The pattern
In 1995, Purdue submitted trial data showing OxyContin had less abuse potential. The FDA approved it. Doctors wrote 80 million prescriptions. By the time Merck pulled Vioxx, the pattern was established: dramatic results, media coverage, public demand, political pressure, regulatory shortcuts, widespread use, adverse effects the small trials were not powered to detect.
We sat on the advisory committee that rejected MDMA. We voted no. We would vote no again.
Psilocybin’s Phase 2 data is extraordinary. We do not dispute it. But the trials excluded patients with psychotic spectrum disorders, bipolar I, family histories of schizophrenia, active suicidality — a substantial portion of the treatment-resistant population the clinical advocates cite as moral urgency.
The blinding problem
Functional unblinding is not a nuisance. When a patient takes psilocybin, they know. The experience is unmistakable — visual distortions, ego dissolution, emotional flooding. The placebo group knows they got placebo. In a depression trial, expectation is a competing explanation.
The MDMA trials had boundary violations — a therapist who had physical contact with a patient during a session. MDMA produces profound trust and vulnerability. A drug that dissolves psychological defenses administered in an intimate dyadic setting with a power differential creates structural risk the clinical advocates treat as isolated incidents.
The tradition holders see what we see from a different angle: the compound without adequate infrastructure is dangerous.
Where we concede ground: Our standard — the double-blind RCT — may be structurally unsuited to evaluating psychedelic therapy.
What would change our mind: Oregon’s five-year real-world data showing adverse rates comparable to existing psychiatric medications.
Read the full synthesis: Are psychedelics actually medicine?