Stream: australia
Topic: Medications Type
Brett Esler (Jun 07 2018 at 06:38):
Hi folks - seeking some comment on whether http://hl7.org/fhir/extension-valueset-reference.html is a better mechanism and should replace the use of http://build.fhir.org/ig/hl7au/au-fhir-base/StructureDefinition-medication-type.html or be use alongside to indicate the AMT refset applicable to an AMT code for Medication.code.coding - we could refer to the TPP,MPP,MPUU,TPUU, etc refsets directly to indicate to naive clients which level an AMT code is from rather than the explicit coding defined in http://build.fhir.org/ig/hl7au/au-fhir-base/ValueSet-medication-type.html @Dion McMurtrie @Liam Barnes ? This was the issue I was trying to solve for AMT but perhaps the explicit coding of type/category is useful for other medication terminologies also so maybe keep?
Brett Esler (Jun 07 2018 at 06:38):
@Richard Townley-O'Neill also..
Dion McMurtrie (Jun 07 2018 at 06:55):
Yes so I had envisaged that I could "tag" a code as being a particular level which may correspond to what we call MPP in AMT that is called something else in RxNorm and something else in dm+d. My intention was to tag it with a generic code system that has one code for each possible "level" in the union of all "levels" across all the drug dictionaries/terminologies.
I've not updated Medserve to do this yet...but I intended to tag an AMT MPP code with both the generic code I invented and the actual MPP code 30513011000036104 - that way a client can consume it AMT specific or generically. I'd intended to do the same for RxNorm and dm+d when I get around to including them. So it would be possible to write code that was agnostic to the actual underlying code system, or specific to it...whichever suited.
If I just used the AMT codes for these levels there'd be no chance to write the code generically...of course this is all experimental and hypothetical too.
@John Grimes may have an opinion too - he's the one that wrote the browser at https://browse.medserve.online and therefore has much more experience being a client to this than I do.
Grahame Grieve (Jun 07 2018 at 06:57):
I don't think that value set reference is a good way to indicate semantics. Which is what I think you're trying to do here
Brett Esler (Jun 07 2018 at 10:50):
Thanks for comments - this has raised my confidence in the coded approach; certainly practically usable that way.
Brett Esler (Jun 22 2018 at 23:54):
with the CodeSystem used for this extension http://build.fhir.org/ig/hl7au/au-fhir-base/CodeSystem-medication-type.html I would like to relate the codes to AMT MP, TPP, MPP, TPUU etc. any suggestions? ConceptMap or other? @Dion McMurtrie @Michael Lawley @John Grimes @Liam Barnes @Jim Steel
Michael Lawley (Jun 23 2018 at 00:39):
ConceptMap seems most appropriate
Dion McMurtrie (Jun 25 2018 at 01:34):
Yes concept map would be appropriate. I'd like to make one which maps to at least AMT, dm+d and RxNorm...but you could use that to translate directly to/from AMT.
Dion McMurtrie (Jun 25 2018 at 01:34):
would it help if I made that and gave it to you?
Dion McMurtrie (Jun 25 2018 at 11:28):
how about this medication-type-amt.json
Last updated: Apr 12 2022 at 19:14 UTC
