Is Europe ready for the European Health Data Space?

The European Health Data Space (EHDS) is one of the most ambitious health data initiatives ever launched in Europe. From 2029 onwards, it will fundamentally change how health data is shared for patient care, research, innovation and public policy. Based on extensive analysis and real-world testing, the AIDAVA project raises a critical question: can EHDS be implemented at scale, affordably, and fairly across Europe?

What EHDS requires from hospitals

EHDS places hospitals and healthcare providers at the very centre of implementation. As data holders, they are legally required to:

Provide patient data for cross-border care in the European Electronic Health Record Exchange Format (EEHRxF)
Maintain and self-certify data quality through the QUANTUM label
Describe their data sources using standardized metadata catalogues
Respond to data access requests for secondary use via Health Data Access Bodies (HDABs)
Support secure processing environments for research and policy analysis

In practice, this means hospitals must extract, curate, validate, and continuously update high-quality, interoperable data from complex and often outdated IT systems. Most of this work must be repeated for every new request, dataset, or reporting obligation.

For many hospitals, especially small and medium-sized ones, this represents a substantial new operational burden with unclear financial return.

Why data quality and interoperability are still broken

European health data is rich, but it is not ready.

Across hospitals and healthcare systems:

Data is fragmented across many systems (EHRs, lab systems, imaging, pharmacy, legacy databases)
Around 80% of health data is unstructured or semi-structured text
Documentation of data sources is often incomplete or outdated
Redundancies, inconsistencies, and errors are common
The same standards (FHIR, SNOMED, LOINC, ICD) are implemented differently across countries and vendors

As a result, health data cannot be reused without extensive manual curation. Today, interoperability is largely reactive: data is cleaned, mapped, and transformed only when needed, often from scratch.

This approach does not scale to EHDS-level requirements and makes reliable data reuse slow, expensive, and error-prone.

Why cost and maturity differ across countries

EHDS assumes a level of digital maturity that does not exist evenly across Europe.

Some Member States operate centralized health data infrastructures with strong semantic standards and national governance. Others rely on fragmented hospital systems, limited interoperability, and scarce technical expertise.

AIDAVA’s analysis shows that:

Digital maturity varies significantly across Member States
Countries with centralized systems are better positioned to meet EHDS requirements
Hospitals in lower-maturity systems face higher costs and longer timelines
Technical connectivity (e.g. MyHealth@EU) does not guarantee semantic interoperability

This means that the same EHDS obligations can cost different amounts, depending on national context and hospital type. Without targeted support and automation, compliance risks becoming disproportionately expensive for less mature systems.

Why this risks a digital health divide

If EHDS implementation follows current approaches, Europe risks creating a two-speed health data space where Some hospitals struggle to comply or delay participation
Data quality and availability will differ across regions
Patients may receive uneven quality of cross-border care (because of uneven data quality)
Research and AI models will rely on biased or incomplete datasets
Innovation will concentrate on countries where data is already strong

Rather than reducing inequalities, EHDS could unintentionally amplify existing digital and health disparities between countries, regions, and healthcare providers, going account the fundamental principles of the EU of (digital and heal) equity and inclusion.

How AIDAVA’s digital twin approach changes the equation

AIDAVA proposes a different foundation for EHDS implementation: AI powered maintenance of a high-quality, interoperable and reusable digital twin of each patient's health record.

This digital twin is stored within each patient, Personal Health Knowledge Graph (PHKG), created and maintained locally by each data holder. Orchestraing a set of AI and non-AI tools for curation, quality enhancement and publishing tools.

AIDAVA:

Extracts data from heterogeneous hospital systems
Harmonizes heterogenous data in variable formats (paper, electronic, structured, semi-structured, narrative, images) and standards (LOINC, ICD, SNOMED, HL7,..) into a single interoperable format, the PHKG compliant with the AIDAVA reference ontology
Detects and corrects inconsistencies and redundancies on the PHKG by triggering data quality checks and therefore ensuring data quality by design
Publishes data automatically in different format from a single PHKG (identifiable data, including the EHDS critical data categories) or from multiple PHKG (anonymous data for registries)

This approach shifts interoperability from a manual task done just-in-time and repeated many times whenever data is needed, to a one-time, proactive process supporting any output (with automated transformation tools). Indeed once the digital twin is in place within an hospitals:

Data quality labels become easier to maintain
Data becomes immediately reusable for care, research, and AI including
Automatic generation of EHDS compliance data such as critical categories in EEHRxF
Smooth extraction and transformation - without any curation - of data across all relevant hospital running a PHKG, in answer to a HDAB query

Crucially, this reduces costs, limits the need for specialized skills, and makes EHDS compliance more realistic for hospitals across Europe, regardless of their starting point.

Why this matters now

EHDS is not just a regulation.It is infrastructure and, if successful, it can boost quality of care in EU while decreasing its cost. Whether it succeeds or fails depends on whether Europe can move from fragmented, manual data handling to automated, high-quality interoperability at scale.

AIDAVA’s work shows that this transition is possible — but only if implementation realities are addressed early, openly, and collectively.