This article paints the picture of the recent evolution of OpenEHR out in the real world.
It is best understood within the context of the multiple large proprietary EHR solutions out there, and the underlying immense challenges of defining and standardising health data and the need for interoperability by design.
(changed to a link by which one of the authors has enabled sharing)
Interesting read. I’ve also found the discussion thread that @NathanK started around the use of AQL, rather than SQL, in openEHR.
It makes the (perhaps) obvious case that AQL is a better equipped to query archetypes than SQL - but both make heavy weather of what would be a simple query (obtaining all body temperature measurements for a patient if the measurement is >37.5°C) in a relational database.
I like this quote:
Using a logical query language like this makes an openEHR datastore tech-neutral i.e id I change my CDR provider to use a new faster/cheaper product which e.g. uses MongoDb rather than a relational database, my application or reporting queries do not change.