A box arrives on a doorstep in July. The gel pack inside is liquid, the carton is at room temperature, and the question that follows is whether the medicine in it is still medicine. It is a reasonable question with a disappointing answer, and the disappointment is not in the direction most people expect.
None of these labels permits an excursion
“Excursion” is a term of art. In United States labeling it usually appears as a USP Controlled Room Temperature clause — store at one range, brief departures permitted to a wider one — and it is the sentence that makes a warm afternoon a non-event. Across the current labels for Wegovy, Ozempic, Zepbound and Mounjaro the word appears exactly once, and it belongs to the Wegovy oral tablet: store at 20°C to 25°C, excursions permitted to 15°C to 30°C.[1]
No injectable presentation of any of the four carries an excursion clause at all.[1][2][3][4] What they carry instead is a budget: a stated number of days the product may spend out of a refrigerator, differing by product and by presentation, after which the instruction is to discard it. Those specific figures belong to the storage article, and they are worth knowing precisely, because they are the only arithmetic a warm shipment can be measured against.
The consequence is narrower than “ruined” and less forgiving than “fine”. A package that traveled at ambient temperature has not breached a rule; it has spent part of an allowance. And the allowance does not reset. A day out of the refrigerator is a day gone whether it was spent in transit, on a porch, or in a bag, and returning the product to the cold does not buy it back.
The ceiling is where the label stops describing the product
All four labels put the same upper bound on the unrefrigerated range: 30°C, which is 86°F.[1][2][3][4] Lilly’s wording is explicit that the allowance is for temperatures not to exceed that figure. That is the number that matters for a summer delivery, because 86°F is an ordinary afternoon across much of the country and an ordinary interior temperature for a parked delivery van.
Above it, the labels say nothing. Not that the product is destroyed — that it is outside the conditions under which the manufacturer characterized it. That silence is the honest limit of what a label can tell a buyer whose box sat in the sun, and it is why a pharmacist with the product in hand can answer a specific excursion question and a website cannot.
What the peptide chemistry supports, and what it does not
Heat tolerance in an injected peptide is routinely underestimated, and the most rigorous evidence for that is about a different molecule. A Cochrane review of the thermal stability of human insulin, drawing on 17 studies plus previously unreleased manufacturer data, concluded that unopened short-acting and intermediate-acting insulin vials and cartridges can be stored at up to 25°C for a maximum of six months, and at up to 37°C for a maximum of two months, without clinically relevant loss of potency. Manufacturer data within it recorded a loss of short-acting insulin activity of 8.3% to 8.6% after three months at 37°C, and oscillating temperatures between 25°C and 37°C for up to three months produced no loss of activity.[5]
That is insulin, not semaglutide, and it is mostly in vitro. It does not license reusing a GLP-1 that has exceeded its label. What it does establish is the shape of the risk: degradation in a refrigerated peptide product is cumulative and slow at the temperatures a mailbox reaches, not a threshold that flips in an afternoon. Work on semaglutide itself points the same way at the low end. A solid-state thermal stress study found the peptide retained its native alpha-helical conformation up to 60°C, with alpha-helical content falling from 49.07% to 43.75% at that temperature and collapsing to 0.2% only at 80°C.[6] Those are oven temperatures, not delivery temperatures.
The failure a cold chain actually has is cold
The direction of the worry is backwards, and the evidence for that is substantial. A systematic literature review of freezing in the vaccine cold chain found accidental freezing pervasive across every segment of distribution. Between 14% and 35% of refrigerators or transport shipments studied had exposed vaccine to freezing temperatures, and in studies that examined all segments of distribution, between 75% and 100% of shipments had been freeze-exposed. More rigorous study designs were associated with higher levels of detected freeze exposure.[7]
That last clause is the one to sit with. The better the measurement, the more freezing it found — which is what a systematically under-detected failure looks like. And freezing is the one condition every GLP-1 label treats as disqualifying rather than as a budget item: all four instruct against freezing and against using a product that has been frozen, and Novo Nordisk adds that the product should not be stored directly adjacent to a refrigerator’s cooling element.[1][2][3][4]
A shipment packed against a frozen gel pack can cross that line inside a cooler that is doing exactly what it was designed to do. The Cochrane review names the same gap from the other side: after examining everything published on insulin thermostability, its authors record that there were no data for cold environmental conditions at all.[5]The warm failure is measured. The cold one mostly is not.
The last segment of the chain is the buyer’s own kitchen, and the labels legislate for it directly. Novo Nordisk’s instruction not to store the product in the freezer or directly adjacent to the cooling element is an instruction about a specific position rather than a general caution, and it exists because that position can run colder than the compartment the dial is set for.[1][2] A carton pushed to the back of a shelf to keep it out of the way is a carton placed in exactly the position the label warns against. The failure that voids a GLP-1 is not only a shipping problem; it is available at home, every week, for free.
What a buyer can conclude from a warm box
Three things, and they are less than the box seems to promise. First, that some of the product’s unrefrigerated allowance has been spent, in an amount nobody can quantify without knowing how long the package was in transit and what it reached. Second, that the shipment was not frozen in the final leg — which is genuine information, and the opposite of what most people take from a melted ice pack. Third, that the seller shipped without temperature monitoring, because a package with a data logger in it would not leave the question open.
And four things a buyer cannot conclude. Not that the drug lost potency: nothing visible about a clear solution reports that, and the labels do not equate an exceeded day count with a measured loss. Not that the product stayed under 30°C. Not that it was never frozen earlier in the chain, at a distribution center or in a refrigerated truck, which is exactly where the vaccine literature finds most of it. And not that a seller who ships in a plain insulated mailer is behaving differently from one who ships in a branded one.
A compounded preparation removes even the label to argue from. It is not an FDA-approved drug, the agency does not review its safety, efficacy or quality before it is dispensed, and there is no reviewed stability program behind its shipping method. Whether a seller names the pharmacy that made it at all is one of the things the seller reviews record, on criteria set out in the methodology. What is inside such a vial is a separate unknown, covered in the contents article.
The question worth asking instead
Not “was it warm”, but “what does this seller do when it is”. A replacement policy for a delayed or heat-exposed shipment is a checkable commitment; a promise of overnight cold shipping is not, since the buyer never sees the carrier’s actual transit time. A seller that answers the first question in writing has said something. One that answers only the second has said nothing that can be tested after the fact.
The specific excursion, on the specific product, belongs to the pharmacist who dispensed it and the prescriber who wrote it, and they are the only parties who can see the lot and the date. Nothing about the subscription that shipped it changes that — though what those subscriptions actually cost, including the fees charged outside the medication price, is set out in the prepay article.