On April 20, Boris Delange (MD, Medical Informatics, Université de Rennes) presented the INDICATE Data Dictionary at the OHDSI Europe Symposium 2026, the yearly meeting of the European OHDSI community, gathering researchers, clinicians and data scientists working on federated health data using the OMOP common data model.

Through a live demo, Boris showed how the Data Dictionary supports multidisciplinary teams working with medical concepts in OMOP — by providing peer-reviewed, versioned concept sets enriched with clinical context and ETL guidance.

Why does this matter? Before any federated analysis can run across hospitals, each site must map its local data to the same shared vocabulary. This “concept mapping” step is essential but notoriously time-consuming, and small inconsistencies between sites can silently bias results. By offering a curated, transparent library of ICU concept sets – with review workflows, semantic versioning, and expert comments – the INDICATE Data Dictionary makes this step faster, more reliable, and easier to share across institutions.

The session also sparked valuable discussions on integration within the OHDSI ecosystem and on extending the approach beyond ICU.

The Data Dictionary is open source and runs entirely in the browser: no server, no login, no install. Anyone can browse the 300+ ICU concept sets, propose reviews on GitHub, or deploy their own instance in minutes.

Explore the INDICATE Data Dictionary via GitHub.

The first session of the INDICATE Training Programme on Legal Framework kicked off! These sessions are running in parallel with and complementing the ongoing Data Models sessions.

This training programme provides participants a comprehensive understanding of the INDICATE legal framework, covering General Data Protection Regulation (GDPR) and European Health Data Space (EHDS) principles. The sessions focus on data protection and privacy-enhancing technologies, governance and rulebook structures, and practical skills to navigate data access processes, requirements, and organizational challenges within INDICATE. 

The first session, led by Ricard Martínez Martínez (Universitat de Valencia), explored how law, technology, and organisation together shape the use of health data in research. A key message was that rules are important, but it is not enough on its own. Trust and responsibility are just as important.

Several important topics were discussed:

• Trust and reputation matter in research. It is not only about following the law. Organisations must also show clearly that they handle data in a responsible way. Without trust, collaboration and research can be at risk.
• Clear governance and roles are essential. In complex data systems, it must be clear who is responsible for what. Researchers, project managers, data protection officers, and platform teams all have different roles. Without clear responsibilities, risks increase.
• Anonymisation is more complex than it seems. Making data truly anonymous is not easy. It requires continuous risk assessment, technical measures, and careful monitoring to reduce the risk of re-identification.
• Balancing innovation and privacy. In healthcare research, there is a need to learn from data while protecting patient privacy. Data minimisation does not mean using as little data as possible, but using the right data for a clear purpose.
• Privacy-enhancing technologies are becoming essential. New technical solutions allow researchers to analyse data without directly accessing raw data. This helps protect privacy while still enabling valuable research.

The session also highlighted that European developments, including the European Health Data Space and emerging AI regulations, will strongly shape how health data research is organised in the coming years. Active engagement from the research community is therefore essential.

A key takeaway from the session is that legal rules and regulations, technical design, security, and ethics are deeply connected. Real progress in health data research happens when all of these elements work together, supported by a strong culture of responsibility and trust.

The INDICATE project receives funding from the European Union’s Digital Europe Programme under Grant Agreement number 101167778.

The second training of the INDICATE Training Programme on Interoperability, OMOP and Vocabularies took place on April 9, 2026. The programme is designed to support data providers in using the INDICATE infrastructure effectively, securely, and in a fully standardised way. It helps participants, such as clinicians and data engineers, build both the conceptual understanding and practical skills needed to work with interoperable health data.

During this second training, led by Maxim Moinat (Researcher, Medical Informatics, Erasmus MC) and moderated by Boris Delange (MD, Medical Informatics, Université de Rennes), participants learned that data from different hospitals and institutions must be made interoperable to enable research at a European level. However, this is only possible when data is structured in a way that makes comparison meaningful and reliable.

This is where standardisation becomes essential. Without a shared structure, data remains fragmented across systems, making large-scale analysis difficult or even impossible. By harmonising data into a common format, researchers can generate evidence that is consistent, reproducible, and scalable across countries.

