Stream: Covid-19 Response
Topic: mcovid.org
Grahame Grieve (Apr 18 2020 at 21:26):
Has anyone here looked at these data sets? Seems to me that FHIR profiles (and v2 and CDA as well) are important. thoughts?
Keith Boone (Apr 19 2020 at 02:39):
Links to them on mcovid.org vocabulary page are broken, I'll track them down eventually.
Keith Boone (Apr 19 2020 at 02:41):
@Mark Kramer or @Chris Moesel Do either of you know who can fix the links on https://mcovid.org/docs/vocabulary/ page? Right now they land on VSAC home page.
Mark Kramer (Apr 19 2020 at 12:31):
I can fix them. Thanks for the heads-up.
Mark Kramer (Apr 19 2020 at 13:17):
OK, fixed: https://mcovid.org/docs/vocabulary/
Keith Boone (Apr 19 2020 at 15:39):
Thanks.
- To answer Graham's question: These meet many of the needs of the SANER Project for value sets that it would recommend in the IG for use with certain measures.
- It would be helpful if the vocabulary link above gave a direct link to the FHIR Value Sets via the NLM FHIR APIs.
- I expect we will include these in the FHIR SANER IG, expanded where possible so that others can readily incorporate them into their systems, with a note that they are current as of the time of publication, and that users should check with the original sources for updates (per the NLM license for reproduction).
- Ideally the NLM would report a canonical url for these value sets, that's a VSAC problem which I assume someone from the HTA could maybe point out to them.
- It would be awesome if the FHIR Terminology service would recognize and expand these.
Keith Boone (Apr 19 2020 at 15:43):
- Does anyone know who would we talk to in order to get other value sets released in this group, rather than having 3 or 4 different groups developing them separately? For example, I'm thinking that it might be good idea to have a valueset for Remdesivir (which seems to have had more success in small clinical trials than Hydroxychloroquine).
Grahame Grieve (Apr 19 2020 at 20:34):
I can't work with vsac in the publisher, since access requires a personal password. I've raised this as an issue before with ONC, but it hasn't made any difference
Grahame Grieve (Apr 19 2020 at 20:36):
this response solely concerns the vocabulary, not the data elements, right?
Bryn Rhodes (Apr 19 2020 at 21:14):
We have been working on profiles to support exactly this scenario (distribution of value sets where the target users wouldn't be expected to be able to perform the expansion): http://build.fhir.org/ig/HL7/cqf-recommendations/branches/master/documentation-terminology.html
Bryn Rhodes (Apr 19 2020 at 21:15):
Specifically the cpg-executablevalueset profile is exactly the case for "implementers can readily incorporate them into their systems, with a note that they are current as of the time of publication".
Bryn Rhodes (Apr 19 2020 at 21:16):
And there is tooling to convert an NLM spreadsheet into a FHIR ValueSet resource.
Grahame Grieve (Apr 19 2020 at 22:23):
NLM publish the value sets through a FHIR end-point. They just prohibit a build process from using that end-point
Grahame Grieve (Apr 19 2020 at 22:24):
I don't see what a profile has to do with that
Grahame Grieve (Apr 24 2020 at 03:28):
@Mark Kramer are you the right person to ask questions about regarding mcovid.org Data Dictionary?
Mark Kramer (Apr 24 2020 at 11:52):
Yes
Grahame Grieve (Apr 24 2020 at 21:28):
ok when I look at the data dictionary, I'm not sure what to make of the Required field - for what purpose are they required?
Also, at the bottom of the page, there's Measure/Measure Result. They have a quite different intent/purpose than the other objects, but they are presented linearly, so there's no context - it feels like some context is missing for each object type
Finally, the page provides no process information - how was this list constructed? what is it trying to achieve? How would you comment on it?
Grahame Grieve (Apr 29 2020 at 19:34):
@Mark Kramer ping on this
Last updated: Apr 12 2022 at 19:14 UTC
