Someone is told to take ten units. They own an insulin syringe, they can count to ten, and the instruction seems about as hard to get wrong as setting a microwave. Then the vial arrives at a different concentration than the last one and those same ten units deliver four times the drug, with nothing about the syringe looking any different.
This is not a hypothetical failure mode. In July 2024 the FDA issued an alert about adverse events involving compounded semaglutide in which patients had administered five to twenty times the intended dose, with some cases requiring hospitalization. The mechanism named in the alert was confusion between units, milliliters and milligrams.
Low dose protocols sit at the exact end of the range where this arithmetic bites hardest, because the smaller the intended dose, the larger a fixed measuring error becomes as a percentage of it. What follows covers what a concentration figure actually determines, why the delivery format changes the risk profile more than the molecule does, what shifted after the FDA declared the shortages over, and which numbers are worth confirming before anything is drawn.
Key Takeaways
- A unit on an insulin syringe is a volume, not a quantity of drug. Ten units is 0.1 mL on a standard U-100 syringe, and that same 0.1 mL carries 0.1 mg from a 1 mg/mL vial or 1.0 mg from a 10 mg/mL vial.
- The FDA alert of July 2024 described compounded semaglutide adverse events in which patients administered five to twenty times the intended dose, and identified unit and milliliter confusion as a mechanism. Some of those events led to hospitalization.
- US poison center calls involving semaglutide rose from a few hundred a year before 2021 to roughly 3,000 in 2023, a rise that tracks the expansion of both prescribed and compounded use.
- Small doses are the most error-sensitive doses. A half-unit misread costs about 4 percent of a 12 unit draw and about 25 percent of a 2 unit draw, which is an argument for choosing a concentration that spreads the intended dose across more of the barrel.
- Semaglutide has an elimination half-life of about one week, so plasma levels approach steady state over roughly four to five weeks. A dosing error made in week one can surface as escalating side effects a month later.
- The FDA declared the semaglutide shortage resolved in February 2025, following tirzepatide in late 2024, which removed the shortage-based justification for compounding copies of those products and triggered wind-down deadlines for 503A pharmacies and 503B outsourcing facilities during 2025.
What Does "Ten Units" Actually Deliver?
Milligrams contained in 10 insulin-syringe units, which is 0.1 mL, at four common vial concentrations
Direct arithmetic rather than measured data, calculated from the definition of a U-100 insulin syringe unit as 0.01 mL. The syringe looks identical in all four cases and the plunger sits in the same place. Only the label on the vial distinguishes a 0.1 mg dose from a 1.0 mg dose.
Why do units and milligrams keep getting confused?
Four points where the measurement chain breaks
The confusion is structural rather than a matter of carelessness. Insulin syringes were designed for a drug whose concentration is standardized by regulation, so the unit marking on the barrel maps reliably to a dose of insulin. Compounded GLP-1 vials carry no such standardization, and the same syringe placed against a non-standard concentration silently loses that meaning.
The syringe was built for a drug with a fixed concentration
U-100 insulin contains 100 units of insulin per milliliter by definition, which is why an insulin syringe can label volume markings as units and have them mean a dose. That equivalence holds only for insulin at that standard strength. Borrowing the same syringe for a drug whose concentration is set by whoever filled the vial keeps the markings and discards the meaning, and nothing on the barrel warns that this has happened.
Concentration varies between suppliers and between batches
Compounded GLP-1 preparations have circulated at concentrations spanning at least 1 mg/mL to 10 mg/mL, and the figure is a formulation choice rather than a fixed property of the drug. A person who has been drawing the same number of units for months can receive a refill at a different strength and continue the habit. This is the failure that produces a several-fold overdose without any change in behavior at the point of injection.
Instructions are frequently written in the wrong currency
A dose is prescribed in milligrams and executed in volume, so someone has to convert between the two, and that conversion needs a concentration to be possible at all. Written guidance that says a number of units without naming the vial strength it assumes is not a complete instruction. The FDA alert on compounded semaglutide specifically described patients unable to convert between units, milliliters and milligrams as a source of the errors it reported.
The syringe supplied is often larger than the measurement requires
Measuring 2 units in a 100 unit barrel places the entire dose inside the first 2 percent of the scale, where the graduations are closest together and the parallax error is proportionally largest. The FDA noted that some compounded products were dispensed with syringes larger than the volume the patient needed to draw. A smaller barrel does not change the dose and it substantially changes how accurately that dose can be measured.
The Same 0.25 mg Dose, Four Different Draws
Insulin-syringe units required to deliver 0.25 mg, by vial concentration
Arithmetic rather than measured data. Read this chart alongside the previous one, because they describe the same relationship from opposite ends. The practical point sits at the right-hand bar: a 10 mg/mL vial forces a 0.25 mg dose into two and a half units, a draw small enough that ordinary reading error becomes a meaningful fraction of the dose.
