Case Study 39.2 — The Enthusiastic Clinic

What this is. A reader is offered something. The person offering it is warm, credentialed, unhurried, and confident. Nothing about the encounter feels like the one in §39.7 — it feels like the opposite, which is precisely why it is worth working through slowly.

This case study runs §39.4's five questions plus §39.8's structural question against a practice that recommends and dispenses. The conclusion is neither credulous nor cynical, and arriving at that conclusion is the exercise.

Everyone here is a role: the reader, the clinic's physician, the reader's regular clinician, a pharmacist.

The encounter

A reader in their fifties, generally well, has been sleeping poorly and has noticed the ordinary thickening around the middle and loss of strength that accompanies a decade. They see an advertisement for a clinic offering "peptide therapy for metabolic health, recovery, and healthy aging," book a consultation, and go.

The consultation lasts fifty minutes. This is the first thing worth noticing, and it is not a criticism of anyone. Fifty minutes is a genuinely different experience from fourteen, and the reader leaves feeling listened to in a way they have not felt in a clinical setting for years. That feeling is real, it is valuable, and it is not evidence about a molecule.

The physician is a licensed doctor. They take a thorough history, order a broad laboratory panel, and at the follow-up present the results with a clear narrative: several markers are "suboptimal, though technically within range," and there is a protocol they have had good results with. It involves a growth-hormone-releasing peptide combination and a second compound described as supporting tissue repair. The clinic supplies both from its own pharmacy. There is a monthly membership that includes follow-up visits and quarterly repeat panels.

The physician is not pushy. They say the reader should think about it. They answer questions patiently. They mention that they take one of the compounds themselves.

Everything about this encounter is more pleasant than every clinical encounter the reader has had in a decade. Hold that thought; it is the thing the analysis has to work around.

Question 1 — "What's the evidence for this in someone like me?"

The reader asks. The answer is thoughtful and comes in three parts.

The physician cites the physiology: the compounds stimulate a well-characterized endocrine axis, the mechanism is understood, and age-related decline in that axis is a real and documented phenomenon. They cite short-term studies showing that the compounds do what they are said to do at the level of hormone measurement. And they cite their own experience: several hundred patients, and they will tell you frankly that most of them report feeling better.

How to read this. Each of the three parts is a different kind of thing, and they should not be weighed together.

The physiology is Chapter 2's territory and this book's most-repeated warning: knowing how something would work is not evidence that it does. Mechanism does not upgrade a rating — that is rule 3 of the system, and it exists because mechanism is the most persuasive-sounding evidence that is not evidence.

The short-term hormone studies are real and are answering a narrower question than the reader asked. They establish that the compounds move a hormone level. That is a surrogate endpoint, and question 2 is about to make that explicit.

The clinical experience is Chapter 6's territory — valuable as a source of hypotheses, weak as evidence of effect, and subject to every selection and expectation mechanism that chapter catalogues. It is not worthless. It is not what the reader asked for.

The honest summary of the answer to question 1: there is no adequately powered randomized trial in a population like this reader, for the outcomes this reader cares about. The physician did not claim there was. They answered a different question well.

Question 2 — "What's the endpoint — what would actually get better?"

This is the most productive question of the five in this setting, and the reader has to press gently to get a specific answer.

The first answer is "you should feel better — more energy, better recovery, better sleep, better body composition." Those are outcomes and they are what the reader wants.

The follow-up: how would we know? The answer is the quarterly panel. The measured thing is a hormone level and a small set of related markers.

Notice what has happened. The promised benefit is a set of outcomes. The measured benefit is a set of surrogates. Chapter 16 spent a chapter on why those come apart, and there is a structural reason a practice like this works in surrogates: a surrogate is measurable at every visit and an outcome is not. A number that moves gives everyone — the physician and the patient equally — something to point at every quarter. This is not a trick. It is the natural gravity of a model built on regular follow-up, and it operates whether or not anyone intends it.

The reader's useful move here is to write down, in advance, what "feeling better" would look like concretely: hours slept, times woken, a specific physical task, a specific measurement. That converts the promised outcome into something that can fail, which is question 4's job.

Question 3 — "What are the alternatives, including doing nothing?"

The reader asks. The physician mentions sleep hygiene and resistance training, briefly and genuinely, and returns to the protocol.

This is the answer worth pausing on. It is not evasive and it is not wrong. But notice that the alternatives were mentioned as adjuncts rather than as comparators — the question would resistance training alone, done properly for six months, produce this? was not engaged, and the answer to it is not obvious.

Doing nothing was not on the list at all. For a well person in their fifties with poor sleep and ordinary age-related changes, "do nothing about the compounds and address the sleep directly, then reassess in six months" is a real option with a real evidence base behind parts of it. A recommendation that never mentions waiting has not been weighed against waiting.

The reader should also notice a cost comparison nobody raised. The membership plus compounds runs to several hundred dollars a month — call it \$400 — which is roughly \$4,800 a year. The alternatives cost approximately nothing. That does not settle anything either, but it belongs in the comparison, and the comparison is what question 3 exists to produce.

