UK pharmacies and wholesalers operate under two overlapping regulatory regimes - MHRA's Good Distribution Practice (GDP) and the General Pharmaceutical Council's pharmacy standards. Both demand evidence, not assurances. This is the regulator-mapped deep dive: what the rules actually say, what inspectors actually ask for, and what your monitoring system must actually produce.
TL;DR: UK pharmacy cold-chain compliance hinges on five evidence types - calibrated continuous data, validated mapping, deviation logs with CAPA, retained records (5 years for GDP), and qualified procedures. MHRA reports continuously cite "inadequate temperature records" as a top-five GDP deficiency (MHRA Inspectorate blog, 2024). A correctly deployed IoT system covers four of the five automatically.
If you want the general introduction to pharmacy temperature monitoring, start with our existing primer on pharmacy temperature monitoring and compliance. This post assumes you already know why monitoring matters - and goes straight into how to satisfy MHRA on the day they arrive.
What does MHRA actually require for cold-chain storage?
MHRA enforces the UK Orange Guide, which folds in EU GDP Chapter 9.2 word-for-word for temperature control (MHRA Orange Guide 2022 edition, 2022). For any medicine labelled "store at 2-8°C", the storage location must be continuously monitored with calibrated equipment, mapped under realistic load, and managed through documented excursion procedures. Inspectors will not accept a fridge thermometer and a paper log.
The Inspectorate has published its annual deficiency trends since 2019, and "inadequate temperature monitoring" or "incomplete records" has appeared in the top ten every single year (MHRA Inspectorate blog - GDP deficiency data, 2023). The pattern is consistent: operators have data, but cannot produce it in an inspector-friendly form within the inspection window.
The five evidence types every inspection touches
- Raw temperature data at the original sample interval, not aggregated.
- Calibration certificates from a UKAS-accredited or ISO 17025-accredited lab.
- Deviation log with root-cause analysis and CAPA (corrective and preventive action).
- Temperature mapping report for each storage location, repeated on a defined cycle.
- Qualified procedures - SOPs, training records, system qualification documents.
[PERSONAL EXPERIENCE] In our deployments across UK wholesalers, the single biggest cause of an MHRA "major" finding is not a sensor failing - it's the operator being unable to produce the raw interval data because their old system only stored 15-minute averages. Averages destroy excursions.
How does EU GDP Chapter 9.2 map to a real IoT deployment?
GDP Chapter 9.2 is short - barely a page - but it sets a high evidentiary bar. It demands "suitable equipment", "regular monitoring", "calibration traceable to national or international standards", and "documented procedures for handling deviations" (European Commission GDP Guidelines, 2013). Every clause has a direct technical counterpart in a properly designed IoT stack.
Requirement-to-evidence mapping table
| GDP / MHRA requirement | Evidence inspectors will ask for | IoT-WorkS deployment artifact |
|---|---|---|
| Continuous monitoring (9.2) | Raw readings at native interval, no gaps | VS-T200 sensors at 1-5 min cadence, gateway store-and-forward |
| Calibrated equipment (9.2) | UKAS/ISO 17025 cert per sensor | Per-device calibration cert PDF stored against sensor ID |
| Temperature mapping (9.2) | Mapping report with summer + winter, full + empty | VS-T200 grid deployment, 14-day study export |
| Door / access events | Door-open log correlated with excursions | VS-D300 reed-switch log joined to temperature stream |
| Deviation procedure (9.2) | Written SOP + populated CAPA log | Alarm tree config + signed CAPA workflow in dashboard |
| 5-year record retention | Exportable archive, no data loss | Immutable cloud archive + on-prem mirror option |
| Power-failure resilience | Continuity of record through outage | VG-O500 gateway with cellular failover + UPS |
| System qualification | IQ/OQ/PQ documents | Qualification pack delivered at commissioning |
[ORIGINAL DATA] Across more than 2,400 deployed sensors in cold-chain environments, the most common deviation type we record is door-open-during-restock - not equipment failure (IoT-WorkS Cold Chain industry data, 2025). This is why door telemetry from the VS-D300 matters: without it, a 30-minute warm event looks like a fridge malfunction and triggers an avoidable CAPA.
What does an MHRA inspector look for during a GDP audit?
Inspectors arrive with a sampling plan. They will pull a calendar week from the last twelve months, ask for raw data across that window, cross-reference any alarm events with the deviation log, and check that every flagged excursion has a closed CAPA. The 2023 MHRA Inspectorate symposium summary noted that 67% of GDP critical and major findings related to documentation rather than physical conditions (MHRA Inspectorate blog, 2023).
