On this page
The Decision You Have to Make Fast
A refrigerated consignment lands at your Nairobi store and the logger shows the line did not stay flat inside 2 to 8C. Now a clock starts. The product is in quarantine by default, customers are waiting, and someone has to decide: release it, or hold it pending the manufacturer. PPB Good Distribution Practice does not let you guess. It expects a deviation record that shows how you assessed the excursion, what evidence you weighed, and why your disposition is justified. The hard part is that "the temperature went out of range" is not one event with one answer. A brief mild warm spell and a short dip below zero are scored very differently, and treating them the same is how good product gets scrapped and damaged product gets released.
The Temperature Excursion Impact Assessor turns the raw logger trace into a structured read: a severity score for each event, a freeze risk flag, an escalation urgency, and a disposition recommendation. It does not replace your QA sign-off or the manufacturer's stability data. It gives you a repeatable first pass so two reviewers on two shifts reach the same starting point instead of arguing from instinct. This guide walks one realistic case end to end and shows how to write it up so it survives inspection. For the temperature math that feeds it, the mean kinetic temperature primer is useful background reading.
Worked Example: Two Events in One Biologic Shipment
Take a 2 to 8C biologic (a vaccine with an explicit do-not-freeze condition on the label) shipped from the airport on a 1-minute logger. The trace shows two separate breaches inside the same leg, not one:
- A heat excursion: the temperature climbed to a peak of 12C and stayed above 8C for 75 minutes during a tarmac hold, then recovered to 5C once the box was back in the cold room.
- A freeze dip: later in the leg, an over-cooled gel pack pulled the sensor down to a low of minus 1C for 20 minutes before it settled back into range.
Run each event through the assessor separately. The heat event scores a moderate severity: the deviation is 4C above the limit for 75 minutes, which is real but mild, and for many 2 to 8C products a short warm excursion of this size is reversible and likely to sit inside the labelled stability budget. The assessor flags it as moderate, escalation urgency medium, with a disposition of "likely acceptable pending stability confirmation." You still have to check it against the actual allowance, which is where the stability budget calculator earns its place: it tells you how much of the product's cumulative out-of-range allowance these 75 minutes consume.
The freeze event looks smaller on paper (only 20 minutes, only 1C below zero) but the assessor scores it as the more serious of the two and raises the freeze risk flag. The reason is physical, not statistical. A brief mild heat excursion is often reversible; freeze damage to a biologic is frequently not. Ice formation can disrupt the protein structure, aggregate the active, or crack a vial, and none of that un-happens when the product warms back up. For a product carrying a do-not-freeze condition, any excursion below 0C is potential primary damage regardless of how short it was. So the assessor returns a high severity, escalation urgency high, and a disposition of "quarantine and consult the manufacturer before any release decision."
Combined, the shipment disposition is driven by the worse event. The consignment stays in quarantine. The heat excursion is logged as likely within budget pending confirmation, and the freeze excursion triggers the manufacturer technical query that actually gates release. That is the value of scoring events separately then escalating on the maximum: it stops the mild, recoverable heat blip from drowning out the short freeze dip that is the real risk to the patient.
What Data You Need and Why
Three inputs drive the assessment. First, the logger trace: the raw interval readings for the leg under investigation, with timestamps, at the native logging interval. Use raw readings, not daily averages or a chart screenshot. Averaging is what hides the 20-minute freeze dip; on a daily mean it disappears entirely, yet it is the event that matters most. Export the full window so the recovery period is included, because the assessor needs to see where each breach starts and ends.
Second, the product limits exactly as approved: the 2 to 8C storage range and, critically, any do-not-freeze condition from the label or product information. The freeze flag depends on the tool knowing the product must never go below 0C. If you only enter the upper limit, a freeze dip reads as a generic low-side deviation rather than the primary-damage event it is for a biologic.
Third, the stability data: the manufacturer's allowed excursion budget, any published time-out-of-range allowance, and whether the product has known freeze sensitivity. This is what converts a severity score into a defensible disposition. If your logger export is messy (duplicate timestamps, mixed time zones, footer rows), clean it through the data logger QA cleaner before you assess it, so a parsing artefact does not get mistaken for an excursion.