How does a compounded vial differ from a brand pen?
The delivery format, the regulatory pathway and the ingredient
A brand pen meters the dose mechanically, so the person injecting turns a dial to a labeled setting rather than measuring a volume by eye. That single design decision removes unit conversion, syringe selection and barrel reading from the process at once. A compounded vial hands all three back to the patient, which is the main reason documented dosing errors cluster around vials rather than pens.
The regulatory pathway differs as well. Approved products go through FDA review of manufacturing, potency and labeling before they reach anyone. Compounded preparations are not FDA-approved, and the two compounder categories are governed differently: 503A pharmacies compound for individually identified patients, while 503B outsourcing facilities register with the FDA and follow current good manufacturing practice requirements. Neither category involves premarket approval of the specific product being dispensed.
The ingredient itself can also differ. The FDA has stated that semaglutide sodium and semaglutide acetate are salt forms distinct from the semaglutide base used in approved products, and that these salt forms are not the approved active ingredient. A milligram figure on a vial describes the amount of whatever compound was weighed out, so equivalent numbers on two labels do not guarantee equivalent amounts of active drug. That is a different problem from measurement error and it sits underneath every calculation in this article.
Why Barrel Size Changes the Dose
Approximate proportional error from a single misread graduation when drawing a small dose, percent of the intended amount
Constructed estimates illustrating how graduation spacing scales with barrel size rather than measurements from a device accuracy study. The highlighted bar is the smallest barrel and the smallest error. Choosing a syringe sized close to the volume being drawn is one of the few dose accuracy improvements available that requires no change to the protocol itself.
Poison Center Calls Involving Semaglutide
Approximate annual exposure calls logged by US poison centers, thousands
Approximate figures reflecting the magnitude and trajectory reported by America's Poison Centers rather than exact published counts. Call volume tracks total exposure to the drug, so the rise reflects how many more people were using it and not a per-patient risk rate. The useful reading is the slope, which steepened in the same period compounded supply expanded.
What changed after the shortages were declared over?
The legal basis for compounding copies narrowed considerably
Compounding a preparation that is essentially a copy of an approved drug is permitted only in specific circumstances, and a drug being listed in shortage is one of them. That listing is what allowed compounded semaglutide and tirzepatide to be produced at scale during 2023 and 2024. Removing the shortage designation removed the justification.
Tirzepatide came off the FDA shortage list in late 2024, and the agency declared the semaglutide shortage resolved in February 2025. The FDA then set wind-down periods rather than an immediate stop, with 503A pharmacies and 503B outsourcing facilities given separate deadlines in the spring of 2025. Compounded product continued circulating past those dates through channels that operate outside the 503A and 503B pathways entirely, including research-labeled material sold without a prescription.
For anyone holding a vial, the practical consequence is that the source now matters more than it did, because the range of possible sources has widened rather than narrowed. A preparation from a registered 503B facility, a preparation from a 503A pharmacy and a vial labeled for research use are three different things with three different levels of oversight behind the number printed on the label. Working out which one is on the counter is a question for the prescribing clinician rather than a forum.
A Dosing Error Takes a Month to Fully Arrive
Approximate plasma semaglutide as a share of eventual steady state on an unchanged weekly dose, percent
Modeled accumulation curve for a drug with a one week elimination half-life rather than a measured pharmacokinetic dataset. The dashed line marks 95 percent of steady state, reached around week five. The consequence worth holding onto is that the first injection of a wrong dose feels manageable, and the same wrong dose four weeks later is delivering roughly three times the exposure.
Where the Delivered Dose Departs From the Intended One
Constructed index of the relative contribution of each factor to a mismatch between intent and delivery
Constructed index reflecting the mechanisms described in FDA and poison center communications rather than a quantified analysis of reported cases, since no such breakdown has been published. The ordering carries the information. The two factors at the top are both label comprehension problems, which means most of the risk sits before the syringe is ever picked up.
What should be confirmed before drawing a dose?
Six checks, none of which replace a prescriber
Everything below describes information worth having and questions worth asking. None of it is a method for setting a dose independently. GLP-1 receptor agonists are prescription medications, and the dose, the concentration and the syringe are decisions that belong with a prescribing clinician and a pharmacist who can see the full clinical picture.
The concentration is the number that converts a volume into a dose, and it is a formulation choice that can change between suppliers and between batches from the same supplier. A refill at a different strength turns a familiar routine into a several-fold error with nothing visibly different about the syringe. Reading the label at each refill rather than at the first one is the check that catches this.
