Skip to content
GLP Loss
← Research
Evidence

Vial or Pen: What the Error Data and the Labels Say

Compounded GLP-1 reports carry 48.9 times the reporting odds of a preparation error — and 0.24 times the odds of a dosing error. The failure is upstream of the syringe, and the multi-dose container is a different formulation.

Owen Castellanos9 min read
Where the errors actually areReporting odds, compounded versus non-compounded, 707 of 81,078 reports1.0Preparation errors48.9×Prescribing errors4.5×Administration errors0.29×Dosing errors0.24×The dosing term points the wrong wayReporting odds are not incidence, and compounded events are underreportedA multi-dose container needs a preservative. A single-dose one carries none.Compounded drugs are not FDA-approved and have no label to read either way.

Almost every approved GLP-1 for weight management reaches a patient as a pen. Almost every compounded one reaches them as a glass vial with a rubber stopper and a bag of syringes. The two presentations are not alternative packaging for one product; they hand the patient different jobs, contain different things, and fail in different places.

What the labels actually say each presentation is

The approved range is more varied than the marketing suggests. Wegovy is supplied as cartons of four single-dose pens, four single-dose syringes with integrated needles, or one single-patient-use FlexTouch pen holding four weekly 2.4 mg doses.[1] Ozempic is a multi-dose pen — three milliliters delivering four doses of 1 mg or of 2 mg, or a mixture of 0.25 mg and 0.5 mg doses at the start — and is also supplied as single-dose prefilled syringes.[2]

Lilly publishes a vial on its own label. Zepbound and Mounjaro are supplied as single-dose pens, single-dose vials of 0.5 mL, and a multi-dose vial or KwikPen holding 2.4 mL, which provides four doses at 0.6 mL each.[3][4] So an FDA-reviewed multi-dose vial of tirzepatide exists, and it is the only published point of comparison for the container most of this market ships. The dosing instruction that comes with it is a volume rather than a device: use a syringe appropriate for the dose, the example given being a 1 mL syringe capable of measuring 0.5 mL or 0.6 mL.[3] Which syringe that should be is the subject of the needle article.

A multi-dose container is a different chemical product

The excipient lists in section 11 of these labels split cleanly along the single-dose and multi-dose line, and the split is not cosmetic. A Zepbound or Mounjaro single-dose pen or single-dose vial contains tirzepatide, sodium chloride, sodium phosphate dibasic heptahydrate and water for injection — and no antimicrobial preservative at all. The multi-dose vial and the KwikPen add, per dose, benzyl alcohol 5.4 mg, glycerin 4.8 mg and phenol 1.08 mg.[3][4]

Novo Nordisk’s labels do the same thing. A Wegovy single-dose pen or single-dose syringe lists disodium phosphate dihydrate, sodium chloride and water. The FlexTouch pen that holds four doses, and the Ozempic multi-dose pen, both add phenol 5.5 mg and propylene glycol 14 mg per milliliter.[1][2]

A preservative is what a container earns by being entered more than once. Every puncture is an opportunity for contamination, and a formulation meant to survive four of them over thirty days needs antimicrobial protection that a use-once device does not. This is the quiet reason a vial and a pen are not interchangeable even when the active ingredient is identical: the vial version of a drug contains substances the pen version does not. What a compounded vial contains is a separate question with a worse answer, because there is no label to read — see the contents article.

The same labels draw a second line around a multi-dose container that a compounded vial rarely carries in writing. Each Ozempic pen is for use by a single patient, and the label states it must never be shared between patients even if the needle is changed, because pen-sharing poses a risk for transmission of blood-borne pathogens.[2] Lilly says the same of the KwikPen, and Novo Nordisk of the Wegovy FlexTouch.[1][3] A container holding four weeks of somebody’s treatment is a container somebody else might be offered a dose from, and that instruction exists because it happens.

The error evidence, and the direction it runs

A pharmacovigilance study of the FDA Adverse Event Reporting System covering 2018 to 2024 isolated 707 reports involving compounded GLP-1 receptor agonists out of 81,078 total and calculated adjusted reporting odds ratios against non-compounded formulations. Compounded products carried higher reporting odds for preparation errors (ROR 48.92, 95% CI 12.63 to 189.6), contamination (19.00, 4.24 to 85.03), compounding or manufacturing issues (8.51, 5.17 to 14.0) and prescribing errors (4.46, 2.49 to 7.98), and higher odds of hospitalization (2.35, 1.94 to 2.83).[5]

The two terms a vial-versus-pen argument would expect to move most went the other way. Compounded products showed lower reporting odds of administration errors (0.29, 0.16 to 0.53) and lower odds of dosing errors (0.24, 0.17 to 0.32).[5] Read carelessly that looks like a defense of the syringe. It is better read as a map of where the failure sits: upstream, in how the preparation was made and how the dose was written, rather than at the moment of injection.

