Stream: argonaut
Topic: DAR for Religion
Jay Lyle (Aug 08 2019 at 15:58):
We have (religion) data using null values. R4 religion extension is extensible, but US Core DSTU2 extension has a required binding.
1. We can't DAR values on a required binding, but I think we can DAR the whole extension, no? If not, we're looking at creating another extension.
2. DAR values are not lining up as well as we'd like. Do these look reasonable?
ASKED BUT DECLINED TO ANSWER:
DSTU2: asked
The definition asserts that this is "unknown" to the respondent; this is incorrect. Does that matter?
Alternatives include 'masked' (by the patient) and 'unknown' (to the system).
V4 has 'asked-declined,' when we get there.
OTHER: unsupported.
OTH would be better; this is lossy but seems to fit as well.
UNKNOWN/NO PREFERENCE: unknown.
A little incorrect but seems ok. 'unsupported' might be closer for 'no pref' ('n/a' isn't in DSTU2) but even that's lossy, and we don't know which it is.
Brett Marquard (Aug 08 2019 at 18:05):
What null values do you currently have in your system? Religion isn't mandatory so I would leave out unless clear value in exchanging the null value
Lloyd McKenzie (Aug 08 2019 at 18:46):
You can't DAR the extension if the extension is properly defined. Extensions are defined to either allow a value or child extensions. So if the extension is present, the value element must be present.
Grahame Grieve (Aug 08 2019 at 18:47):
the DAR goes on the extension value, since that's what's unknown
Lloyd McKenzie (Aug 08 2019 at 18:47):
And you can't have an extension in place of a value attribute if the value has a required binding
Jay Lyle (Aug 08 2019 at 18:52):
OK. So we could just not include the extension in the null cases (values in CAPS above). But if the client needs them, we have to create another extension, unless we can find a way to use the extension that asserts it's R4 (and has extensible binding).
Last updated: Apr 12 2022 at 19:14 UTC
