FHIR Chat · DetectedIssue vs OperationOutcome · implementers

Stream: implementers

Topic: DetectedIssue vs OperationOutcome


view this post on Zulip Eric Haas (Dec 02 2017 at 17:54):

If I'm doing some sort of CDS operation and the input code is wrong or out of date -to me that is a technical error but I think would still be reported using DetectedIssue. The DetectedIssues Boundaries suggest is limited to clinical issues and to use OperationOutcome instead. I think that some more clarity is the documentation needed here. I had assumed that OO is constrained to a direct response to an interaction and not some other issue discovered downstream in the workflow. ( for example I successfully post a resource ...yay... and it is discovered later that it is using a deprecated code..drat!... )

view this post on Zulip Lloyd McKenzie (Dec 02 2017 at 18:28):

DetectedIssue is something that's part of the patient's clinical record. I wouldn't expect the fact you'd used an obsolete code to be captured in the patient record.

view this post on Zulip Eric Haas (Dec 03 2017 at 16:06):

My use case is public health case reporting it and the determination of reportability and using trigger codes. So (thinking out loud) I think that in a restful approach it would be OO to the post. Reversing my earlier position. thanks Lloyd

view this post on Zulip Eric Haas (Dec 03 2017 at 16:52):

So if I have a Communication that I am sharing with clinical partners and I want to say "hey... the Sender sent this and it failed and here is why" I would have a status of 'aborted' and a .statusReason extension ( assuming R3) which could reference the OO resource. (There are lots of overlap with AuditEvent here but is being used in a clinical context)


Last updated: Apr 12 2022 at 19:14 UTC