A data logger alarm is not a diagnosis. It is a prompt to start a short, disciplined sequence of checks that decides whether a refrigerator’s contents are still usable or need to be written off. Field teams that skip straight from alarm to a use-or-discard guess either waste potent stock out of caution or, worse, distribute product that has quietly lost potency. The response protocol matters as much as the logger itself, and it is a different question from which logger to buy or how to calibrate one.
What Counts as a Temperature Excursion
Most vaccines and many other cold chain products are specified to stay between +2°C and +8°C from the point of manufacture to the point of use. A temperature excursion is any period the stored product spends outside that band, in either direction. Heat excursions get more attention, but freezing is often the quieter failure — several widely used vaccines lose potency permanently once frozen, with no way to reverse the damage by rewarming. WHO and UNICEF cold chain guidance, including material from the WHO Vaccine Management Handbook on monitoring temperatures in the supply chain, treats both directions as equally serious and both as irreversible once potency is lost. A logger’s alarm threshold is set to flag exactly this: a departure from the validated range, not a judgment about whether the product is still good.
Isolate First, Investigate Second
The first action on an alarm has nothing to do with reading numbers. It is to physically separate the affected stock from anything already confirmed safe — a clearly labeled quarantine shelf or a second unit, with a note on what triggered the move and when. This single step prevents the most common and most damaging error in cold chain response: someone reaching into the same fridge an hour later, seeing product that looks fine, and using it before anyone has actually reviewed the excursion data. Isolation costs nothing and buys the time needed to do the rest of the sequence properly. Only after stock is physically set aside should the team turn to the logger itself.
Duration and Degree: Reading the Logger’s Full Story
A single alarm timestamp is the least useful piece of information a logger provides. What actually determines risk is the combination of how far the temperature moved outside the band and how long it stayed there. A brief door-open spike to 10°C for four minutes during restocking is a routine event most stability data tolerates without comment. A steady climb to 10°C sustained over six hours because a compressor failed overnight is a different problem entirely, even though both produce an "excursion" flag. This is why pulling the full time-temperature trace — not just the alert — is the actual second step, and why a logger that only reports a pass/fail alarm without a retrievable history leaves a team guessing. Where the product carries a vaccine vial monitor, that heat-sensitive label offers a second, independent read on cumulative heat exposure and should be checked alongside the logger data rather than instead of it.
| Excursion pattern | What it usually indicates | Next step |
|---|---|---|
| Brief spike, minutes, single event | Door-open or restock handling | Log the event; usually no action against manufacturer limits |
| Sustained elevation, hours | Equipment fault, power loss, ambient heat | Pull full logger trace; check VVM; consult stability data before any use |
| Any freezing reading | Overcooling, packing too close to cold walls, defrost cycle fault | Treat as potentially unrecoverable; isolate and escalate regardless of duration |
| Repeated short alarms, same time daily | Routine operational pattern, not equipment failure | Adjust workflow or alarm threshold rather than discarding stock each time |
The Use-or-Discard Decision
Once duration, degree, and any VVM reading are in hand, the decision itself should not be improvised at the fridge door. It belongs to whoever holds responsibility for stock management, working from the manufacturer’s published stability data for that specific product — the document that states how many hours or days a formulation tolerates a given deviation before potency is affected. If the excursion falls inside that documented tolerance and any VVM present shows no change, the stock can generally be returned to use. If it exceeds the documented tolerance, or a VVM indicates heat damage, the guidance is unambiguous: discard rather than gamble on partial potency, because degraded product cannot be identified by appearance alone. Teams that keep the relevant manufacturer inserts on hand, rather than trying to locate them after an alarm has already fired, cut the time this decision takes from hours to minutes.
Building the Response Into Routine, Not Improvisation
None of this works if it is invented fresh every time an alarm sounds. The sequence — isolate, pull the trace, check the VVM, consult stability data, decide, document — is worth writing down as a short standing procedure posted near the storage unit, with names assigned to who is authorized to make the final use-or-discard call. Some organizations pair this with periodic tabletop drills: simulate an overnight power failure, have the on-duty staff walk the actual steps, and see where the sequence stalls. The value of the exercise is finding the gap in daylight — a missing stability sheet, an unclear chain of authority — before a real excursion forces the same gap to be discovered under pressure, with product on the line.
Conclusion
A temperature excursion alarm is the start of a process, not an answer in itself. Programs that isolate stock immediately, read the full duration and degree of the excursion rather than just the alert, weigh any VVM signal alongside the logger data, and check every decision against the manufacturer’s actual stability limits make fewer costly mistakes in both directions — needless waste and needless risk. The logger’s job is to flag the event accurately. The response protocol is what turns that flag into a defensible decision.