When a patient is referred to a specialist clinic, the receiving physician typically gets almost no information — just the referral slip and a short note explaining why the patient was sent. Electronic health records often contain only a few sentences of usable history: past analyses, examinations, treatments and laboratory results are scattered across providers, locked in non-interoperable systems, or never digitised at all.

The consequences are repeated tests, missed clinically relevant history, slower and lower-quality clinical decisions, and unnecessary cost. The problem is equally acute in Slovenia, the Netherlands and most other European health systems, because the root cause is not missing digitalisation but the missing intelligent selection, structuring and transfer of data that already exists.

The project will build a privacy-preserving platform that generates a structured, context-aware summary of the patient's health record at the moment of referral. Instead of transferring the full history, the platform intelligently selects and organises what is relevant to the reason for referral: completed analyses and examinations, past treatments, relevant laboratory results and trends, and tests that would be pointless to repeat.

An AI document-understanding module converts handwritten and non-digitised records into usable data. The core summarisation and transfer logic is shared across the consortium, while each partner adapts the solution to its national health system — making the project modular and resilient by design, with GDPR-compliant exchange and transparent access control as first-class requirements.

The consortium will develop a context-aware summarisation engine that ranks record content by relevance to the referral reason, together with multi-source health data collection and structuring for the receiving system based on HL7 FHIR and IHE interoperability. Trend highlighting, clinical alerts and duplicate-test prevention will surface what matters — for example flagging repeated laboratory visits in the past two months. An AI module for reading handwritten and non-digitised medical records will recover otherwise inaccessible history. Secure, GDPR-compliant health-data exchange with transparent access control will be designed in from the start, and national-system adaptation modules will be built per partner country. Referral-pathway pilots will run in Slovenia and the Netherlands, integrated with multidisciplinary council workflows.

• Open Line Vitaly d.o.o. — Industrial coordinator (project coordinator)