Patient Diary De-Identification with anonym.plus

Clear names and places from the diary entries while the symptom log stays.

Patient diary de-identification is the removal of personal detail that a participant writes in free text. It supports GDPR Art. 9 on special-category health data. anonym.plus runs offline and keeps the symptom record readable.

When this applies

A team analyses daily symptom entries. Participants often write a spouse's name, a town, or a workplace in the free-text fields.

How anonym.plus handles it

  1. Open the diary export (CSV, DOCX, or PDF) in anonym.plus.
  2. The tool scans every free-text entry for names and places.
  3. Local OCR reads a scanned paper diary if you add one.
  4. Confirm the flagged people, places, and contact details.
  5. Replace each with a neutral label, or remove it.
  6. Save the cleaned entries locally with no upload.

What you need to provide

PHI entity types detected

Categoryanonym.plus entity typeExample
Self / familyPERSON"my husband Olav" → [RELATIVE]
PlaceLOCATION"at work in Aalborg" → [PLACE]
DatesDATE_TIMEfelt dizzy 12 May → [DATE]
WorkplaceORGANIZATIONthe Aalborg depot → [WORKPLACE]
ContactPHONE_NUMBER+45 555 0177 → [PHONE]
EmailEMAIL_ADDRESSdiary@example.dk → [EMAIL]

Compliance achieved

Anonymize patient diaries offline — see plans & start free →

Limitations & cautions

Diaries are pure free text, so phrasing varies a lot. A nickname or an odd spelling can slip past once in a while. Always read the flagged entries before export, since a rare personal detail in a sentence can still identify the writer.

Frequently asked questions

Why are diaries special-category data?

They record health symptoms tied to one person, which Art. 9 treats as sensitive. They also hold free-text personal asides. Cleaning both is needed before the entries are analysed or shared.

Does the symptom log survive?

Yes. The ratings, the dates of symptoms, and the notes about how the person felt stay. Only names, places, and contacts are swapped.

How does it handle nicknames?

It flags likely names from context. A rare nickname may not be caught, so a human review of the entries is still important.