Why a temperature excursion needs a defined process, not a judgment call
A temperature excursion is one of the most common events in a pharmaceutical warehouse, hospital pharmacy or vaccine store, and one of the least consistently handled. In the 2019 IQVIA Institute and Peli BioThermal survey of biopharma cold-chain professionals, 44.6% of respondents reported multiple temperature excursions per year and 16% said excursions were a monthly occurrence; the same report put losses from temperature-controlled logistics failures at roughly $35 billion a year.[2] Most of those events are not dramatic freezer failures. They are a door left open during a stock take, a defrost cycle that runs long, or a two-hour power cut on a Sunday night.
What separates a well-run site from a poorly run one is not the absence of excursions but whether each one is detected quickly, documented completely, assessed against real stability data, and closed with a corrective action. This guide covers what counts as an excursion under USP <1079>, WHO and EU GDP guidance, the first-hour actions, the fields a temperature excursion report must contain, how to run the impact assessment (including where mean kinetic temperature helps and where it does not), and how to set alarm thresholds and escalation.
What counts as a temperature excursion (USP <1079>, WHO, EU GDP)
The definitions are aligned. USP General Chapter <1079> defines a temperature excursion as "an event in which a pharmaceutical product is exposed to temperatures outside of the range(s) prescribed for storage and/or transport", adding that those ranges "are determined by the product manufacturer, based on stability data".[1] WHO Technical Report Series 961, Annex 9 uses the same wording for time- and temperature-sensitive pharmaceutical products (TTSPPs).[4] For vaccines, the CDC Vaccine Storage and Handling Toolkit is blunt: "Any temperature reading outside ranges recommended in the manufacturers' package inserts is considered a temperature excursion."[6]
Three practical points follow:
- The reference is the labelled range, not your alarm set-point. A reading of 8.3 °C is an excursion for a 2 °C to 8 °C product even if the alarm was set at 8.5 °C and never fired.
- A brief excursion may be acceptable, but only after assessment. USP <1079> states that excursions "for brief periods, may be acceptable provided that stability data and scientific/technical justification exist" showing quality is unaffected.[1] Acceptable is a conclusion, not a reason to skip the record.
- Alarms are expected. EU GDP (2013/C 343/01), section 3.2.1, requires that "appropriate alarm systems should be in place to provide alerts when there are excursions from pre-defined storage conditions" and that "alarm levels should be appropriately set and alarms should be regularly tested".[3]
Excursion vs deviation vs alarm
An alarm is a notification from the monitoring system when a set-point, delay or event condition (door, power) is met; it may be a pre-alarm inside the labelled range. A temperature excursion is product exposed outside its prescribed range, whether or not an alarm fired. A deviation is the quality-system record of a departure from an approved procedure or condition. USP <1079> says each excursion "should be documented and handled with a deviation or appropriate risk assessment",[1] and EU GDP section 1.2 requires that "deviations from established procedures are documented and investigated" and that "appropriate corrective and preventive actions (commonly known as 'CAPA') are taken".[3] So an alarm may or may not become an excursion; every excursion becomes a deviation or a documented risk assessment.
Photo: U.S. Army RHCE / Russell Toof, via Wikimedia Commons, public domain
Immediate actions: the first 60 minutes
Your excursion SOP should be short enough for a night-shift operator to follow from memory. The CDC toolkit's response steps are a good template for any temperature-sensitive product.[6]
- Acknowledge and escalate. Whoever hears the alarm or sees the reading notifies the named responsible person immediately.[6]
- Contain, do not discard. Label exposed stock "DO NOT USE" and isolate it while the assessment runs; the CDC wording is "do not discard these vaccines".[6] WHO TRS 961 says products with unacceptable excursions should be quarantined and not released "until checks have been completed satisfactorily".[4]
- Restore control. Close the door, restart the unit, switch to backup power or move stock to a qualified backup unit; never leave product in a failed unit "for an extended period of time".[6]
- Preserve the evidence. Export the continuous log for the whole event window, min/max values, alarm timestamps and door or power events before anyone resets anything.
