Discussion Guide — Chapter 39

Six prompts, all answerable without any student describing their own use.

1. Argue the other chair (§39.1)

Half the room describes the appointment from the clinician's side: what are they braced for, what are they triaging against, what does "I'm not comfortable with that" actually mean? The other half says what the patient heard.

What to listen for. Attribution to structure: eight minutes, a base rate of unsupportable requests, an incomplete chart, documentation demands, the professional cost of "I don't know." They should reach the four-meanings point — the evidence is bad, I don't know it and won't pretend, my institution won't let me, I'd need forty minutes — and see that asking which costs nothing.

Weak responses attribute everything to personality, or make one side the villain. Watch too for the student who calls the clinician burned out and stops there — sympathy, not analysis, and it yields no move. Redirect: what can the patient do differently, given fixed constraints?

2. Why clinical framing, not moral (§39.2)

The chapter argues for disclosure on clinical grounds and explicitly refuses the moral argument. Why? What does the moral framing do to the behavior it is trying to produce?

What to listen for. The mechanism, not the sentiment: a person who experiences disclosure as a confession will minimize, and minimizing produces exactly the failures disclosure was for. They should name the four reasons — differential diagnosis, interactions, peri-operative safety, monitoring — and see that each takes the medication history as an input and fails silently when it is incomplete.

On the peri-operative case, listen for precision: GLP-1 receptor agonists slow gastric emptying as part of the mechanism, so a correctly fasted patient may still have retained contents, raising aspiration risk under sedation. The move that matters is that the patient does not need to know current guidance — the anesthesiologist needs to know about the exposure.

Red flag to correct on the spot: any student reciting a hold interval or protocol. The chapter gives none deliberately, and guidance has moved toward individualized assessment.

Weak responses reduce to "honesty is the best policy" — the framing the section rejects.

3. Take the five questions somewhere else (§39.4)

Assign each group a claim with nothing to do with peptides — a screening test, a knee operation, a widely used supplement, a psychiatric medication. Run all five. Report which was most uncomfortable to ask.

What to listen for. Whether question 4 produces a real falsifiable prediction — a magnitude and a date — rather than a restated hope. Whether "including doing nothing" is treated as a genuine arm, connected to natural history and regression toward the mean. Whether question 5 is understood as something to ask before starting, since it is nearly impossible to ask honestly once invested.

Listen for the transfer: these are medical-claim questions, not peptide questions. A group that cannot run them on a knee operation has not understood them. Common weak move: collapsing question 1 into "does it work for me," dropping the population reasoning entirely.

4. Which sentence did you just hear? (§39.6)

Give three responses in a clinician's voice and have students classify them: "we don't know," "it doesn't work," "I don't know." What does each license you to conclude, and what is the next question?

What to listen for. That "it doesn't work" is the strongest and rarest — it asserts evidence exists and is negative — and that the follow-up is which trial? That "I don't know" says nothing about the state of the field. That "we don't know" is a real and common state of the world, and that saying it costs the clinician something.

The best responses connect this to the rating system: a ❌ is a statement about the evidence, not a verdict on the molecule, and never-tested and tested-and-failed are different situations with different futures.

Weak responses treat all three as versions of "no." Watch for the student who hears "we don't know" as a brush-off — the misreading the section exists to fix.

5. The enthusiastic clinician (§39.8)

A clinician is warm, unhurried, interested, and recommends the compound. Your last one brushed you off. What do you do differently — and what do you do the same?

What to listen for. Symmetry: the five questions get asked identically, and interrogating a no while relaxing at a yes is motivated reasoning wearing skepticism's clothes. Listen for students who can say why enthusiasm is cheaper to produce than restraint.

Then the structural question — does this practice sell what it recommends? Strong responses treat dispensing, membership models, in-house lines, and purchase-linked follow-up intervals as knowable structural facts, not accusations, and land on the chapter's conclusion: a conflict of interest raises the value of checking and settles nothing.

Two failure directions. Cynicism ("dispensing doctors are in it for the money") is the more common and needs immediate correction; practices dispense for sound reasons. Credulity is quieter — the student who cannot say why a recommendation is not evidence about a molecule has learned only the easy half. Ask directly: what makes a white coat different from a forum post, and what does it not change?

6. What supervision adds, and why report back (§39.9, §39.10)

Why is "medical supervision reduces the risk of using an unapproved compound" rated ⚠️ rather than ✅ or ❌? Separately: why does a clinician's impression of a compound drift toward optimism, and what one sentence fixes a piece of it?

What to listen for. On the rating: supervision adds baseline and interval monitoring, interaction checking, and a response pathway — the Chapter 19 §19.7 list — while adding no evidence and unable to detect harms nobody knows to look for. The ⚠️ marks a sound direction of unquantified magnitude. Listen for the guard clause: ⚠️ rates what supervision adds and is not an endorsement of the underlying use. A student who drops it has converted a rating into permission.

On reporting back: the selection mechanism. Patients who improve say so; patients who notice nothing quietly stop and never mention it; patients who stop for a side effect often just stop. The impression is assembled from a non-random sample biased toward success — Chapter 6's mechanism inside a clinic. Strong responses name the fix: come back and say what happened, including when nothing happened, including when you stopped.

Close on the chapter's own closing point: these one-patient-at-a-time observations are weak evidence and, for many of these compounds, the only evidence there will ever be.