What was decided
On 23 September 2026 the Italian authority adopted decision no. 710 against IQVIA Solutions Italy S.r.l., made public on 2 October. The fine is seven million euro. The inquiry began after inspections in April 2025. The database covered around one million patients of some eight hundred general practitioners and was used for studies partly commissioned by pharmaceutical companies. Besides the fine, the authority ordered the processing to be brought into line within one hundred and twenty days if the activity continues.
The company's defence was the most common one in this field: the data is anonymous, so the GDPR does not apply. The reasoning is formally sound — recital 26 does say the Regulation does not concern anonymous information — and almost always wrong in fact. The point is not whether the name is there: it is whether the person can be found again.
The line between anonymous and pseudonymised
In the database each patient had a code. Not a name, a code. But that code was stable: it served to follow the same person over time, because that is exactly what a longitudinal study must do. An identifier that allows one person's data to be linked together is not an anonymisation measure: it is the definition of pseudonymisation in Article 4(5). And pseudonymised data is still personal data.
- Recital 26 asks you to consider the means reasonably likely to be used to identify: cost, time, available technology — not absolute impossibility.
- A stable per-person code allows events to be linked over time, and that is precisely what makes the data attributable.
- Year of birth, sex, diagnoses, prescriptions and a territorial reference, taken together, narrow the field down to very few people.
- Whoever claims data is anonymous must be able to demonstrate it, and the demonstration is technical: a re-identification analysis, not a sentence in a privacy notice.
- If the data is not anonymous you need a legal basis and a notice, and for health data also a condition under Article 9(2).
The question to ask the supplier, and the answer that does not count
When a supplier says it processes «anonymous data», the question is not «is it anonymous?» — they will say yes. The question is: «does the same patient have the same code in two extracts six months apart?». If the answer is yes, the data is not anonymous, whatever the contract says. It is a technical question answered in thirty seconds, and it separates a statement from a verification.
Then there is the upstream side. The data came from general practitioners. For a DPO advising a medical practice, a laboratory or a healthcare provider, today's question is whether any of their data ends up in research or commercial-information databases, on what legal basis, and whether patients know. The fact that a third party does the processing does not move the responsibility of whoever collected it.
What to take away, concretely. The word «anonymous» in a contract or in a register of processing is not a legal qualification: it is a claim that must be proved. Wherever it appears, it is worth asking three things — who verified it, by what method, and when. If those three answers are missing, that line of the register describes a processing operation believed to be outside the GDPR which is in fact inside it, with everything that follows.
Official source:Garante per la protezione dei dati personali — Health data: a seven million euro fine (2 October 2026)Official source:Cybersecurity360 — When health data is truly anonymous, according to the authorityOfficial source:Il Sole 24 Ore — Health data, IQVIA fined seven million euroLooking for a workspace for your DPO work?
DPO Workspace is built by a certified DPO. 30-day free trial.
Start free