A written instruction that names only units is incomplete because it silently assumes a concentration. An instruction that names the milligram dose, the vial concentration and the resulting volume can be checked for internal consistency by anyone, including a pharmacist who did not write it. The FDA alert on compounded semaglutide identified exactly this conversion as a source of reported overdoses.
A dose occupying the first few graduations of a 100 unit barrel is measured across the least readable part of the scale, and the same dose in a 30 unit barrel spreads over three times the distance. The FDA specifically noted compounded products dispensed with syringes larger than the volume required. Requesting an appropriately sized syringe changes the accuracy of every dose without changing the protocol.
The FDA has stated that semaglutide sodium and semaglutide acetate differ from the semaglutide base in approved products and are not the approved active ingredient. A label naming a salt form is describing a different compound from the one in the registration trials, so the milligram figure does not carry the same meaning. This is a question for the prescriber and the dispensing pharmacy, not something resolvable by inspection.
A 503B outsourcing facility registers with the FDA and operates under current good manufacturing practice requirements, a 503A pharmacy compounds for individually identified patients under different rules, and research-labeled material sold without a prescription sits outside both. These are three different levels of assurance behind the same printed number. After the shortage designations ended in 2024 and 2025, the mix of what circulates has become more varied rather than less.
Semaglutide accumulates toward steady state over roughly four to five weeks because of its one week half-life, so an error made early presents as escalating symptoms later with no obvious trigger. Severe or persistent vomiting, abdominal pain and dehydration are the presentations that brought people in the FDA reports to hospital. In the United States, Poison Control is reachable at 1-800-222-1222 and emergency services for anything severe.
What is genuinely unresolved here?
Three gaps worth naming
The true rate of dosing errors is unknown. Poison center calls and FDA adverse event reports are passive surveillance systems that capture events serious enough for someone to report, so they establish that a problem exists without measuring how often it occurs. A person who drew twice the intended dose, felt unusually nauseated for a week and never told anyone appears in neither dataset. The published counts are a floor rather than an estimate.
Potency and content of compounded preparations have not been systematically characterized. Independent testing of compounded GLP-1 products has been limited and has reported variable findings, and no comprehensive survey of what is actually in circulation has been published. That gap makes the milligram figure on a compounded label a claim rather than a verified measurement, which is a different kind of uncertainty from the measuring error this article has mostly been about.
Reduced dose protocols have not been studied as protocols. The doses below the approved starting range have no trial data behind them describing efficacy, safety or the appropriate way to prepare and measure them. Everything written above about accuracy at small volumes is arithmetic and device geometry, both of which are solid, and none of it constitutes evidence that a particular small dose is a good idea for a particular person. That question has not been answered by anyone.
Frequently Asked Questions
Selected References
- US Food and Drug Administration. FDA alerts health care providers, compounders and patients of dosing errors associated with compounded injectable semaglutide products, July 2024.
- US Food and Drug Administration. FDA's concerns with unapproved GLP-1 drugs used for weight loss, including statements on semaglutide sodium and semaglutide acetate salt forms.
- US Food and Drug Administration. Drug shortages, resolution of the semaglutide injection shortage, February 2025, and resolution of the tirzepatide injection shortage, 2024.
- US Food and Drug Administration. Compounding and the FD&C Act, sections 503A and 503B, requirements for compounding pharmacies and outsourcing facilities.
- America's Poison Centers. National Poison Data System reporting on semaglutide exposure calls.
- Novo Nordisk. Prescribing information for semaglutide injection, sections on pharmacokinetics, elimination half-life and dose administration.
- Institute for Safe Medication Practices. Reports on insulin syringe unit confusion and errors arising from non-standard concentrations.
- United States Pharmacopeia. General chapter 797, pharmaceutical compounding for sterile preparations, on beyond-use dating and labeling.
This article is for general educational purposes only and does not constitute medical advice, diagnosis, or treatment. It is not dosing instructions, and none of the calculations shown here should be used to prepare, measure or self-administer any medication. GLP-1 medications are prescription drugs, and the dose, the product concentration, the syringe and any low dose or microdosing approach should be decided and monitored by a qualified healthcare provider and dispensed by a licensed pharmacy. Nothing here is intended to diagnose, prevent, treat, or cure obesity, diabetes, or any other condition, and no low dose or microdosing approach has been shown in trials to be safe or effective. Compounded preparations are not FDA-approved and their potency and content are not verified by the FDA. Several charts here are constructed, modeled or arithmetic illustrations rather than measured outcomes, and the surveillance figures cited are approximations of published reporting rather than exact counts. Anyone who suspects a dosing error should contact their prescriber, contact Poison Control at 1-800-222-1222 in the United States, or seek emergency care for severe symptoms. Individual results vary.
Smaller Doses. Sharper Protocols.
Metabolic health you can measure, built around a dose designed to work with your body rather than against it.
Find Your Protocol