Stream: terminology
Topic: Code "Translations"
Gay Dolin (Nov 01 2018 at 17:57):
Is it correct that one can assert required bindings to codeable concepts, and that, so long as one coding is drawn from the required set, it still allows for codes from other codes systems in coding"
Gay Dolin (Nov 01 2018 at 17:59):
As it seems to assert here: https://www.hl7.org/fhir/datatypes.html#CodeableConcept
Lloyd McKenzie (Nov 01 2018 at 18:08):
Yes, that's correct.
Lloyd McKenzie (Nov 01 2018 at 18:09):
Further, you can slice CodeableConcept.coding and have bindings to additional value sets - so if you wanted you could, for example, require a SNOMED, and ICD9 and an ICD10 code all be present for the same CodeableConcept.
Gay Dolin (Nov 01 2018 at 18:10):
Thanks -- so that means the only time you need to slice - is if you want to require additional bindings to other code systems?
Lloyd McKenzie (Nov 01 2018 at 18:12):
More specifically, to other value sets. You could have codes from different code systems in one value set. And in theory, you could also slice and require codes from different value sets that draw from the same code system. E.g. requiring both pre-coordinated and post-coordinated SNOMED codes.
Lloyd McKenzie (Nov 01 2018 at 18:12):
But yes, that's the only time you'd need to slice
Gay Dolin (Nov 01 2018 at 18:22):
Discussion on a CIMI call - they are not understanding how this can be modeled
Gay Dolin (Nov 01 2018 at 18:22):
or represented
Lloyd McKenzie (Nov 01 2018 at 18:22):
I can join if you PM me the call info
Gay Dolin (Nov 01 2018 at 18:24):
https://global.gotomeeting.com/join/116193445
Gay Dolin (Nov 01 2018 at 18:24):
Great
Rob Hausam (Nov 01 2018 at 18:24):
the call is now?
Gay Dolin (Nov 01 2018 at 18:24):
yep
Rob Hausam (Nov 01 2018 at 18:24):
I'll join
Gay Dolin (Nov 01 2018 at 18:43):
So - I just want to confirm of codableConcept is constrained to a particular code, value set or code system - you can still send, for example, an IMO lexical (also can be identified as a "User entered term") in coding
Rob Hausam (Nov 01 2018 at 18:49):
I assume by constrain you mean there is a required binding to a particular code or value set (you don't bind directly to a code system itself). The IMO lexical code can be flagged as user selected and can also be sent as long as it's in some identified code system - i.e. has a value for 'system' (which could be defined by IMO).
Grahame Grieve (Nov 01 2018 at 18:51):
it doesn't have flagged as user selected (but can be)
Gay Dolin (Nov 01 2018 at 18:54):
Right - which we have.
I just want to be sure that the IMO lexical code (or other code/code system) is not precluded from being sent in coding in any codableConcept where there codeableConcept has been specified
Lloyd McKenzie (Nov 01 2018 at 18:55):
Not sure what you mean by "where the codeableConcept has been specified"
Lloyd McKenzie (Nov 01 2018 at 18:56):
It's certainly possible to profile a codeableConcept so it can only send a single code from a single code system - but that's generally a very bad thing for a profile to do
Lloyd McKenzie (Nov 01 2018 at 18:56):
If you use the pattern approach or the slicing approach, you can always set expectations that a particular code (or particular set of codes) is present while allowing room for other stuff to be present too.
Gay Dolin (Nov 01 2018 at 18:57):
just like in CDA - where you can send multiple Translations where say, the observation/code is bound to a particular value set or code or code system
Gay Dolin (Nov 01 2018 at 18:57):
Agree!, "It's certainly possible to profile a codeableConcept so it can only send a single code from a single code system - but that's generally a very bad thing for a profile to do"
Lloyd McKenzie (Nov 01 2018 at 19:16):
Not quite the same. In CDA, there's a root code and a bunch of translations. You then run into the problem of differerent templates saying the "root code must be X", where X is different for each template. With FHIR, all the codings are at the same level and order has no meaning. As a result, there should be no issue simply sending all the codings required by all the profiles and leaving it to the receiver to pay attention to the ones they care about.
Last updated: Apr 12 2022 at 19:14 UTC
