OMOP vs openEHR: What Good is AI Without Good Data?

Why This Debate Matters for the Future of Ethical AI in Healthcare

By Health Data Forum
📍 Cardiff, 24–26 September 2025
🔗 Join the livestream

As artificial intelligence accelerates its influence on healthcare—from clinical decision support to population health analytics—there is a quiet but crucial architectural decision underpinning it all:

How are we structuring the data that will shape the intelligence we deploy?

Two of the most prominent data models used in healthcare—OMOP and openEHR—offer contrasting approaches to this foundational question. At the Health Data Forum Global Hybrid Summit 2025 in Wales, we invite our global community to move beyond the hype of AI and focus on the integrity of the systems that feed it.

OMOP: Standardized for Analysis

Originally developed through the Observational Medical Outcomes Partnership, OMOP is widely used in research networks like OHDSI. It provides a relational data schema and common vocabulary, enabling the harmonization of large, siloed datasets for multi-center analytics.

✅ Great for: epidemiology, real-world evidence (RWE), large-scale data mining
❌ Limitations: rigid structure, loss of clinical context, post-hoc harmonization

openEHR: Modeled for Care

openEHR is built on a semantically rich, archetype-based approach. It is designed to support clinical recording at source, enabling fine-grained, reusable, and interoperable clinical data models that reflect real-world care.

✅ Great for: longitudinal EHRs, clinical decision support, interoperability
❌ Limitations: less mature in research analytics, implementation complexity

Why Does This Matter for the "Data First, AI Later" Movement?

The Health Data Forum co-leads the "Data First, AI Later" movement alongside Health Parliament India and other global partners. This movement is built on a simple but urgent idea:

Before we build artificial intelligence, we must first build data integrity.

Choosing between OMOP and openEHR is not a technical footnote—it is a reflection of values:

Do we treat data as a secondary byproduct of care, or as a primary asset of trust?
Are we building systems for analysis, or for care—and can we do both?

A Constructive Tension at the Heart of Cardiff 2025

We won't reduce this to a binary choice. In fact, the future may require both. openEHR may serve as the semantic backbone for data capture, while OMOP can remain the workhorse for analysis and regulatory research.

These architectural questions will be brought to life in sessions like:

🧩 "OpenEHR and AI: Driving Data-Driven Healthcare" – with Rachel Dunscombe, CEO of openEHR International and former NHS CIO

📊 "Real-World Evidence for the Real World" – a keynote by Tiago Taveira Gomes, CEO of Sigil Scientific Enterprises, exploring the gap between data systems and clinical reality and how to close it through AI and real-world data.

Meet the Speakers

Rachel Dunscombe

Professor Rachel Dunscombe, CEO of openEHR International, is a globally recognised digital health leader with over 30 years of experience. Previously CEO of the NHS Digital Academy (2017–2022), she led the Salford/NCA Group to become the NHS's most digitally mature organisation. A Visiting Professor at Imperial College London, she has advised the UK Secretary of State for Health, served on the UK AI Council, and sits on the Digital Health Society Board. Her work drives interoperable, open-source health data standards for improved patient care.

Session: Panel Discussion – Who Owns Cloud Data? Infrastructure and Governance
Thursday, 25 September 2025, 11:30–12:15 BST
With Prabs Arumugam (AWS), Rafael Almeida (SPMS), and moderated by Ifan Evans (DHCW)

Tiago Taveira-Gomes

Dr. Tiago Taveira-Gomes, MD, PhD, is the Founder of MTG Research and Development Lab, OPVANCE DevSecOps, and Sigil Scientific Enterprises. A specialist in Family Medicine with a PhD in Medical Education & Medical Informatics and a Master's in Information Engineering, he bridges clinical practice with AI and software innovation. He serves as an Adjunct Professor at the University of Porto and other institutions, focusing on real-world evidence, implementation science, and AI-driven healthcare solutions.

Keynote: Advancing Personalised Healthcare with Real-World Evidence and AI
Friday, 26 September 2025, 09:10–09:30 BST

Our Take: Don't Pick a Model. Pick a Mindset.

At Health Data Forum, we are not evangelists for one model over another. We are facilitators of a deeper conversation:

What kind of data culture are we cultivating?
How do we balance semantic depth and statistical breadth?
Who gets to decide how health data is captured, structured, and used?

These questions don't live in code—they live in policy, design, and trust. And that's what Cardiff 2025 is all about.

Join Us in Wales

Dates: 24–26 September 2025

Venue: voco St. David's Hotel & Cardiff Innovation Campus
On-site registrations are now closed. But we've opened universal free access to all sessions via livestream.
🌐 Watch live on Zoom: Register here.

#OMOPvsOpenEHR #DataFirstAILater #HDFWales2025 #RachelDunscombe #TiagoTaveiraGomes #EthicalAI #DigitalHealthInfrastructure #HealthDataMatters