Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Health information exchange security protects the interfaces, standards, and trust arrangements that move patient records between organisations, where identity matching errors cause clinical harm.

What Is Health Information Exchange Security?
Patients move between organisations and their records have to follow. A GP refers to a hospital, a hospital sends results to a laboratory, an ambulance service needs allergies before arrival, a regional shared record pulls from a dozen systems at once. Health information exchange is the plumbing that makes that possible, and securing it is a different discipline from securing any one of the systems it connects, because the trust decisions happen between organisations rather than inside them.
HL7 version 2 still carries the majority of clinical messaging worldwide. It is a pipe-delimited format normally transported over a minimal protocol on a private network, with no authentication, no encryption, and no integrity protection in its base form. It works because everyone agreed the network was trusted, an assumption that stopped being true some time ago.
FHIR, the modern successor, is a REST API with resources, OAuth 2.0 authorisation, and a defined approach to scopes through the SMART application framework. That is a substantial improvement, and it means health interoperability inherits the whole of API security: token scope that is too broad, missing checks that a request for a patient’s record belongs to that requester, and endpoints exposed without the gateway anyone assumed was in front of them.
IHE profiles supply the cross-organisational layer: document sharing and query across community boundaries, patient identity services, user assertions carried between domains, and a consistent audit trail with node authentication. DICOM moves imaging, and it carries an unusual amount of identifying metadata inside the files themselves.
The first is identity matching. Linking the wrong records together is the most dangerous failure in this field, because it puts one patient’s allergies, medication, and history into another patient’s chart. It is also mundane: similar names, transposed dates of birth, duplicate registrations, and partial identifiers produce false matches at scale. Matching algorithms need measured false positive and false negative rates, manual review for ambiguous cases, and a defined process for unmerging records when a mistake is found. Treat it as a safety-critical control with the governance that implies.
The second is exposure. Imaging servers have a long history of being reachable from the internet without authentication. Research published by Greenbone Networks in 2019 identified millions of patient imaging studies accessible worldwide through unprotected servers, complete with names, dates of birth, and clinical detail embedded in the files. The cause was never exotic: a server stood up for a project, a firewall rule added for a supplier, an appliance with its default configuration intact. Internet-facing exposure assessment of the whole clinical estate, repeated regularly, finds these before researchers do.
An exchange works because participating nodes accept each other’s assertions about who is asking and why. That trust needs structure: mutual certificate-based authentication between nodes, signed assertions carrying the requesting user’s role and purpose, and a federated audit trail so that an organisation can answer what was accessed from its records and by whom, even when the access happened elsewhere. Without federated audit, each participant can only see its own half of an interaction, which makes investigating inappropriate access across the network effectively impossible.
Consent is the hardest part. Where a patient has restricted the sharing of some information, that restriction must travel with the data and be honoured by systems the originating organisation does not control. In England, the National Data Opt-Out governs secondary uses of confidential patient information, shared care records operate under local information sharing agreements, and national services route through NHS infrastructure with its own access controls. In Turkey, the consolidation described under e-Nabiz and MHRS puts the permission decision with the patient directly, which is a different architecture reaching for the same goal.
A breach here is a personal data breach under UK GDPR involving special category data, reportable to the ICO within 72 hours where individuals’ rights are at risk, and the organisations involved must establish quickly which of them is the controller for the affected processing. Those roles are discussed under controller and processor, and they should be documented in the sharing agreement rather than negotiated during an incident. At Privia Security, we assess interoperability estates and their exposure through our infrastructure security testing services.
Because in its base form it has no authentication, encryption, or integrity protection. It was designed for a private clinical network where every connected system was assumed trustworthy. Any device able to reach the listening port can send a message that will be processed as clinical data. Transport-layer security and endpoint authentication can be added, and frequently have not been on long-standing interfaces.
It is a patient safety incident. Merged records place one person’s allergies, medication, and diagnoses into another person’s chart, where clinicians will rely on them. Matching needs measured error rates, manual review of ambiguous cases, and a documented unmerge process, and the governance should reflect its safety criticality rather than treating it as a data quality issue.
Through ordinary operational decisions: a server deployed for a project, a firewall rule opened for a supplier and never closed, an appliance left with default settings, or a research system outside the main estate. Research in 2019 found millions of patient imaging studies reachable this way worldwide. Regular external exposure assessment across the whole clinical estate is the control that finds them.
It must, and making that work is the hardest part of exchange design. A restriction applied in one organisation has to be honoured by systems that organisation does not control, which requires the preference to be carried with the request and enforced at the receiving end. Arrangements differ: England uses the National Data Opt-Out and local sharing agreements, while Turkey places the permission decision with the patient directly.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.