The switch usually happens for one reason, and it is a good one: the approved product costs several times what a compounded vial costs, and the person making the change has already tolerated the drug for months. The mental model that comes with it is that nothing changes except the price. Almost everything changes except the molecule’s name, and the part that changes hardest is the arithmetic that turns a prescription into an injection. The general comparison between the two markets is in the compounding article; this is about the day of the handover.
The reassuring sentence is on FDA’s warning list
FDA publishes a list of telehealth red flags for consumers buying these drugs. The first item on it is a company that makes claims such as the compounded drug being the same as an FDA-approved drug. The second is a price that seems too good to be true.[1]
That is an uncomfortable pairing, because those two sentences are together the entire pitch for switching. The agency’s reasoning is not rhetorical: compounded drugs are not FDA-approved, which means the agency does not review them for safety, effectiveness or quality before they are marketed.[1] Sameness is a conclusion that requires evidence, and the review that would generate that evidence has not happened.
Since the shortages ended, a copy is the one thing it cannot be
The legal ground shifted underneath this market. While semaglutide and tirzepatide were on the shortage list, compounders could make what were effectively copies. With the shortages resolved, that route closed, and what remains lawful is a product that differs from the approved one.
A 2026 survey searched compounder websites between February and March 2025 and recorded what the market had become. It identified 33 distinct semaglutide or tirzepatide products. Two-thirds contained semaglutide and one-third tirzepatide. Forty-eight percent combined the peptide with some mixture of cyanocobalamin, glycine, niacinamide, docusate or ondansetron. Of the 17 single-ingredient products, 82% were sublingual and 18% were orally disintegrating tablets — neither of which is a route any approved product uses.[2]
Its assessment of those additions is blunt. The authors found little justification for adding nutrients or docusate sodium to an injectable peptide, note that a rationale exists for ondansetron but that evidence for giving it subcutaneously is lacking, and that whether a sublingual or disintegrating formulation offers any advantage over the approved oral tablet has not been determined.[2]
So the buyer is usually switching to something structurally different from what they were taking, and the difference is the reason it is on the market at all. Whether a particular vial is a lawful preparation for a particular patient is a question for the prescriber and the pharmacy, and the regulatory split that governs it is in the 503A and 503B article.
The milligram number survives. The way it is measured does not.
A pen resolves the dose by construction. The number on the device is the number delivered, and there is nothing to calculate. A vial resolves the dose by volume, and volume only becomes a dose once the concentration is known.
FDA states the problem in its own words: compounders offer these products in various containers, product concentrations may vary depending on the compounder, and a single compounder may offer multiple concentrations. The instructions that accompany the drug, where instructions are provided, may direct administration in “units” — the volume of which depends on the concentration.[3]
That last clause is the whole hazard in one line. A unit is a mark on a barrel, not a quantity of drug. Two vials labeled with the same number of units can contain different amounts of peptide, and nothing on the outside of either announces it. Anyone arriving from a pen has never had to think about this, because a pen never asked them to.
The approved vial shows how little a label commits to here
There is one FDA-reviewed vial in this class, and it is instructive. The Zepbound and Mounjaro single-dose vials are supplied at a fixed strength in 0.5 mL, and the multi-dose vial holds 2.4 mL providing four doses of 0.6 mL each. The administration instruction is to use a syringe appropriate for the dose, with the example given as a 1 mL syringe capable of measuring 0.5 mL or 0.6 mL.[4]
Even there, the label specifies a volume rather than a device, because volume is what a vial dose is. The difference is that the concentration behind that volume is fixed, published and reviewed. Strip that out and the same instruction becomes unresolvable. Which syringe belongs to which preparation is the subject of the needle article.
What the error record shows, and what it cannot
This is documented rather than hypothetical. A disproportionality analysis of the FDA Adverse Event Reporting System from the fourth quarter of 2003 through the first quarter of 2024 identified 3,348 reports of accidental overdose involving GLP-1 receptor agonists, with reporting odds ratios significant for every agent in the class, ranging from 2.64 to 61.12 against the comparator, all at P < 0.008.[5] The authors attribute the pattern to access pressure pushing patients toward online and compounding channels, and state directly that a pharmacovigilance database can establish association and not causation.
