Stream: implementers
Topic: No 'Update' or 'Amend' for immunization.status
Maria Hu (Jul 30 2021 at 20:41):
Resource Immunization.status only has valueset 'complete|entered-in-error|not-done' with binding strength required. How would the updated or amended immz record be mapped to this limited status code? Any suggestions, thanks.
Lloyd McKenzie (Jul 30 2021 at 20:44):
Whether the record had been amended or not is orthogonal to whether it represents a completed, not-done or entered-in-error record.
Lloyd McKenzie (Jul 30 2021 at 20:45):
If you wanted to have a separate flag indicating that the record had been changed (e.g. to capture a lot number not in the original record or record an adverse event), that could be captured as an extension. You could also look at Provenance or history to see if something had been changed.
Lloyd McKenzie (Jul 30 2021 at 20:45):
(You wouldn't want a status of 'updated' that obscured whether the record was of an immunization that had occurred or had been refused.)
Maria Hu (Jul 30 2021 at 20:51):
Lloyd McKenzie said:
(You wouldn't want a status of 'updated' that obscured whether the record was of an immunization that had occurred or had been refused.)
ok thanks Lloyd
Eric Haas (Jul 30 2021 at 23:16):
observation and diagnostic report have amended and corrected, how is that different?
Lloyd McKenzie (Jul 31 2021 at 01:52):
Those are specializations of 'completed'.
Eric Puster (Aug 03 2021 at 14:24):
But isn't entered-in-error a specialization of not-done? What led to giving it a spot on the top line?
Lloyd McKenzie (Aug 03 2021 at 14:55):
Not done is a positive record saying "this was supposed to happen, but didn't for some reason - refused, bad batch, ineligibility, whatever". Entered in error is a flag that is put on a record that was erroneously created. The erroneous record could well have been one where the status had been 'not done'. It's saying "please pretend this record never existed, but by the way, this erroneous record might have been acted on by others".
John Silva (Aug 03 2021 at 15:59):
In MedicationAdminstration the not-done status is VERY important -- clinicians need to know if a Med was given or not. Maybe it's not so strict for Observations I suppose.
https://www.hl7.org/fhir/valueset-medication-admin-status.html
Eric Puster (Aug 03 2021 at 20:28):
Maybe this is only true outside of FHIR, but I have seen many cases where an order was canceled for ineligibility or lack of availability and the user codes "entered-in-error" because they made an error - they didn't realize the winkus wasn't available until after they'd placed the order, or because they misunderstood a verbal order that was being given. The existence of the record itself is a positive phenomenon, where 'not-done' instantly conjures the question "why not?"
Lloyd McKenzie (Aug 03 2021 at 20:35):
Entered-in-error really conveys "no record" - i.e. no record that anything was done, wasn't done, or was even contemplated being done should have existed. As such, it's orthogonal from the other status codes.
John Silva (Aug 03 2021 at 20:56):
I've heard of 'entered-in-error' being used for the use case of a clinician entering data for the wrong patient. In that case something was done for sure, it was just documented for the wrong patient. This is (as I think was pointed out) a "BIG deal" --- because if another clinician made a decision about this data before the error was detected, it could have a drastic effect on the patient. In countries like the US, this could be used for litigation or even in non-litigation situations for process improvement (e.g. lots of incorrect data entry).
Lloyd McKenzie (Aug 03 2021 at 23:00):
Right - which is why we flag the record as having been "in error" rather than just removing it entirely - so that it's clear in the medical record that there was something recorded that may have been acted upon.
Last updated: Apr 12 2022 at 19:14 UTC
