I’d be interested in openEHR community’s thoughts on the NIH initiative to create a standard set of “Common Data Elements” mainly for the purpose of having a common set of data elements for clinical research and trials. There’s a recent RFI with good description. As I can see it’s a small set of very specific atomic data elements (Cost Delayed Healthcare Occurrence Indicator) but they also create bundles to create concepts like Vital Signs!
Obviously there’s a lot of overlap / reinvention with not just openEHR but FHIR, OMOP, ISO/IEC 11179 based METEOR, SNOMED, LOINC etc.
You can use openEHR for research, all the models are available —much more than what you find in a typical common data model or in OMOP. EHRbase is also open source. You could also use FHIR for that depending on what you want, resources are in the end a common data model. On top of that, we already have OMOP.
So, why would you need this? Does it provide anything that openEHR or OMOP doesn’t?
Thanks @SevKohler , I personally have no intention to use it! It came up during a research meeting and I promised to study it and prepare a robust response. I’m hoping to gather excellent insights from this forum.
I always wonder why people do that, i mean you can just use openEHR (or OMOP if its about having more research tooling) and have 0 effort to define most of the stuff you need.
Less resources required, and the template are semantically scalable, ah we need more fields like this and that yeah sure just add them in they are provided by the archetype (if they are).
AQL …
Looking forward to your findings, also would be nice to know why they do this lol.
Agree.
There are overlapping resources everywhere, and the key is what you want to achieve, and then choose the right resource(s)/standard(s). Each standard has its own intended purpose and scope of application.
Research requires a lot of data that is not collected in EHR systems. This is the reason there need to be standards for NIH research data. Our challenge is to try to insure that the 1) Research data elements are interoperable to the fullest extent possible with the EHR data elements and 2) interoperable across NIH.
We are working on these objectives at NIH, given the constraints and specific requirements for each research project.
Hi @warzeld I looked into OMOP in more detail and understood the use cases better now. It’s a pragmatic and simple way to manage and access research data. I still have some reservations as to how different data types are organised in a relational database format but I’m guessing most end users would already be familiar working with relational databases using SQL etc. At any rate thanks for your feedback. And good luck with budget cuts!