A regional poison-center review adds the shape of the curve. Analyzing 1,047 human exposure cases reported to a statewide system between December 2017 and December 2023, an interrupted time-series found reported exposures rising by 1.16 per month (95% CI 0.57 to 1.80; P < 0.001) and hospital utilization from those exposures rising by 0.351 per month (95% CI 0.159 to 0.544; P = 0.001) after the weight-management approval. Most cases were managed at home: 66.5%, with 21.0% treated in an emergency department and 4.4% admitted.[6]
Two readings follow, and they cut against each other. Errors of this kind are common and rising, and the usual outcome is several unpleasant days rather than an emergency. What neither dataset can see is the opposite error. A dose drawn ten times too small produces no symptom, no report and no call — only a patient who concludes after two months that the drug has stopped working, which is exactly what a plateau looks like from the inside.
Salt forms are a different active ingredient
One substitution happens at the chemistry level rather than the dosing level. FDA states that some semaglutide products sold by compounders may be salt forms — semaglutide sodium and semaglutide acetate — and that these are different active ingredients than are used in the approved drugs. The agency adds that it does not have information on whether those salts share the chemical and pharmacologic properties of the approved active ingredient, and that it is not aware of any lawful basis for their use in compounding.[1]
A patient cannot detect this from a vial. It is a question the pharmacy can answer and the buyer cannot, which puts it on the short list of things worth asking in writing before the first shipment — alongside the questions in the vetting article.
Two laboratories have opened the vials
Both analytical studies published so far were run by the originator manufacturers, and that belongs in the reading of them.
A 2026 analysis of follow-on and compounded semaglutide and liraglutide samples, authored by Novo Nordisk employees and shareholders, reported impurity profiles distinct from the originators — amino acid deletions and additions plus unidentified impurities — and significant disparity in strength, impurity sum and high-molecular-weight protein level between compounded semaglutide and originator product on light exposure. Using an antigen-presentation assay, it found potentially immunogenic peptides presented on cells stimulated with those impurities.[7]
A 2026 analysis by Eli Lilly employees tested compounded tirzepatide combined with B12 analogs, the combination category the survey above found in nearly half the market. It identified a previously unreported impurity arising from a chemical reaction between tirzepatide and certain B12 analogs, present at substantial levels, and states plainly that the clinical effects of that impurity are unknown.[8]
Manufacturer authorship is a real conflict and does not dissolve a mass spectrum. What both papers establish is narrower than their framing: the contents differ from the approved product in ways that were not anticipated, in categories of product sold at scale, and no clinical consequence has been measured either way. A broader review of follow-on and compounded manufacturing reached the same structural conclusion — that process and sourcing drive the differences.[9]
The list of things a buyer cannot check
Potency, purity and sterility are the obvious three, none of which is visible or inferable from a shipment. Beyond that, the survey of 33 products found that only 18% published any information about a beyond-use date and 28% stated preferred storage conditions.[2]
That matters more for a multi-dose vial than for a pen. FDA recommends discarding a multi-dose vial of sterile medication within 28 days after first use because of contamination risk, and says explicitly not to keep using it after 28 days even if the compounder’s instructions say the medication can be used for longer.[1] A recommendation phrased to override the dispenser’s own paperwork is a recommendation written after somebody read the paperwork. How the vial travels before it gets there is covered in the cold-chain article.
The surveillance record is thin for a structural reason rather than a reassuring one. As of May 31, 2026 the agency had received 990 adverse event reports associated with compounded semaglutide and more than 730 associated with compounded tirzepatide — while noting that federal law does not require state-licensed pharmacies that are not outsourcing facilities to submit adverse events at all, so events from compounded versions are likely underreported.[1] A count drawn from a channel with no mandatory reporting is a floor, not a rate.
The class has exactly one labeled switching instruction
It is worth seeing what a switch looks like when somebody has actually studied it. The Wegovy label covers moving between its own two presentations, and the instruction is specific in both directions: one week after discontinuing the 2.4 mg injection, initiate the 25 mg tablet once daily; and the day after discontinuing the 25 mg tablet, initiate the 2.4 mg injection once weekly.[10]
A week in one direction, a day in the other, between two forms of one molecule from one manufacturer. That asymmetry exists because somebody measured it. No equivalent sentence exists for moving from an approved pen to a compounded vial, in either direction, and the absence is not an oversight — it is what an unreviewed product means. The timing rules that do have labeled answers are in the dose-timing article.
None of this argues that the switch is wrong for any particular person. Cost is a clinical variable, and a drug nobody can afford has an effectiveness of zero. What it argues is that the switch is a change of product rather than a change of vendor, that the questions it raises have answers held by the prescriber and the pharmacy rather than by a comparison page, and that the one claim most likely to be offered as reassurance is the claim the regulator lists first among its warning signs. What each seller discloses about its own preparation is recorded in the provider write-ups.