Stream: omop + fhir
Topic: FHIR Terminology Services for OMOP
Michael Lawley (Nov 08 2021 at 21:42):
This is a brief report we put together based on initial exploration of how terminology is handled within OMOP and where we see opportunities for alignment with FHIR and exploiting its powerful and dynamic Terminology Service capabilities.
FHIR-OMOP-Opportunity-Report.pdf
Morten Ernebjerg (Nov 09 2021 at 06:39):
Hi @Michael Lawley, sounds interesting but the link seems to be broken.
Michael Lawley (Nov 09 2021 at 06:41):
It looks like the initial upload failed. I've re-attached it now.
Morten Ernebjerg (Nov 09 2021 at 06:42):
Thanks, works :confetti:
Jake Gillberg (Nov 09 2021 at 14:25):
Michael Lawley said:
This is a brief report we put together based on initial exploration of how terminology is handled within OMOP and where we see opportunities for alignment with FHIR and exploiting its powerful and dynamic Terminology Service capabilities.
FHIR-OMOP-Opportunity-Report.pdf
Ah, this is awesome!
I'd love to see a similar collected description of the FHIR Terminology capabilities. Does anyone know if this exists, and where I could find it? happy to be pointed to a collection of wiki pages that require a bit of digging as well.
One small question that I had: were the authors of this aware of the the OMOP concept_relationship table, and the other sorts of relationships that can be found between concepts? I think it addresses the below point. If not, I'd be interested to know why not.
[OMOP] is distinctly less powerful because it does not
retain information relating to “role grouping”. Hence it may not be possible to determine which
body structure a particular morphology applies to...
For example, SNOMED 275476006 [Foreign body under eyelid] retains the relationships "Has associated morphology (SNOMED) - SNOMED 125245002 [Retained foreign body]" and "Has finding site (SNOMED) - SNOMED 80243003 [Eyelid structure]"
Davera Gabriel (Nov 09 2021 at 14:57):
Hello @Jake Gillberg some of the example terminology use case resources I will show later today in the OHDSI CDM Vocab meeting you may find useful, albeit this will fall short of a complete picture I think you will desire. There are YouTube videos available with background & instruction on FHIR terminologies. Also @Rob Hausam has a workshop coming up in a couple of weeks where he provides terminology basics: http://www.hl7.org/events/webinar/fhir/terminology/2021/
Im looking forward to our meeting later today!
Jake Gillberg (Nov 09 2021 at 15:14):
Shoot, I don't know that I can swing that registration price.
Michael Lawley (Nov 10 2021 at 23:21):
@Jake Gillberg yes we were aware. A simple example of the issue would be a concept like 361209006 where there are two finding sites but the morphology only applies to one of them; pupura applies to the skin but not the peripheral vascular system. Our understanding was that this linkage between the relationships is lost.
Michael Lawley (Nov 10 2021 at 23:25):
For an intro to FHIR Terminology services there are videos on YouTube from DevDays by both @Rob Hausam and @Jim Steel (I really like this one https://youtu.be/40Lvv2t8OxU)
Rob Hausam (Nov 10 2021 at 23:36):
I really like Jim's presentation, too. It's very creative.
Jake Gillberg (Nov 11 2021 at 05:21):
Michael Lawley said:
Jake Gillberg yes we were aware. A simple example of the issue would be a concept like 361209006 where there are two finding sites but the morphology only applies to one of them; pupura applies to the skin but not the peripheral vascular system. Our understanding was that this linkage between the relationships is lost.
Ah, okay, this makes sense now. Thanks for clearing that up! and for the links!
Last updated: Apr 12 2022 at 19:14 UTC
