Patient records: show why an account accessed the record
An authorised account can open a patient record without a legitimate reason. The ICO warning helps a clinic define the evidence needed to assess that access.
Czytaj po portugalskuBy Pentesty
What the warning establishes
The ICO announcement records more than 300 unauthorised-access reports received and assessed from the UK health sector between January 1 and mid-October 2026. Reports awaiting assessment are excluded. The total does not count patients, external intrusions or confirmed violations.
The regulator wrote to more than 200 health-sector leaders. Its November 12 letter seeks assurances about permissions, monitoring, training and responses to inappropriate access. It asks for confirmation of a recent review or a review where needed, while preserving existing legal responsibilities. For Brazilian clinics, this is a management reference; the UK exercise does not automatically apply to them.
Account permission and the purpose of a lookup
Pentesty's analysis starts with an operational distinction: being able to open a record does not explain why it was consulted. A receptionist may need to confirm an appointment; a care team may need to follow a patient. The account's role should be examined alongside the activity that would justify access.
Choose a bounded workflow, such as a referral between two teams. Identify who grants permission, when it should change and who can explain a disputed lookup. This helps locate where responsibility becomes unclear between the clinic, a service provider and the system supplier, without assigning blame in advance.
Include cover arrangements, shifts and urgent care in the review. A lookup outside usual hours may have a care-related justification. The process should let the responsible person provide that context, rather than treating an exception as sufficient proof of abuse or obstructing care to produce an audit result.
A small review that produces useful evidence
Agree an authorised assessment with clinical, operational and privacy owners. Where possible, use test accounts and synthetic records to check an expected lookup and an attempt outside the permitted scope. Set boundaries, stopping conditions and evidence handling in advance. Curiosity about a real patient is not a valid testing method.
Propose a minimal trail connecting an identifiable account, the record accessed, time, action and work context. Establishing purpose may require checking another source, such as a referral, without copying clinical content into the report. These fields are a Pentesty assessment proposal; they do not imply that the ICO requires a particular log format.
Check whether the team can gather that information using the access it already holds. An event may record a screen opening without showing which data was viewed. Document that limitation and distinguish what the system records, what the team could verify and what remains unanswered.
Check context before alleging abuse
An alert should begin a proportionate inquiry. Before concluding that access was inappropriate, check account attribution, shift arrangements, the connection to care and the reliability of timestamps. Shared accounts, poorly documented delegation or integrations can make it difficult to attribute an action to a person.
Keep the explanation and evidence that challenges an allegation under restricted access, with a named owner and retention period suited to the scope. Record uncertainty when information is missing. Refer consistent indications to the appropriate decision-makers. A technical review alone determines neither blame, notification duties nor disciplinary action, and cannot guarantee compliance.
What leadership should be able to demonstrate
The letter asks recipients to confirm a review within the previous six months and, if none occurred or the arrangements are unsatisfactory, to review them within the following three months. November 27, 2026 is the assurance response date for the letter's recipients. That timetable is not a new general deadline for hospitals or clinics.
A useful assessment ends with a supported decision: which access was tested, what evidence explained it, what gap remains and who owns remediation. Repeat the check after relevant workflow or permission changes. This makes it possible to evaluate a concrete improvement without confusing the existence of a policy with evidence that it works.
Frequently asked questions
Can an authorised account make an inappropriate lookup?
Yes. An account may have technical permission to open the system without every lookup having a legitimate purpose. Assess the record accessed, the activity and the context while preserving access needed for care.
Are the 300-plus reports confirmed violations?
The ICO describes reports received and assessed between January and mid-October, excluding those awaiting assessment. The figure does not give a patient count or establish that each report resulted in a confirmed violation.
Does the November 27 deadline apply to every clinic?
The date concerns replies from recipients of the UK letter. The document says the exercise neither replaces nor changes existing legal responsibilities. It is not a new general deadline applying to Brazilian clinics.
How can a test avoid exposing real patient records?
Agree an authorised scope, owners and stopping conditions. Prefer synthetic accounts and records, collect only necessary evidence and preserve urgent-care access. The test should assess controls without creating unjustified access to clinical data.
Sources and references
- ICO: warning to the health sector, November 12, 2026
- ICO: letter on protecting patient records, November 12, 2026
Original Pentesty analysis, published November 15, 2026. The proposed scenarios are editorial guidance, with scope and validation to be defined for each environment.