These are reporting odds, not rates. FAERS counts reports, not patients, and the reporting channel differs by construction: federal law does not require state-licensed pharmacies that are not outsourcing facilities to submit adverse events to FDA, and the agency states that events from compounded versions are therefore likely underreported.[6] A ratio built on a channel that under-counts one arm cannot be read as an incidence comparison in either direction. The regulatory split behind that asymmetry is set out in the 503A and 503B article.

What the poison-center record shows

Case-level reporting fills in what a ratio cannot. Three cases reported to one regional poison control center involved incorrect administration of semaglutide obtained from compounding pharmacies and an aesthetic spa; two were tenfold dosing errors. One patient described receiving a vial with syringes and no pharmacist counseling on administration, and one reported dosing in milliliters and units rather than milligrams. The authors’ conclusion is the sharpest sentence in this literature: vials of compounded semaglutide do not use the safety features provided by prefilled manufactured pens, and allow for large overdoses.[7]

A second series described three unintentional overdoses at initiation — 2 mg instead of 0.1 mg, 2.4 mg instead of 0.25 mg, and 1.7 mg. All three developed nonspecific gastrointestinal symptoms and none experienced hypoglycemia.[8] A nine-year retrospective cohort at another regional center counted 237 GLP-1 exposures, of which 164 (69.2%) were unintentional therapeutic errors; semaglutide was the most frequent agent at 36.0%, and hypoglycemia appeared in eight patients (3.4%), whose lowest mean blood glucose was 49.6 ± 23.7 mg/dL.[9]

Two things follow, and they cut in opposite directions. The feared outcome is rare: a tenfold GLP-1 overdose produces days of vomiting far more often than it produces a hypoglycemic emergency. And a pen is not immunity — that 69.2% therapeutic-error majority accumulated across 2014 to 2023, a period dominated by approved prefilled devices. A device removes the milligram-to-unit conversion. It does not remove the possibility of using it wrongly.

What FDA says produced the large errors

The agency’s own account is specific about mechanism. Most of the reports it received described patients drawing more than the prescribed dose from a multiple-dose vial and administering five to 20 times the intended amount; in several, patients told to draw a 5-unit (0.05 mL) dose into a 1 mL U-100 insulin syringe administered 50 units. Providers converting milligrams to units made the same class of mistake in the other direction, prescribing 25 units for an intended 0.25 mg dose and 20 units for an intended 2.[10]

FDA also names the structural cause rather than blaming the patient: compounded semaglutide is offered in various containers and at concentrations that differ between compounders and sometimes within one compounder, and accompanying instructions may specify “units” whose volume depends on a concentration the patient never sees.[10] One approved concentration and one fixed device makes that arithmetic impossible to get wrong. Several unpublished concentrations and a generic syringe makes it routine.

There is also a quieter asymmetry in who bears the consequence. A tenfold overdose from a vial shows up as days of vomiting and a poison center call. A tenfold underdose shows up as nothing at all — no symptom, no alarm, and a patient who concludes after two months that the drug does not work for them. Neither the FAERS series nor the poison-center cohorts can count that second failure, because nobody reports it. What a plateau does and does not mean is covered in the plateau article.

What the difference costs

The presentation is the main reason the two markets do not converge on a price. A device has to be manufactured, filled and tested; a vial is glass and a stopper. The approved products are sold in counted doses and the manufacturers define a month accordingly, which is worked through in the price-variation article alongside the full distribution of what sellers here advertise.

The saving is real and so is what it buys. A secret-shopper cross-sectional study of 75 weight-loss clinics and medical spas in two states, conducted from August to October 2025 after the semaglutide and tirzepatide shortages ended, identified 23 compounding facilities supplying them. Four of 21 were not licensed to perform sterile compounding, one of 23 had received multiple FDA warning letters since 2023, and 3 of 22 had been subject to state-level disciplinary action. Forty-two of the 75 businesses (56.0%) offered compounded GLP-1 products combined with B vitamins.[11] None of that is visible in a monthly figure, and the ordered field of what sellers here charge is the price board.

Compounded drugs are not FDA-approved. The agency does not review their safety, effectiveness or quality before they are dispensed, and that applies to the container and the concentration as much as to the molecule.[6] A pen is a manufacturing decision that has been reviewed. A vial with a bag of syringes is a manufacturing decision that has not, and the questions that separate one compounder from another are in the vetting article.

