Chapter 12 — Key Takeaways
The one-sentence version
The molecule is settled and the society around it is not: a drug with excellent evidence sits inside a pricing structure, a supply chain, a legal exception, and an argument about what obesity is, and none of those four is decided by a trial.
Core claims
The word "price" hides four numbers. List (public, paid by almost nobody insured), net (confidential, what the plan pays after rebates), out-of-pocket (set by plan design), and cash (what the uninsured are quoted, historically close to list). Rebates lower net and out-of-pocket; they do nothing for the uninsured. Identify which number an argument is about before joining it.
The list price is driven by exclusivity, population size, and who is doing the buying — not by cost of manufacture. The unit price is a pharmaceutical problem; price × population is a fiscal one, and they need different arguments.
The public funded the physiology and pays again for the drug. Both the complaint and the answer are true. Basic science is not a drug, and the private development effort was real, expensive, and mostly failed. What form recognition of the public contribution should take is a policy question with no option free of failure modes.
Payers had five reasons, four of them real. Budget impact at population scale; indefinite obligation for chronic therapy; the wrong-pockets problem in a market where members churn; a genuinely bad history of withdrawn weight drugs; and a category framing that coded weight loss as cosmetic. Only the last was mostly prejudice — and it made the other four unnecessary to argue honestly.
The shortage was a factory problem, not a chemistry problem. The binding constraint was aseptic fill-finish capacity and injector-pen assembly. Those are buildings, validation campaigns, and line-by-line regulatory approvals. A capital announcement is not capacity.
Compounding is legitimate; compounded copies of approved drugs generally are not. The exception is the FDA shortage list. Semaglutide was listed, an industry formed inside the exception, and the permission narrowed when the shortage resolved — amid litigation and transition periods. Compounded products are not FDA-approved and are not reviewed for safety or efficacy.
503A ≠ 503B, and neither is approved. 503A: patient-specific, state boards. 503B: larger batches, no patient-specific prescription required, CGMP, FDA inspection. Inspected and approved are different words for different things.
Off-label is legal, common, often the standard of care — and it moves the evidentiary burden onto the prescriber. It is not the same as unapproved. The label's closing clause, "as an adjunct to a reduced-calorie diet and increased physical activity," is the trial protocol written into the label, not a character test.
Bundled telehealth expanded genuine access and aligned revenue with prescribing at the same time. Both are true. Read the incentive by reading the interface: does the intake have a failure mode, can a person tell which answer qualifies them, is there follow-up that could result in stopping, is stopping as easy as starting.
Chapter 11's prediction, answered. GLP-1 agonists share insulin's ③ inelastic demand, ④ residual-payer structure, and ⑤ discontinuous coverage — but not ① acute lethality on interruption. Consequence: cost-driven discontinuation produces no deaths anyone can attribute and no legislation, while plausibly generating an aggregate loss of benefit across tens of millions that exceeds insulin rationing several times over. Large and illegible at the same time.
Regain is expected physiology, not treatment failure. The drug overrides an intact system defending a set point rather than replacing an absent hormone. That makes it chronic therapy. The expectation that a weight drug should produce permanent change after stopping is applied to no other chronic therapy in medicine, and it comes from a moral framework rather than a scientific one.
And the load-bearing logical point: a drug that works on a system is not evidence that the system caused the condition. Semaglutide's effectiveness shows GLP-1 signaling is a lever. It does not show what pulled it.
The four positions on obesity — stated, not adjudicated
| Position | Core claim | Strongest evidence | Most serious difficulty |
|---|---|---|---|
| Disease | Weight is regulated around a defended set point; susceptibility is heritable | Metabolic adaptation, monogenic obesity, response to pharmacotherapy | BMI is a crude marker; classification has interested parties |
| Behavior | Intake and expenditure are the proximate causes; medicalizing displaces agency | People do achieve durable change; self-efficacy predicts health behavior | Proximate ≠ ultimate; supplies the vocabulary of stigma |
| Food environment | Population-scale change in decades cannot be individual willpower | Genes did not change; the food supply did | Slow, mixed evidence base; under-explains within-household variation |
| Weight-neutral | Target health outcomes, not weight; weight-focused care carries its own harms | Fitness gains partly independent of weight; stigma and ED harms are real | In tension with SELECT's hard endpoint |
They are not mutually exclusive. An environment that changed, acting through behavior, on a population with biologically variable susceptibility, is a coherent single account containing all three — and heritability within a generation says nothing about what moved a population average.
Ratings issued
| Claim | Rating | Why |
|---|---|---|
| "Compounded semaglutide is equivalent to the branded product" (general claim) | ❌ | The category spans a registered 503B facility using genuine semaglutide base under inspected quality systems and material of unclear origin supplied as a salt form without the delivery device. No equivalence claim survives that range. Conceded: the legitimate end produces the same active molecule, and during a genuine shortage it solved a real problem for real patients. |
No new compound ratings were issued in this chapter. The rating above attaches to a claim about a production route, not to semaglutide.
Key terms
list price · net price · rebate · pharmacy benefit manager · formulary · prior authorization · step therapy · out-of-pocket cost · residual payer · monopsony · fill-finish · aseptic processing · drug shortage list · compounding · 503A pharmacy · 503B outsourcing facility · salt form · active pharmaceutical ingredient · off-label use · approved indication · direct-to-consumer telehealth · inverse care law · set point · metabolic adaptation · chronic therapy · relative risk reduction · absolute risk reduction · number needed to treat
What you can now evaluate
- Whether a cost claim is about list, net, out-of-pocket, or cash — and therefore whom it describes
- Whether a headline reporting a relative risk reduction has withheld the number you actually need
- Whether a compounded product is at the regulated or the unverifiable end of an enormous range, using six questions whose answers should be documents
- Whether an off-label recommendation sits at the "published trials" end of the spectrum or the "the clinic sells it" end
- Whether a prescribing pathway is an assessment or a sorting mechanism, by reading its interface
- Whether an argument about obesity is making a claim about physiology or a claim about desert
- Whether an "absence of a safety signal" is reassurance or an artifact of an exclusion criterion
Dossier progress
Field 10 (Status) — regulatory status, approved indications, availability, cost structure, who is selling it, what is not known, date checked, and where to re-check. Demonstrated in the chapter on compounded semaglutide and contrasted with the branded product. This is the field that goes stale fastest; an undated Field 10 is not information.
Next
Chapter 13 — Weight-Loss Peptides Beyond GLP-1. Amylin analogs, the combination products, and the far larger market of things promising to raise your own GLP-1 without a prescription. The question shifts from which molecule works to which claim is being made, and by whom. Everything in Chapter 12 applies to those products too — and applies hardest where the marketing is loudest.