A storm takes the power out, and the first calculation anyone makes is in hours. How long has it been off. How long can it stay off. That is the wrong unit for this class of drug, and the arithmetic explains why within a line or two: the allowances these labels grant are counted in weeks, and the condition that actually voids one is not measured in time at all.
How long an unpowered refrigerator holds
This has been measured properly. Work at the National Institute of Standards and Technology tested thermal ballast — water containers placed inside a domestic refrigerator to add mass — against temperature stability under several failure modes. A ballast load of ten to fifteen percent of total storage volume kept contents between 2°C and 8°C for four to six hours during a power outage.[1]
The same study found the tactic does not generalize. Ballast did not reliably reduce the frequency or the severity of excursions caused by repeated door opening, by a door left open, or by the refrigerator’s own defrost cycle.[1] The mitigation everyone hears about protects against the rare event and does nothing about the routine ones, which is close to the opposite of how it is usually described. The per-product limits that apply once the door is open at all are collected in the storage article.
The budget an outage is spending
Every approved product in this class permits a defined period outside a refrigerator, and the periods are long. A Wegovy single-dose pen or syringe may be kept from 8°C to 30°C for up to 28 days before the cap is removed, and an Ozempic single-dose prefilled syringe carries the same 28.[2][3] A Wegovy FlexTouch or an Ozempic pen already in use carries 56 days at 15°C to 30°C or refrigerated.[2][3] A Zepbound or Mounjaro single-dose pen or vial carries 21 days at temperatures not exceeding 30°C; an unopened multi-dose vial or KwikPen held at room temperature is discarded after 30 days.[4][5]
Set an outage against the shortest of those. Twenty-one days is 504 hours. A twelve-hour failure consumes about 2% of it; a two-day failure, under 10%. Unless a product has already spent most of its allowance on a countertop, the duration of a domestic outage is essentially never the number that disqualifies a dose. The clock continues to run afterward, because these allowances are cumulative totals rather than per-episode ones, but an outage is a small withdrawal from a large account.
The ceiling is the rule that binds
The allowance is a storage window between two temperatures, not a tolerance for going above the upper one. A house with no power in summer does not hold 30°C, and that is where the labels start to diverge sharply.
One product converts the event into a disposition. Wegovy instructs throwing away the single-dose pen or syringe if it has been frozen, has been exposed to light or temperatures above 86°F (30°C), has been out of the refrigerator for 28 days or longer, or after use; the same instruction is repeated for the FlexTouch pen at its own 56-day figure.[2] Exposure above the ceiling disqualifies the product, with no duration attached and no grace attached.
The other three do not say that. Ozempic prohibits freezing and prohibits using a product that has been frozen, and states its room temperature range, but attaches no discard instruction to the range being exceeded.[3] Zepbound and Mounjaro say not to freeze, say to throw away a pen that has been frozen, and otherwise trigger every discard on a count: 21 days, 30 days, 30 days after first use, four weekly doses.[4][5]
The control on that census is the day counts themselves. Reading the four current labels for discard triggers returns all of them — every 21-day, 30-day, 56-day and four-dose rule on all four products — so a heat-exceeded trigger appearing on only one of them is a genuine asymmetry rather than a reading that missed something. For three of the four approved products, a label that otherwise specifies dating to the day has nothing to say about a temperature having gone past its own stated maximum.
What an outage does not threaten, and what does
One instruction is unanimous. All four labels prohibit freezing and prohibit use of a product that has been frozen.[2][3][4][5] An unpowered refrigerator warms, so that is not the failure an outage produces — it is the failure the appliance produces while working normally, and it is also the failure a cooler produces on the way home. Those two settings are covered in the travel article and the cold chain article.
What an outage genuinely produces is an unknown. The maximum temperature reached while nobody was watching is not recoverable afterward, and none of these products reports its own history. A pen that spent nine hours at 34°C looks exactly like a pen that spent nine hours at 24°C, and on the Wegovy label those two pens have opposite dispositions. A refrigerator thermometer that records a maximum is the only thing that converts the question into an answerable one, and it has to be in place beforehand.
The pharmacy literature treats this as an acknowledged gap rather than a solved problem. A review of power outages and refrigerated medicines notes that after an outage the stability of vaccines and other medicines can be difficult to ascertain, that many public health guidelines therefore recommend discarding potentially compromised pharmaceuticals unless the cold chain can be guaranteed, and that the usefulness of the government guidance aimed at limiting heat exposure is uncertain.[6] Discarding is expensive advice, and it is the advice that exists because the alternative requires data nobody has.
What is unknown for a compounded vial
Every allowance, ceiling and discard trigger above belongs to a product the FDA reviewed. A compounded preparation 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 label to carry a room-temperature allowance, an upper limit or an instruction about exceeding one.[7]
None of the numbers above transfers to it, and that is worth stating flatly because the transfer is tempting and easy. A compounded vial of semaglutide is not entitled to Wegovy’s 28 days, and a compounded vial of tirzepatide is not entitled to Zepbound’s 21. Those figures were established for particular formulations in particular containers. The date on a compounded vial is a beyond-use date assigned by the pharmacy that made it, under a different rulebook, and what that rulebook allows is set out in the storage article. A beyond-use date also assumes the preparation was held at the storage condition it was assigned for, which is exactly the assumption an outage breaks.
So a compounded vial after a power failure sits in a position with no published answer at all. There is no labeled ceiling to have exceeded, no excursion budget to have spent, no manufacturer stability program behind the container, and frequently no disclosed formulation to reason about — the composition problem is the subject of the contents article. A multi-dose vial is also the presentation that has been open longest by the time anything goes wrong, which is one of the differences set out in the vial-versus-pen comparison.
Where the question actually goes
The dispensing pharmacist is the person who can answer it, and the reason is not deference. They know which product was dispensed, which presentation it came in, when it left refrigeration, and for a compounded preparation they hold the beyond-use date and the storage condition it was assigned under. A website has none of that. For an approved product the manufacturer also maintains a line that can be asked about a specific excursion against a specific lot.
What a reader can take from this before anything fails is narrower and more useful than an hour count. The duration of an ordinary outage is very unlikely to exhaust a labeled allowance. A maximum-reading thermometer in the refrigerator converts an unanswerable question into an answerable one. And the product whose storage rules are most exposed to the whole problem is the one with no rules published — which is also the one most of this market sells, on criteria set out in the methodology.