- Open the record while details are fresh, plus the QMS deviation if your SOP requires it.
- Contact the manufacturer or immunization program if you do not hold the stability data; disposition advice depends on the magnitude and duration you report.[6]
Excursion report: what to document
A reviewer who was not present should be able to reconstruct the event and verify the disposition from the report alone. These fields merge the CDC documentation list, WHO's requirement to keep "records of excursions" per batch and USP's basis for disposition (temperature reached and time exposed).[6][4][1]
| # | Field | What to capture |
|---|---|---|
| 1 | Event identifier | Excursion or deviation number, site, storage unit ID, zone. |
| 2 | Date and time | Start (last in-range reading), end (first stable in-range reading), total duration out of range.[6] |
| 3 | Temperatures | Labelled range, peak or trough reached, min/max over the event, room ambient if available.[6] |
| 4 | Detection | How it was found (alarm, routine check, audit), who acknowledged it and when. |
| 5 | Description | What happened: door open, defrost, power failure, compressor fault, warm stock loaded, HVAC failure.[6] |
| 6 | Affected inventory | Products, batch/lot numbers, quantities, expiry dates, non-product items in the unit.[6][4] |
| 7 | Prior history | Earlier excursions for the same lots (cumulative exposure) and known problems with the unit.[6] |
| 8 | Evidence attached | Continuous data export (CSV or PDF), trend graph, alarm log, door/power event log. |
| 9 | Immediate actions | What was done with the product and when; who was contacted; instructions received.[6] |
| 10 | Impact assessment | Stability data or manufacturer statement used, MKT for the exposure period if applicable, conclusion. |
| 11 | Disposition | Released, released with shortened expiry, returned or destroyed, and by whose authority.[6] |
| 12 | Root cause and CAPA | Root cause, corrective and preventive action, owner, due date, effectiveness check.[3] |
| 13 | Sign-off | Prepared by, reviewed by QA, date; electronic signatures if the record is electronic. |
Photo: Phil Hearing / Unsplash
Impact assessment: stability data, MKT and the disposition decision
USP <1079> sets out the logic: disposition "should be established on the basis of an assessment of the excursion (i.e., the temperature to which the material or product was exposed, and for how long)", combined with the manufacturer's stability data.[1] A practical sequence:
- Quantify the exposure. State the peak (or trough) temperature and total minutes outside range from the continuous log, using the actual curve rather than assuming the peak applied throughout.
- Find the applicable stability statement. Many labels or manufacturer letters state permitted excursions. If you do not hold this data, ask the manufacturer or, for vaccines, the immunization program;[6] downstream handlers "may rely on the manufacturer's product disposition instructions".[1]
- Calculate MKT for the exposure period where appropriate. ICH Q1A(R2) defines mean kinetic temperature as "a single derived temperature that, if maintained over a defined period of time, affords the same thermal challenge to a drug substance or drug product as would be experienced over a range of both higher and lower temperatures".[7] USP <1079> adds a guard-rail: MKT "should be calculated for the period of time that a drug is in residence at a warehouse and/or in transit on a truck to avoid the problem of diluting the impact of excursions by calculating annual MKT values".[1]
- Know when MKT does not apply. MKT is a heat-stress tool for Arrhenius-type degradation; it says nothing about freezing. WHO's review of vaccine temperature sensitivity notes that freeze-sensitive vaccines such as hepatitis B, DTP-containing and IPV vaccines are irreversibly damaged at or below 0 °C.[10] A freeze event needs a product-specific assessment, not an average.
- Decide and document. Record the conclusion (fit for use, shortened expiry, return, destroy) with the supporting data. ICH Q9(R1) says the "level of effort, formality and documentation of the quality risk management process should be commensurate with the level of risk"; a 20-minute drift to 8.4 °C on a robust tablet needs less paperwork than a six-hour freeze on a biologic.[9]
Cumulative exposure is the point sites most often miss: a lot that used part of its allowance in transit has less headroom in storage, which is why field 7 belongs in the report.