Reading the Outputs
The assessor returns four signals. Read them together, never in isolation:
- Severity score: a per-event rating built from how far the temperature went out of range and for how long. A 75-minute heat excursion to 12C and a 20-minute dip to minus 1C get separate scores, because the disposition follows the worst event, not the average of both.
- Freeze risk flag: a yes/no marker that fires when any reading falls below 0C on a product with a do-not-freeze condition. When this flag is set, weight it heavily. Freeze damage is frequently irreversible, so a flagged event usually outranks a longer, milder heat excursion in the same shipment.
- Escalation urgency: how quickly this needs a decision and who to involve. Medium means assess against stability data and let QA dispose. High means quarantine now and open a manufacturer query before release.
- Disposition guidance: the recommended next step, such as "likely acceptable pending stability confirmation" or "quarantine and consult manufacturer." This is guidance for your QA sign-off, not the sign-off itself.
Pair every score with operational context: the lane and its known delay points, the gel pack configuration, the season, and where the sensor sat in the box. A freeze dip caused by an over-cooled coolant pack against the probe is a packing problem to fix, not just a product to disposition. If the same lane keeps producing freeze flags, the issue is the shipper design, and the MKT calculator on the full leg helps you show whether the cumulative thermal stress is trending the wrong way across shipments.
Documenting the Decision for PPB and GDP Audits
An excursion assessment is a deviation event, and the disposition is only as good as the record behind it. Capture it the way an inspector expects to see it:
- Deviation record: attach the raw logger export and the cleaned file, state each event separately (peak, duration, severity score, freeze flag), and note the approved limits used.
- Manufacturer consultation: when the freeze flag fires or a heat excursion exceeds the stability budget, log the technical query you raised and keep the manufacturer's written response. That response, not your own judgement, is what justifies releasing a frozen-flagged biologic if release is ever appropriate.
- Stability justification: tie the heat-event disposition to the actual time-out-of-range allowance you consumed, so "likely within budget" is backed by a number, not a feeling.
- Change control and CAPA: if the cause is a packing or lane problem, route the fix through change control and open a CAPA. Recurring freeze flags on one lane are a finding waiting to happen.
Done this way, the disposition is defensible: an inspector can follow the chain from the trace, to the per-event scores, to the freeze flag, to the manufacturer's response, to the final QA decision. If you are still building these records from scratch, the GMP readiness checklist covers the deviation and documentation an inspector will ask for.
Frequently Asked Questions
Why does a short freeze dip score higher than a longer heat excursion?
Because the damage mechanism differs. A brief mild heat excursion is often reversible and may sit inside the product's stability budget. Freeze damage to a biologic is frequently irreversible: ice can disrupt the active or crack the vial, and that does not un-happen when the product warms up. For a do-not-freeze product, any reading below 0C is treated as potential primary damage.
Should I score the heat event and the freeze event separately?
Yes. Score each event on its own, then drive the shipment disposition from the worst result. Averaging the two, or treating the leg as one excursion, can let a short freeze dip disappear behind a longer mild heat excursion and produce the wrong release decision.
Does the assessor make the release decision for me?
No. It gives a repeatable first-pass severity score, freeze flag, escalation urgency, and disposition guidance. Final disposition must follow your approved SOP, the manufacturer's stability data, and authorised QA sign-off.
When must I contact the manufacturer before releasing?
Whenever the freeze risk flag fires on a do-not-freeze product, or a heat excursion exceeds the labelled stability budget. Raise a technical query, keep the written response in the deviation record, and let that response gate the release rather than your own judgement.
Why must I use raw logger data instead of daily averages?
Because averaging hides short events. A 20-minute dip to minus 1C vanishes on a daily mean, yet for a biologic it is the most serious part of the excursion. Always export raw interval readings for the full window, including the recovery period.
Run the Tool with Your Own Data
Upload the logger trace for the affected leg, enter your storage limits and any do-not-freeze condition, and get a per-event severity score, freeze flag, escalation urgency, and disposition guidance you can take straight into your deviation record.
Open Excursion Impact Assessor