Stream: argonaut
Topic: Must support + an invariant?
Cooper Thompson (Dec 02 2021 at 19:49):
What does it mean when there is a property tagged as must-support, but that same property has an invariant that requires the property only in certain conditions? Does the invariant overrule must-support?
Example: Procedure.performed.
Related tracker: FHIR#33243. That says applied, but it isn't clear whether the MS flag should have been removed when the invariant was added.
Cooper Thompson (Dec 02 2021 at 19:52):
And in general, should all the performed-style timestamp properties on all profiles be invariants instead of MS? We're running into issues where anything that is canceled / not-done isn't passing, since we don't have performed dates for stuff that was never performed...
Brett Marquard (Dec 02 2021 at 19:52):
'Performed SHALL be present if the status is 'completed' or 'in-progress''
Cooper Thompson (Dec 02 2021 at 19:53):
Ah - sorry. I see. Since it is now 0..1 in 4.0.1+, then invariant+MS works. The issue is in 3.1.1 where it is 1..1.
Brett Marquard (Dec 02 2021 at 19:53):
you got it!
Cooper Thompson (Dec 02 2021 at 19:55):
@Johnny Bender , FYI, ONC clarification request coming your way :grinning:
Cooper Thompson (Dec 02 2021 at 20:07):
Although... what about stuff like patient-reported surgical history, where the patient reports it and you don't have performed dates?
Brett Marquard (Dec 02 2021 at 20:13):
performedAge
Brett Marquard (Dec 02 2021 at 20:13):
or string -- it's not a mandatory dateTime
Cooper Thompson (Dec 02 2021 at 20:15):
I see. Thanks! It took me a minute to untangle the two different must-support flags + invariant + cardinality. Whew.
Cooper Thompson (Dec 02 2021 at 20:30):
CFS-3549 submitted for ONC clarification.
Cooper Thompson (Dec 02 2021 at 20:38):
Ugh. Actually for patient reported surgical history, we don't require any time info. It can be provided, but we don't require it. So we have a bunch of historical data that has no "performed" timing info at all.
Cooper Thompson (Dec 02 2021 at 20:39):
I suppose at this point I should just write a ballot comment?
Cooper Thompson (Dec 02 2021 at 20:50):
Eric Haas (Dec 03 2021 at 03:42):
DAR that
Cooper Thompson (Dec 03 2021 at 14:28):
I swear I know all of this stuff in my head. I just get it all mixed up when juggling DAR rules, MS, invariants, cardinality, and US Core 3.1.1 vs. newer versions. Thanks for bearing with me.
Last updated: Apr 12 2022 at 19:14 UTC