Root cause and CAPA: why the same excursion keeps happening
ICH Q10 expects a CAPA system that acts on "the investigation of complaints, product rejections, non-conformances, recalls, deviations, audits, regulatory inspections and findings, and trends".[8] For US manufacturers, 21 CFR 211.192 requires that unexplained discrepancies be "thoroughly investigated" and that "a written record of the investigation shall be made and shall include the conclusions and followup".[11] Repeat excursions from the same unit with the same cause and no preventive action are exactly what inspectors look for. The usual causes:
| Root cause | Signature in the data | Preventive action |
|---|---|---|
| Door openings (picking, stock-takes, deliveries) | Short spikes in working hours; door-side sensor reacts first | Door-open event alarms and time limits; sensors away from transient zones;[5] load scheduling |
| Defrost cycles | Regular, timed rises in freezers and cold rooms, often overnight | Confirm defrost parameters, verify product-zone temperature stays in range, set alarm delay accordingly |
| Power failure | Steady rise from the outage; data gaps if the logger has no battery | UPS or generator, battery-backed loggers with power-failure alerts, emergency transfer plan |
| Refrigeration or HVAC fault | Slow drift over hours or days | Preventive maintenance, pre-alarms inside the labelled range, spare capacity |
| Warm stock loaded or seasonal ambient extremes | Rise on receipt days; hot spots only in summer | Stage receipts; re-map in both seasons and move sensors or stock |
| Wrong alarm configuration | Excursion in the data but no alarm, or a late one | Review set-points, delays and recipients after every event; test alarms regularly[3] |
WHO Technical Supplement 6 adds that if door-driven "transient events" in an area generate too many out-of-range alarms and cannot be fixed technically or operationally, "these areas should not be used to store TTSPPs".[5] Sometimes the corrective action is to stop storing product in that zone.
Photo: Luke Chesser / Unsplash
Alert thresholds and escalation that prevent excursions
WHO Technical Supplement 6 describes the alarm structure a monitoring system should offer: "low and high alarm threshold settings, triggered before temperature goes out of range" (pre-alarms), "low and high alarm settings, triggered after temperature goes out of range", and "event alarms triggered by events such as mains power failure or door open"; limits "should be set only by authorized users" and the system "should automatically alert responsible staff by email, text (SMS) message or other communication medium".[5] A workable configuration for a 2 °C to 8 °C unit:
| Level | Condition | Notified | Expected response |
|---|---|---|---|
| Pre-alarm | Above 7 °C or below 3 °C for more than 15 minutes | Storekeeper on duty (app push) | Check door, load, unit; acknowledge |
| Alarm | Above 8 °C or below 2 °C beyond the agreed delay (for example 10 to 15 minutes, from qualification data) | Storekeeper + supervisor (SMS + email) | Follow excursion SOP; open record |
| Escalation 1 | Not acknowledged within 15 minutes | QA on call / facility manager | Take over response |
| Escalation 2 | Not acknowledged within 30 minutes, or power-failure / door-open event alarm | Site head | Trigger emergency transfer plan |
Two rules prevent alert fatigue: base delays on your own qualification data (how long the unit holds range with the door open or power off), and review alarm counts monthly, because a unit that generates dozens of nuisance alarms trains staff to ignore the real one. EU GDP's "regularly tested" is the minimum.[3]
How Ideabytes IoT helps
Ideabytes IoT loggers and the CMARS21 cloud platform are built around this workflow.
- Alarms with escalation and acknowledgement. CMARS21 sends live alerts by SMS (optional), email and app push, escalates when an alarm is not acknowledged, and keeps an alarm log next to the trend graph for every device.
- Evidence on demand. Each logger stores 90 days of readings locally at the default 15-minute interval and runs up to about 20 hours on battery, so a power cut does not create a gap. Scheduled and on-demand PDF reports are 21 CFR Part 11-compliant, include the MKT value and carry an audit trail; CAPA notes can be attached to events.
