Stream: implementers
Topic: conflation of "state of X" vs "state of statement of X"
Erich Schulz (May 26 2016 at 02:19):
from allergy negation discussion...
Erich Schulz (May 26 2016 at 02:20):
it seems that in several places "written in error" appears as a value along side more specific values that are qualifiers on the record
Erich Schulz (May 26 2016 at 02:22):
I guess the question is: is this a good thing?
Erich Schulz (May 26 2016 at 02:27):
Observation seems very clean
Erich Schulz (May 26 2016 at 02:27):
but Condition and AllergyIntolerance not so much
Lloyd McKenzie (May 26 2016 at 22:58):
We haven't come up with a use-case where the "state" of a resource was relevant once it was determined to be "entered in error", so the complexity of having a separate flag didn't seem worth it.
Erich Schulz (May 26 2016 at 23:03):
it means you are obliterating a field by marking it as in-error...
Erich Schulz (May 26 2016 at 23:04):
isnt a primary directive to not do that...
Erich Schulz (May 26 2016 at 23:04):
and with some of the values there are opposing meanings
Lloyd McKenzie (May 26 2016 at 23:05):
Obliterating a field that's not relevant isn't necessarily a problem. Whether it was an assertion that the allergy existed or it didn't, the key message now is that this allergy record should never have existed. (You can always look at history to see what it was.)
Erich Schulz (May 26 2016 at 23:13):
i guess this raises the question of why this value even exist then
Erich Schulz (May 26 2016 at 23:14):
if it doesn't allow meeting of basic audit/version requirements and still dependant on logs why not just allow deleting the record? (because you keep an audit trail right?)
Lloyd McKenzie (May 26 2016 at 23:18):
Because there's a clinical need to have something come back in the search response that says "hey, there used to be a record here, but isn't anymore, so if you might have made decisions based on this, you may want to revisit them". If that use-case doesn't exist, deleting is certainly preferable.
Erich Schulz (May 26 2016 at 23:20):
but you're keeping an audit trail right of all the versions?
Erich Schulz (May 26 2016 at 23:20):
now you need to be aware of both an audit trail and this value
Lloyd McKenzie (May 26 2016 at 23:22):
People generally only bother to look at the audit trail of something if they know it exists in the first place.
Erich Schulz (May 26 2016 at 23:31):
still not convinced allowing "written-in-error" to over-write clinical statements is dependable
Last updated: Apr 12 2022 at 19:14 UTC
