Stream: fmg
Topic: QA - GF#13059 - Approved
Lloyd McKenzie (Mar 15 2017 at 17:29):
GF#13059 - Context type is pointing to OIDs instead of value set URLs
Lloyd McKenzie (Mar 15 2017 at 17:29):
+1
Grahame Grieve (Mar 15 2017 at 18:36):
I don't understand this one. where is the problem?
Lloyd McKenzie (Mar 15 2017 at 18:47):
The context codes are listing what value sets to use, but they're doing it by OID
Grahame Grieve (Mar 15 2017 at 18:50):
that's what it sounded like, but I couldn't see where
David Hay (Mar 15 2017 at 23:58):
+1 (assuming they are oids)
Lloyd McKenzie (Mar 16 2017 at 00:30):
The where is here: http://build.fhir.org/valueset-usage-context-type.html
Grahame Grieve (Mar 16 2017 at 00:33):
so this is just a documentation change then. Those OIDs aren't ever actually referenced
Lloyd McKenzie (Mar 16 2017 at 00:34):
Except that in some cases, value sets may actually need to be defined I think
Grahame Grieve (Mar 16 2017 at 00:35):
not going to define the NUCC ones in the next few hours. There's outstanding issues there. And the others are already defined
John Moehrke (Mar 16 2017 at 00:59):
I would be happy to vote, but I look at the CR and can't determine what the change would be. I see only OIDs in descripiton text, which is odd but not substatitive. Is this CR recommending the creation of multiple valuesets from those OIDs?
Lloyd McKenzie (Mar 16 2017 at 00:59):
Bryn says most have code systems, but only 2 have value sets
Lloyd McKenzie (Mar 16 2017 at 00:59):
CR is recommending the creation of a value set for each row that applies to something coded and replacing the OIDs with value set URLs
Grahame Grieve (Mar 16 2017 at 01:04):
ok +1
Brian Postlethwaite (Mar 16 2017 at 01:06):
NUCC is US Core, don't think is appropriate to be including that.
Lloyd McKenzie (Mar 16 2017 at 01:07):
Agree. Need to remove the reference to NUCC
Lloyd McKenzie (Mar 16 2017 at 01:11):
Bryn just confirmed that these are intended to be example bindings - and that will be made clear as part of the update
Lloyd McKenzie (Mar 16 2017 at 01:11):
So NUCC should be ok
John Moehrke (Mar 16 2017 at 01:12):
+1
Brian Postlethwaite (Mar 16 2017 at 01:13):
is extensible binding ok then?
Grahame Grieve (Mar 16 2017 at 01:17):
this has nothing to do with binding
Lloyd McKenzie (Mar 16 2017 at 01:26):
Well, it's sort of binding. It's a recommended binding documented with the code. To have the actual bindings you'd need to have a profile that sliced by code
Brian Postlethwaite (Mar 16 2017 at 01:52):
The UsageContext.code is extensibly bound to this valueset you're referring to.
Grahame Grieve (Mar 16 2017 at 01:52):
right. but the references we're talking about are in the definitions of the codes, and relate to UsageContext.value
Brian Postlethwaite (Mar 16 2017 at 01:56):
That's why I didn't want NUCC in there, if that' not included, then extensible is fine by me.
If NUCC is in, then that's US codings forced for everyone.
Grahame Grieve (Mar 16 2017 at 01:59):
it would only be example right now at best. but we're not even going to bind it. Lloyd just wants to refer to them by URL
Brian Postlethwaite (Mar 16 2017 at 02:06):
It's already bound to the UsageContext resource as extensible
Lloyd McKenzie (Mar 16 2017 at 02:10):
The UsageContext.code is bound to a value set of 6 codes which are HL7-maintained. That bit is fine. The problem is that those codes talk about what codes should be used for UsageContext.valueCodeableConcept. That discussion isn't currently clear whether it's example or not, plus the recommended value sets are done by code system OID.
Grahame Grieve (Mar 16 2017 at 02:10):
well, are they examples?
Lloyd McKenzie (Mar 16 2017 at 02:10):
So we want the HL7 codes to no longer refer to OIDs, we want them to be clear that the listed codes are examples only, and to use URLs to point to real value sets.
Lloyd McKenzie (Mar 16 2017 at 02:11):
They are examples. There's no intent to bind - at the international level
Brian Postlethwaite (Mar 16 2017 at 02:16):
If they are examples, then why the extensible binding and not an example binding?
Brian Postlethwaite (Mar 16 2017 at 02:16):
(Then I'd have no issue)
Grahame Grieve (Mar 16 2017 at 02:17):
there's 2 different elements at play here.
The value set itself is a normal value set with an extensible binding, all good
Grahame Grieve (Mar 16 2017 at 02:17):
thats on UsageContext.code
Brian Postlethwaite (Mar 16 2017 at 02:18):
ok, thanks
+1
Grahame Grieve (Mar 16 2017 at 02:18):
umm ok\
Last updated: Apr 12 2022 at 19:14 UTC