- IBI-WTH120 Wi-Fi Temperature & Humidity Data Logger: -40 °C to +120 °C, 0 to 100% RH (condensing), ±0.5 °C (-10 °C to +50 °C), IP64, for cold rooms and warehouses where humidity also matters.
- IBI-MT120 4G M2M Temperature Data Logger: -50 °C to +120 °C, 4G LTE Cat1 with 2G fallback on a built-in M2M SIM, for sites without reliable Wi-Fi.
- IBI-MT120-CN 4G M2M Temperature + GPS Data Logger: -40 °C to +110 °C with GPS, a door sensor and a local buzzer, so "door open" becomes a timestamped event alongside location and temperature.
See also our guides on 21 CFR Part 11 automated temperature monitoring for pharmaceutical warehouses and blood bank and vaccine storage monitoring with real-time alerts, and the companion article on how mean kinetic temperature is calculated.
Frequently asked questions
Is every alarm a temperature excursion?
No. A pre-alarm inside the labelled range is a warning. An excursion is product exposed outside its prescribed range, whether or not an alarm fired.[1]
Can I use MKT to justify a refrigerated or frozen product excursion?
Only within what the manufacturer's stability data supports, and never for freeze events. USP <1079> directs that MKT be calculated for the actual residence period, and freeze-sensitive products are damaged irreversibly regardless of the average.[1][10]
How long should excursion records be kept?
The longer of your national requirement and your product-retention rule. EU GDP requires monitoring data to be backed up and retained "for the period stated in national legislation but at least five years".[3]
Conclusion
Temperature excursions will happen. What you control is detection time, the completeness of the record, the rigor of the impact assessment, and whether the root cause is fixed. Write the SOP, build the 13-field report into your system, set pre-alarms and escalation from qualification data, and review trends monthly. If you would like help configuring alarms, escalation chains and 21 CFR Part 11 reports for your storage units, talk to the Ideabytes IoT team.
References
- USP General Chapter <1079> Risks and Mitigation Strategies for the Storage and Transportation of Finished Drug Products (authorized reprint, 28 Oct 2024). U.S. Pharmacopeia. usp.org (PDF)
- Failures in temperature-controlled logistics cost biopharma industry billions (IQVIA Institute / Peli BioThermal 2019 survey). Air Cargo News, 26 Jul 2019. aircargonews.net
- Guidelines on Good Distribution Practice of medicinal products for human use (2013/C 343/01), sections 1.2, 3.2.1, 3.3.1 and 9.2. European Commission, 5 Nov 2013. eur-lex.europa.eu
- WHO Technical Report Series 961, Annex 9: Model guidance for the storage and transport of time- and temperature-sensitive pharmaceutical products. World Health Organization, 2011. who.int
- WHO Technical Supplement 6 to TRS 961 Annex 9: Temperature and humidity monitoring systems for fixed storage areas. World Health Organization, 2015. cdn.who.int (PDF)
- Vaccine Storage and Handling Toolkit, "Temperature Excursions (Out-of-Range Temperatures)". U.S. Centers for Disease Control and Prevention. cdc.gov (PDF) (mirror: tn.gov)
- ICH Q1A(R2) Stability Testing of New Drug Substances and Products, Glossary: Mean Kinetic Temperature. International Council for Harmonisation, Feb 2003. database.ich.org (PDF)
- ICH Q10 Pharmaceutical Quality System, section 3.2.2 Corrective Action and Preventive Action (CAPA) System. International Council for Harmonisation, June 2008. database.ich.org (PDF)
- ICH Q9(R1) Quality Risk Management. International Council for Harmonisation, Jan 2023. database.ich.org (PDF)
- Temperature sensitivity of vaccines (WHO/IVB/06.10). World Health Organization, 2006. iris.who.int
- 21 CFR 211.192 Production record review (investigation of discrepancies). U.S. FDA via eCFR. ecfr.gov