotoxic and neurotoxic biological potential. After mass deployment, surveillance established a causal association between mRNA vaccination and myocarditis, particularly in younger males. Research continues into the roles of vaccine-derived spike, immune activation, and lipid nanoparticles in particular injuries.

None of that means that approval supplied no evidence. It means approval could not supply time that had not elapsed or observations that had not yet occurred.

The approved product contained the larger horizon of exposure-derived unknowns. The unapproved uses involved substances with decades of human history. Public language commonly reversed those facts:

approved novel intervention = safe
        unapproved use of known medicine = dangerous

That was the category error.

Its importance does not depend on settling another endless argument about whether hydroxychloroquine, ivermectin, or vaccination was effective. Efficacy is its own evidentiary axis. The error occurred when efficacy, intrinsic toxicity, accumulated exposure, residual uncertainty, and regulatory status were collapsed into one approved-or-unapproved judgment.

Informed consent requires material information about benefits, risks, alternatives, and uncertainty. Approval can contribute to that conversation. It cannot replace it.

A truthful comparison in 2020 and 2021 would have sounded something like this:

This new intervention has COVID-specific trial evidence, but limited accumulated human exposure. Rare, delayed, and population-specific harms may emerge only after widespread use.

And:

This established medicine has extensive prior human exposure and known toxicities under its ordinary uses. Its effectiveness, dose, timing, and benefit-risk relationship for COVID-19 remain uncertain.

Those two statements can coexist. Neither tells a patient what decision to make. Together they tell the patient what kind of decision is actually being made.

Instead, approval was widely allowed to mean that the novel intervention's safety question had been answered. Nonapproval was allowed to mean that the known medicine was outside responsible consideration. The first hid residual uncertainty. The second hid accumulated knowledge.

Then institutions went further and required the approved exposure as a condition of employment, education, military service, or participation in ordinary life.

The mistake had moved from language into bodies.

Keep the field open

The governing principle in a pandemic should be almost embarrassingly obvious: maximize the number of potentially useful interventions entering disciplined evaluation. Do not confuse making a treatment available with declaring it effective. Do not confuse testing it with endorsing it. Do not eliminate a known substance merely because no one owns enough of it to finance the regulatory identity demanded of it.

Run adaptive trials. Create monitored off-label protocols. Record doses, timing, interactions, outcomes, and adverse events. Reject candidates because evidence rejects them, not because a category prevented the evidence from being produced.

Fauci's diary matters because it shows that the relevant distinction was visible inside the room. On March 18, he knew that extensive experience with hydroxychloroquine and limited experience with remdesivir were materially different facts. The public did not consistently receive that distinction as part of its understanding of safety or its consent to the narrowing of alternatives.

The problem is larger than Fauci and larger than COVID-19. The same error appears whenever a newly approved proprietary treatment is assumed safer than an unapproved use of a familiar substance solely because one development pathway produced a regulatory stamp and the other did not.

That error is common because the systems enclosing it benefit from its simplicity.

It is dangerous because simplicity is not consent.

Approval is evidence. Exposure is evidence. A person deciding what may enter their body is entitled to know the difference.

Sources