The OMOP Common Data Model provides exactly this; a shared way of organising patient data and a shared vocabulary for describing clinical events, so that hospitals across Europe can describe the same reality in the same terms. Maxim walked participants through the main building blocks of the model and showed how they apply to ICU data, with concrete examples detailed during the session. He also presented the wider OHDSI community and European networks such as EHDEN and DARWIN EU, which already federate data on hundreds of millions of patients.

Maxim then walked participants through the full journey from raw hospital data to interoperable, OMOP-formatted data, step by step, from the initial exploration of the source system to the final validation of the mapped database. At each stage, he introduced the corresponding tools from the OHDSI ecosystem, a suite of open-source resources designed to support data providers throughout the process. He also showed how the INDICATE Data Dictionary, presented in Session 1, fits into this journey by guiding data providers on which clinical concepts to prioritise for mapping.

The session concluded with key take-home messages on the importance of clear mapping specifications, vocabulary alignment, and the value of a shared data model for enabling collaborative research across institutions and countries.

Overall, the training provided participants with both a conceptual and practical understanding of how the OMOP CDM and the surrounding OHDSI ecosystem support interoperable and scalable health data research within INDICATE.

The next training will focus on the ETL Workflow, data preparation requirements, and data quality expectations and is planned on May 7 2026. 

Read more about the first training.

In order to support all consortium members in using the INDICATE infrastructure effectively, correctly, and securely, we are organizing a three-session series of the INDICATE Training Programme on Legal Framework, running in parallel with and complementing the ongoing Data Models sessions. 

The programme will give participants a comprehensive understanding of the INDICATE legal framework, covering GDPR and EHDS principles, data protection and privacy-enhancing technologies, governance and rulebook structures, and practical skills to navigate data access processes, compliance requirements, and organizational implementation challenges within INDICATE.

Session dates

All sessions will be held from 14.00 – 16.00 (CEST) via Zoom.

• May 4 – Session 1 | Understanding GDPR
• June 24 – Session 2 | Understanding and using Data Access
• September 10 – Session 3 |  Understanding the Rulebook and legal onboarding steps

This week marked the first session of the INDICATE Training Programme, designed to support data providers in using the INDICATE infrastructure effectively, securely, and in a fully standardized way.

The session, guided by moderator Maxim Moinat (Data Engineer, Erasmus MC) and co-moderated by Maarten Ligtenberg (Co-founder Cradeq), provided a solid introduction to key building blocks on the INDICATE onboarding framework, interoperability in complex healthcare data environments and federated data infrastructure and secure data sharing principles.

Jan van den Brand (technical lead INDICATE) highlighted key challenges in ICU clinical decision-making and innovation, driven by fragmented data, a lack of standardized data-sharing agreements, and limited secure infrastructure. He illustrated this using a metaphor: hospitals today resemble a house with different types of power sockets, where every device requires its own adapter to function.

In this analogy, medical and AI software represent the appliances, while hospital systems such as electronic health records and laboratory databases represent the power sources. Without a shared standard, hospitals are often forced to build and maintain these “adapters” themselves, increasing complexity, cost, and operational risk. This underlines the need for shared standards and interoperable data models.

A central theme, introduced by our presenter Boris Delange (Doctor in Medical Informatics, Université de Rennes), was the reality of hospital data: each institution often uses its own “language” to describe the same clinical concepts. This creates significant challenges for interoperability and data integration, while also highlighting the importance of standardization for enabling meaningful reuse of healthcare data in research and innovation. 

Boris also addressed the broader context of Hospital Information Systems and Clinical Data Warehouses, focusing on challenges related to data quality, semantic alignment, and making heterogeneous data usable beyond clinical care. Despite its value, a large proportion of hospital data (97%!) remains underutilized for research purposes.

INDICATE addresses this challenge by developing a federated data infrastructure, where data remains securely stored within its original institution (the data never leaves the hospital) while becoming interoperable and accessible for analysis across organisations  through shared standards.

The training programme consists of five sessions. The next session will take place on April 9, 14:00–16:00 CEST.

The training sessions are organised by Maarten Ligtenberg, Melania Istrate, Elisa Vera, Jan van den Brand, Aliza Bos, Maaike van Zuilen, and Irene Gebuis, a collaboration between Work Packages 1 and 5 and the INDICATE Training and Education Workgroup.