The three audit-day questions a monitoring system must answer instantly
- "Show me every excursion for fridge F-03 between 14 and 20 March." - the system must filter raw data by asset and date range, including readings either side of the excursion for context.
- "Where is the calibration certificate for sensor S-119?" - one click from the reading to the PDF cert. Auditors lose patience with shared drives.
- "What was the root cause of the alarm at 03:14 on 8 February?" - the deviation record must show acknowledgement time, investigator, root cause, and CAPA reference.
[UNIQUE INSIGHT] The systems that fail audits are not the ones with bad data - they are the ones where the data lives in three places (sensor logger CSVs, an email inbox, a SharePoint folder of deviation forms) that cannot be joined. The fix is architectural: one stream, one timeline, one query.
For the AI-driven side of this - automated root-cause prompts on every deviation - see our work on AI telemetry and anomaly detection.
What hardware actually satisfies these requirements?
There is no MHRA-approved hardware list. There are characteristics inspectors expect: accuracy under ±0.5°C across the operating range, calibration traceability, sample intervals fast enough to catch realistic excursions, and resilience to power and network failure. Most pharmacy fridge excursions develop within 10-15 minutes during restock, so a 5-minute sample interval is the practical minimum (WHO PQS performance specification E006, 2020).
Reference deployment for a typical UK pharmacy
- VS-T200 temperature/humidity sensors (±0.3°C, 7-year battery) - one per fridge minimum, two for any unit over 200 litres, plus a mapping campaign with up to nine probes during validation.
- VS-D300 door sensors on every refrigerated unit - door telemetry is the difference between a closed CAPA and an open one.
- VG-O500 outdoor industrial gateway with cellular backhaul - relevant for pharmacy chains running cross-site monitoring or for warehouses with poor indoor coverage.
- Calibration cycle - 12 months for in-use sensors, with a fresh UKAS cert filed against the sensor ID in the dashboard.
For broader cold-chain deployment patterns including reefer and warehouse use cases, see our field guide for UK cold-chain operators and the full product catalogue.
How long must records be retained, and in what format?
UK GDP requires records to be retained for at least one year beyond the expiry of the last medicinal product they relate to, and "in any event for at least five years" (MHRA Orange Guide, 2022). For dispensing pharmacies under GPhC, two years is the minimum but five is the prudent floor if any wholesale activity is involved. Format must allow inspector readability without proprietary tooling.
Practical implications: CSV or PDF export must work for any date range, sensor metadata (calibration cert, location, asset ID) must travel with the readings, and the archive must survive a vendor relationship ending. We deliver an annual signed PDF audit pack alongside the live system so pharmacies are never dependent on continued software access to defend a five-year-old reading.
Frequently Asked Questions
What is the MHRA Orange Guide and is it legally binding?
The Orange Guide is MHRA's compendium of UK law plus the GMP and GDP guidelines that have force of guidance under the Human Medicines Regulations 2012. While the guide itself is guidance, the regulations it interprets are statutory. MHRA inspectors assess compliance against the Orange Guide's expectations.
Do I need a wholesale dealer's licence to be GDP-compliant?
Only if you wholesale - that is, supply medicines to anyone other than the patient. A dispensing pharmacy that occasionally transfers stock to another pharmacy may need a WDA(H) licence, at which point full GDP applies. MHRA publishes licensing guidance for borderline cases.
How quickly must a temperature excursion be reported?
GDP does not set a fixed hour-count. It requires that excursions are detected, investigated, and managed in a manner proportionate to risk. In practice, inspectors expect real-time alarming, acknowledgement within the shift, root-cause investigation within 48 hours, and CAPA closure within a defined period set by the quality system.
Conclusion: every degree, every door, every audit-ready
MHRA compliance is not about owning the most accurate sensor in the room. It's about being able to prove - on the inspector's clock, in the inspector's format - that every degree was recorded, every door was logged, every deviation was investigated, and every record will still exist in five years. The hardware is the easy part. The architecture is the rest.
If you're scoping a new deployment or replacing a legacy logger fleet, the typical engagement looks like: site walk, mapping study, sensor and gateway specification, system qualification pack, and inspector-ready dashboard configuration. Most UK pharmacy clients see audit preparation time fall by around 92% within the first year (IoT-WorkS Cold Chain industry data, 2025). Start with the cold-chain industry page or talk to our engineering team about a site assessment.
IoT-WorkS Engineering Team - UK-based IoT systems engineers specialising in regulated cold-chain monitoring for pharma, food, and life sciences. We deliver MHRA-aligned deployments with full IQ/OQ/PQ documentation.