Stream: implementers
Topic: AllergyIntolerance.reaction.certainty feedback
Michelle (Moseman) Miller (Oct 20 2016 at 23:17):
The Patient Care work group would like implementer feedback whether AllergyIntolerance.reaction.certainty is supported in your system. Patient Care is trying to assess whether 80% of systems capture a “certainty” at both the AllergyIntolerance level as well as the reaction level.
For example, in DSTU2:
- AllergyIntolerance.status value set [1] contains both unconfirmed and confirmed.
- AllergyIntolerance.reaction.certainty value set [2] contains both likely, unlikely, and confirmed.
We understand the functional difference between these and the desire to know both, but the question is really about whether most systems support documentation of a certainty/verification at both levels?
[1] http://hl7.org/fhir/DSTU2/valueset-allergy-intolerance-status.html
[2] http://hl7.org/fhir/DSTU2/valueset-reaction-event-certainty.html
Grahame Grieve (Oct 20 2016 at 23:21):
none of the systems here: http://www.healthintersections.com.au/?p=2267 nor any of the 8 others I was able to get data for but not publish have anything like AllergyIntolerance.reaction.certainty
Stephen Royce (Oct 20 2016 at 23:31):
The public hospital systems in both South Australia and Western Australian have started sending reaction certainty in their documents to the centralised My Health Record system in Australia.
Grahame Grieve (Oct 21 2016 at 00:02):
are they being populated?
Stephen Royce (Oct 21 2016 at 00:11):
Yep. They are currently fairly basic: likely, possible, unknown and are free text, but we're in the process of releasing an addendum to the Discharge Summary CDA IG along with a SNOMED CT reference set of 6 codes: certain, probable, possible, unlikely, conditional and unclassifiable. The terms and their meanings are taken from the WHO recommendations on certainty of drug-related adverse events and re-worded slightly to suit adverse reactions.
Peter Jordan (Oct 21 2016 at 00:23):
The clinicians that profiled the AllergyIntolerance resource at last year's HL7NZ Workshop wanted to retain reaction.certainty, but with a different valueset - possible | probable | unknown.
Stephen Royce (Oct 21 2016 at 00:50):
So, basically "conditional" = "unknown" and we're still in discussions with our terminology folks about the utility of "unclassifiable". If we drop that, then we'll have 5 codes: certain, probable, possible, unlikely & unknown. The additional ones are useful in the context of challenges, de-challenges, re-challenges &c.
Marten Smits (Oct 21 2016 at 12:19):
In the Netherlands we populate it as well using the codes: Certain, Probable, Possible, Unlikely, Ruled-Out, Unknown.
Last updated: Apr 12 2022 at 19:14 UTC
