Handling temperature excursions in storage and distribution

An excursion is any period where the recorded temperature left the labelled storage range. Finding one is not a failure of the monitoring system; it is the monitoring system working. What happens next is a quality decision with a defined route, and the most common way it goes wrong is that a cumulative average is used to make the event disappear.

Why MKT cannot answer this question

Mean kinetic temperature summarises cumulative chemical stress over a storage period. It is the right tool for that and the wrong tool for an excursion, because a short event is a negligible fraction of the period. The table is a 40-day 2 to 8 °C cold-chain profile at 4 °C, with a single excursion inserted.

Six hours above the limit shifts MKT by eight hundredths of a degree. Report that alongside an in-range MKT and the excursion is invisible, which is exactly why an in-limit MKT is not a disposition.

The last two rows carry the other lesson. Six hours at 25 °C moves MKT further than 48 hours at 8 °C, despite being an eighth of the duration, because the exponential weighting cares far more about how hot than how long. Ranking excursions by duration gets the order wrong.

The misuse to refuse: reporting an in-limit MKT as evidence that an out-of-limit period was acceptable. They answer different questions, and an excursion may have caused damage no cumulative figure models at all, such as freezing a vaccine or melting a suppository.

The route

• Quarantine the affected product immediately, physically or by system status. Everything else can be reconsidered; a release cannot.
• Establish the facts from the record: start and end times, peak, duration above and below limit, which pallets or consignments were in the affected space, and whether the sensor itself was calibrated and in date.
• Rule out a monitoring artefact before treating it as a product event. A sensor out of calibration, a logger in the wrong place or a unit-confirmation error produces excursions that never happened.
• Assess against the product's stability data and its marketing authorisation, including any permitted excursion allowance. This is where the decision actually lives.
• Escalate to the Responsible Person or the equivalent named role, and to the manufacturer where the product's own data does not cover the event.
• Record the disposition with its reasoning, not just its outcome. "Released" without the basis is not a record.
• Investigate the cause and raise corrective action. An excursion with a disposition but no root cause will recur.

What the assessment needs to have available

A decision made in the hour after an excursion is only as good as what was prepared before it. That means the mapping study that shows where the affected location sits relative to the hot and cold spots, the sensor's calibration certificate, the product's permitted excursion data, and a named person authorised to decide. Assembling those during the event is how product sits in quarantine for a week.

Excursions in transit

Transport excursions add two problems: the shipment has usually arrived and been unpacked before anyone reads the logger, and the packaging's qualification data is what has to answer for the interior temperature. A shipper qualified for 72 hours at a defined ambient profile has been characterised for that profile and no other; an excursion outside it is outside the qualification, whatever the interior logger recorded.