DUR anonymisation is the removal of patient IDs from a utilisation dataset. Using identifiable NHS data without consent would need support under s.251 of the NHS Act 2006; properly anonymising the extract avoids that route entirely. anonym.plus does this locally, so the prescribing patterns stay while no row names a person.
When this applies
A utilisation dataset ties fill patterns to named individuals. A prescribing quality study needs the cohort to hold together across rows — the same shape the MHRA’s CPRD releases to researchers, minus the identities.
How anonym.plus handles it
- Point anonym.plus at the dataset on your device.
- It scans ID columns and any free-text fields.
- Patient names, NHS numbers, and dates get flagged.
- Confirm the flags; the medicines stay as non-IDs.
- Swap IDs with a steady map to keep cohort links.
- Save the clean data; the source stays on your machine.
What you need to provide
- The dataset as CSV, JSON, or a bundle of files.
- A column map for patient and NHS number fields.
- Replace with a steady map to keep cohort joins.
Patient data entity types detected
| Category | anonym.plus entity type | Example |
|---|---|---|
| Names | PERSON | patient_name → [PATIENT_n] |
| Record IDs | MEDICAL_RECORD_NUMBER | mrn field → [MRN_n] |
| Dates | DATE_TIME | fill dates → shifted [DATE] |
| Location | LOCATION | postcode → [POSTCODE] |
| Age | AGE | age 92 → [AGE_BAND] |
| Identifiers | NHS_NUMBER | NHS no. → [NHS_NUMBER] |
Compliance achieved
- Utilisation research fits UK GDPR Art. 9(2)(j) with DPA 2018 Schedule 1, Part 1, paragraph 4 and the safeguards in Art. 89(1).
- Using identifiable NHS data without consent would need support under s.251 of the NHS Act 2006 and the Health Service (Control of Patient Information) Regulations 2002 — proper anonymisation avoids that route.
- Mirrors how the MHRA’s Clinical Practice Research Datalink (CPRD) releases de-identified prescribing data for study.
- Age banding and postcode truncation follow the motivated-intruder test in the ICO anonymisation guidance.
Anonymise utilisation datasets offline — see plans & start free →
Limitations & cautions
Ages over 89 should be banded and a full postcode truncated, since both single people out in small groups. The medicines and patterns stay. Check that date-shift keeps the gaps your study needs before a full run. If you keep the alias map, the dataset is pseudonymised rather than anonymous and stays in scope.
Frequently asked questions
Can cohorts stay linkable after the swap?
Yes. A steady label map swaps each ID the same way, so an individual's rows still group while no real identity is left in the data.
How are very old ages handled?
The ICO anonymisation guidance treats very small groups as a heightened re-identification risk, and ages above 89 are one, so band them. The tool flags them for review.
Do I still need s.251 support?
Not if the extract is genuinely anonymous — s.251 of the NHS Act 2006 exists for identifiable data used without consent. Record the assessment that got you there.