Audio Debate / 23 minutes
FDA approval versus biological safety.
The debate starts where the draft starts: approval and accumulated human exposure are both evidence, but they are not evidence of the same thing.
I am preparing to make a cross-disciplinary assertion into medical policy. That sentence should make a person slow down. I am not a regulator or a physician, and I am not trying to become one by vocabulary. I am trying to take a distinction that is visible in the evidence, keep it inside its lawful limits, and ask whether current consent language sometimes collapses two different kinds of knowledge into one category.
The draft is called Approval Is Not Exposure. FDA approval establishes that a particular product met an applicable benefit-risk and quality standard for a defined use, population, dose, route, and evidence package. Accumulated human exposure establishes something else: how much opportunity common, rare, delayed, interaction-dependent, and population-specific effects have had to become visible over time.
I am deliberately trying not to describe the paper in Nested Causal Modelling terms. The method is underneath it, but the medical-policy claim should survive in ordinary evidentiary language. If it cannot, the claim is not ready to cross disciplines.
Draft-process archive: this field note and its feedback form preserve the public working stage. The revised paper is now a DOI publication; neither version is medical advice, a therapeutic recommendation, or a claim that historical use proves efficacy or safety.
Anonymous Draft Review
Keep the feedback on the claim.
Feedback is anonymous. This form does not request or store your name, email address, IP address, or user-agent string in the feedback record. Do not include identifying information in your answers. Anonymous feedback cannot receive a reply; use Contact or Subscribe if you want to stay in contact.
The draft does not argue that an unapproved treatment works. It does not turn historical use into proof of safety, and it does not weaken the value of an approval program. It argues that approval status and exposure history answer different questions, and that informed consent becomes less accurate when one is allowed to impersonate the other.
The proposed correction is a two-axis disclosure: evidence for the proposed therapeutic use on one axis, accumulated human-exposure evidence on the other. The reason evidence is missing should also be named. A failed trial, a safety failure, no sponsor, no investigation, and commercial abandonment are not the same state.
Feedback would be appreciated, if anyone out there is reading these posts. I am especially interested in where the evidence is insufficient, where the language crosses beyond the claim, and whether the disclosure standard would actually improve a clinical or policy decision. Praise is pleasant. A boundary I failed to see is more useful.