Question 4 — "What would you expect to see, and by when?"

The physician answers this one well: most people notice sleep and recovery changes within six to eight weeks, and body composition over three to six months.

This is a good answer and the reader should write it down verbatim, with the date. It is a falsifiable prediction. It converts an open-ended arrangement into something that can be evaluated.

The reader should then add their own specification, since the physician's version is still soft: sleep means what, measured how? The reader decides on two things — times woken per night, averaged over a week, and whether they can complete a specific physical task they currently cannot. Both recorded now, before anything starts.

Question 5 — "What would make you stop this?"

The reader asks. The answer is the least satisfying of the five: adverse effects would stop it, and otherwise the protocol is a long-term one — this is maintenance, not a course.

Sit with that. An intervention that is maintenance by design has no natural end, and therefore no point at which the honest question is this doing anything? is forced. Question 5's whole purpose is to install that point, and here it did not get installed.

This is the single most informative answer the reader received, and it did not come from suspicion or confrontation. It came from asking a standard question and listening to the answer.

The reader can install the stopping rule themselves, and should: if the two specified measures have not moved by the date the physician named, I stop. Written down. Dated. Set by the version of them that is still neutral.

The structural question — "does this practice sell what it is recommending?"

The reader does not need to ask anyone. The answer was on the website: the clinic dispenses from its own pharmacy, operates a monthly membership, and includes quarterly panels in the fee.

Three structural facts. All knowable. None an accusation.

  WHAT THE STRUCTURE TELLS YOU        WHAT IT DOES NOT TELL YOU

  A recommendation to continue is     Whether the physician believes what
  also a recommendation to keep       they are saying. (Most do.)
  paying.
                                      Whether the compounds work.
  The measured endpoints are the      (Unaffected by who sells them.)
  ones generating the follow-up
  schedule.                           Whether this physician is dishonest.
                                      (Nothing here is evidence of that.)
  There is no internal party whose
  interest is served by the reader    Whether dispensing practices are
  concluding "this isn't working."    illegitimate. (They are not, and many
                                      dispense for sound reasons.)

Chapter 42's conclusion applies exactly: a conflict of interest is a reason to check the evidence, not evidence of dishonesty. The reader who converts this into "that clinic is a scam" has made an error in the opposite direction from credulity and has learned nothing usable, because the cynical reading and the credulous reading are both substitutes for checking.

And the deeper point, from §39.8 and Chapter 38 §38.10: "a doctor recommended it" is not evidence about a molecule. The physician's recommendation is a clinical judgment. So was the reader's regular clinician's reluctance. Neither is data. A reader who discounts a stranger's testimonial but not a physician's recommendation has learned half the lesson, and it is the easy half.

Where the reader lands

Not "no." Not "yes." Something more useful than either.

What the reader now knows that they did not before the questions:

• The evidence for the specific outcomes they care about, in a population like theirs, does not exist
• The measured endpoint is a surrogate, and the promised endpoint is not the measured one
• Doing nothing was never compared against
• There is a date-stamped prediction, which they have written down
• There is no stopping rule in the protocol, so they have written their own
• The practice's structure means continuation is the default and nobody inside it is positioned to argue for stopping

What the reader does next, whichever way they decide:

They tell their regular clinician. All of it — the compounds, the panel, the plan. Not for permission and not for a second opinion on the protocol, but because §39.2 applies here exactly as much as it applies to something bought from an overseas supplier. A compound obtained from a licensed physician who dispenses it is still a compound their regular clinician's differential needs to contain.

They take the panel to a pharmacist and ask what the compounds interact with.

And whatever happens, they come back at the date they wrote down — to the enthusiastic clinic and to their regular clinician both — and say what actually happened. Including if it was nothing.

Discussion questions

1. The physician's answer to question 1 had three parts: mechanism, short-term hormone studies, and clinical experience. Rate the evidentiary weight of each, and explain why combining them into a single impression is the error rule 3 of the rating system is designed to prevent.

2. The case study argues that a practice built on regular follow-up has a natural gravity toward surrogate endpoints, and that this operates whether or not anyone intends it. Explain the mechanism. Then describe a follow-up structure that would resist the gravity, and say why it is harder to run.

3. Question 5 produced the most informative answer of the five. State what the answer was, why it was informative, and what it would have taken for the reader to notice this after six months of membership rather than before starting. Refer to Chapter 6.

4. The reader converts the physician's soft prediction ("sleep and recovery within six to eight weeks") into two specific measures recorded in advance. Write the equivalent conversion for a different promised benefit of your choosing. What makes a measure suitable for this purpose, and what makes one unsuitable?

5. Write two verdicts on this clinic — one credulous, one cynical — in a sentence each. Then write the verdict this case study actually reaches. Explain what the third one can do that neither of the first two can, in terms of what the reader would do differently tomorrow.

6. The reader's regular clinician was reluctant; the clinic's physician was enthusiastic. §39.8 says neither is evidence about a molecule. Does that mean the two are equally uninformative? Construct the strongest case that one of them should carry more weight, then evaluate whether that case survives the chapter's argument.