The NHS holds one of the most valuable clinical datasets in the world. The challenge has never been a shortage of data — it's that most of it sits locked in proprietary formats, accessible only through expensive complex interfaces that can be difficult to build and maintain. The growing conversation around an NHS Shared Patient Record brings it back into sharp focus — the ambition of a coherent, longitudinal record spanning care settings is compelling, but it raises hard questions about data standards, governance, and how the existing system landscape connects to something that large and complex. What has been missing, perhaps, is a clinically credible data standard that the whole ecosystem can build around. OpenEHR is one candidate worth serious attention, and it's something I've been exploring at ProtoFlex.
What is OpenEHR?
OpenEHR is an open specification for storing and exchanging electronic health records. Where standards like FHIR define how data moves between systems, OpenEHR defines how clinical information should be modelled and persisted. It uses archetypes — structured, clinician-authored definitions of clinical concepts such as observations, assessments, and care plans — to create a canonical data model (CDM) that is vendor-neutral and internationally maintained.
The appeal in an NHS context is that it doesn't require replacing existing systems. Instead, it can act as a shared data layer that individual systems write to and read from, each continuing to develop the specialist tooling that makes them valuable in their own domain.
OpenEHR is already in production at national scale in Norway and parts of Sweden and Australia - I sat in on an interesting talk that at Rewired last year that discussed it. There is growing interest within NHS trusts and integrated care systems, and the NHS's current interoperability agenda creates a more receptive environment than has existed before.
Where This Might Fit for ProtoFlex
ProtoFlex is a platform for configuring clinical pathways and processes, capturing structured records of activity as patients move through care. Getting that data into other systems — or pulling data from other systems into ProtoFlex views — has historically required bespoke integration work on both sides.
This is already on our roadmap, and I've done enough with it to know it's worth pursuing. Last year I put together an initial proof of concept using an adapter pattern that intercepts saves within ProtoFlex and writes data to an OpenEHR repository alongside our native data store. The core mapping of data items to archetypes was more straightforward than I expected. The harder problem, which we'll need to work through when I come back to this, is getting accurate mappings right where data is recorded across more complex workflows: branching protocols, conditional steps, pathways where the context can shift depending on which route a patient takes.
The potential benefits, if we get that right, are interesting:
- Operational data availability — records from pathways becoming available to other authorised systems in near real-time, via the CDM rather than bespoke extracts
- Analytics and research — compatible tools querying the CDM directly, reducing the data engineering burden that currently makes NHS research access so slow
- Consuming data from other systems — if other systems are also writing to the CDM, ProtoFlex could surface that data in its views without custom mapping work on either side
- Reduced interfacing overhead — fewer point-to-point integrations to build and maintain across a trust or ICS
None of this is straightforward. Archetype governance requires genuine clinical engagement, trusts need CDM infrastructure, and the benefits only compound meaningfully when critical mass is reached across the ecosystem. We're not claiming this is the answer — but it feels like the right question to be asking.
If OpenEHR is on your radar — whether you're in an NHS informatics team, working in a trust, or building in the same space — I'd be happy to compare notes.