Frequently asked

Is a compounded vial more dangerous than a pen?
The only study measuring both found compounded GLP-1 reports carried 48.9 times the reporting odds of a preparation error and 2.35 times the odds of hospitalization, but 0.24 times the odds of a dosing error. Those are reporting odds rather than rates, and compounded events are likely underreported because state-licensed pharmacies are not required to send them to FDA, so neither figure is an incidence comparison.
Do multi-dose GLP-1 vials contain preservatives?
The approved multi-dose presentations do. A Zepbound or Mounjaro multi-dose vial or KwikPen contains benzyl alcohol 5.4 mg, glycerin 4.8 mg and phenol 1.08 mg per dose; the single-dose vial and pen of the same drug contain none of those. Ozempic and the four-dose Wegovy FlexTouch carry phenol 5.5 mg and propylene glycol 14 mg per mL, while single-dose Wegovy does not.
Does an FDA-approved tirzepatide vial exist?
Yes. Zepbound and Mounjaro are supplied as single-dose vials of 0.5 mL and as a 2.4 mL multi-dose vial providing four weekly doses of 0.6 mL, alongside the pens. That makes the Lilly multi-dose vial the only FDA-reviewed point of comparison for the container most compounded sellers ship.
How large were the reported dosing errors with vials?
FDA reports that most of the errors it received involved patients drawing five to 20 times the intended dose from a multiple-dose vial, including patients instructed to measure a 5-unit (0.05 mL) dose in a 1 mL U-100 insulin syringe who administered 50 units. Providers converting milligrams to units made errors of five to ten times in the same direction.
Does a pen prevent dosing mistakes?
It removes one of them. A pen fixes the concentration and the delivered volume, so the milligram-to-unit conversion that produced the largest reported errors cannot occur. It does not remove error generally: in a nine-year poison-center cohort of 237 GLP-1 exposures spanning a period dominated by approved prefilled devices, 164 cases (69.2%) were unintentional therapeutic errors.
Why is compounded product almost always a vial?
Because a pen is a manufactured device and a vial is glass and a stopper, so the vial route carries far less unit cost. That cost difference is a real part of why compounded and approved prices do not converge, and it is also what transfers the measuring job from the manufacturer to the patient.

Sources

  1. [1] Novo Nordisk Pharmaceutical Industries, LP (2026). WEGOVY (semaglutide) injection, solution — Description and How Supplied: single-dose presentations, FlexTouch four-dose pen, excipients (read September 15, 2026) DailyMed, U.S. National Library of Medicine. Source
  2. [2] Novo Nordisk Pharmaceutical Industries, LP (2026). OZEMPIC (semaglutide) injection, solution — Description and How Supplied: multi-dose pen, single-dose prefilled syringes, phenol and propylene glycol content (read September 15, 2026) DailyMed, U.S. National Library of Medicine. Source
  3. [3] Eli Lilly and Company (2026). ZEPBOUND (tirzepatide) injection, solution — Description: single-dose vial and pen excipients versus multi-dose vial and KwikPen benzyl alcohol, glycerin and phenol (read September 15, 2026) DailyMed, U.S. National Library of Medicine. Source
  4. [4] Eli Lilly and Company (2026). MOUNJARO (tirzepatide) injection, solution — Description: single-dose and multi-dose presentation excipients (read September 15, 2026) DailyMed, U.S. National Library of Medicine. Source
  5. [5] McCall KL, Mastro Dwyer KA, Casey RT, Samana TN, Sulicz EK, Tso SY, Yalanzhi ER, Piper BJ (2026). Safety analysis of compounded GLP-1 receptor agonists: a pharmacovigilance study using the FDA adverse event reporting system. Expert Opin Drug Saf. PMID 40285721
  6. [6] U.S. Food and Drug Administration (2026). Medications Containing Semaglutide Marketed for Type 2 Diabetes or Weight Loss — adverse event reporting is not required of state-licensed pharmacies, and compounded events are likely underreported (read September 15, 2026) U.S. Food and Drug Administration. Source
  7. [7] Lambson JE, Flegal SC, Johnson AR (2023). Administration errors of compounded semaglutide reported to a poison control center-Case series. J Am Pharm Assoc (2003). PMID 37392810
  8. [8] Wiener BG, Gnirke M, Vassallo S, Smith SW, Su MK (2024). Challenges with glucagon-like peptide-1 (GLP-1) agonist initiation: a case series of semaglutide overdose administration errors. Clin Toxicol (Phila). PMID 38470137
  9. [9] Muschler K, Muschalek R, Hoyte C (2025). Characterization of glucagon-like peptide-1 (GLP-1) agonist exposures reported to a single United States poison center. Clin Toxicol (Phila). PMID 39803696
  10. [10] U.S. Food and Drug Administration (2024). FDA alerts health care providers, compounders and patients of dosing errors associated with compounded injectable semaglutide products (read September 15, 2026) U.S. Food and Drug Administration. Source
  11. [11] DiStefano MJ, Tilley A, Paratane D, Gyimah Gyamfi H, Moore GD, Nair KV (2026). Postshortage Compounded GLP-1 RA Market in 2 States With Potentially High Demand. JAMA Health Forum. PMID 42467450

Where to get it

Best GLP-1 injections

Every injectable seller we can verify, with the price each one publishes and an honest read of what the trials measured.

Compare providers →

More in Evidence