FHIR Chat · Verification of specimen source · smart/health-cards

Stream: smart/health-cards

Topic: Verification of specimen source


view this post on Zulip Max Masnick (Nov 30 2021 at 19:46):

I used our audience with OO to ask about how to handle the method of specimen collection (https://github.com/HL7/fhir-shc-vaccination-ig/issues/165), meaning one of lab-collect | individual-collect-unobserved | individual-collect-virtual-observed | individual-collect-in-person-observed

They thought Observation.component was a reasonable place for this. We would need to find a code for Observation.component[].code to indicate "method of collection", and then figure out the value set for the concepts listed above (create our own, get SNOMED codes, etc.)

view this post on Zulip Grahame Grieve (Nov 30 2021 at 19:48):

I really don't think that this should be called 'method of collection' - that will make any body connected to the process think about venepuncture etc

view this post on Zulip Max Masnick (Nov 30 2021 at 19:48):

Yeah I was struggling with how to describe this...suggestions appreciated!

view this post on Zulip Grahame Grieve (Nov 30 2021 at 19:50):

looks like LOINC doesn't have a code for this:

Details about the exact source and collection method (e.g., blood drawn from the right arm and maintained on ice) are not a proper part of the test name

(and that's not even about this method)

view this post on Zulip Grahame Grieve (Nov 30 2021 at 19:50):

'verification of specimen source'

view this post on Zulip Max Masnick (Nov 30 2021 at 19:50):

@Neelima Karipineni did a survey of SNOMED as well and didn't find anything that really captured this

view this post on Zulip Rob Hausam (Nov 30 2021 at 19:52):

I agree this is not the traditional "collection method".

view this post on Zulip Rob Hausam (Nov 30 2021 at 19:55):

I don't think "verification of specimen source" conveys this idea to me very well, either.

view this post on Zulip Neelima Karipineni (Nov 30 2021 at 19:57):

Copy/paste from my comment on the github issue in July -

I did a LOINC and SNOMED search and did not find codes that really convey the options above, either single codes or even combinations that could be used together. There isn't a code that generically represents a patient-performed procedure such as sample collection, and did not find anything close to 'observed', 'supervised', 'monitored', etc.

There are encounter status codes that sort-of represent in-person vs remote encounter, but that was the closest I could find.

view this post on Zulip Rob Hausam (Nov 30 2021 at 20:00):

This seems sort of like "collection process type" (not that I particularly like that term, either).

view this post on Zulip Max Masnick (Nov 30 2021 at 20:00):

Yes, would very much appreciate more suggestions on how to name/describe this

view this post on Zulip Vitor Pamplona (Nov 30 2021 at 20:18):

specimen collection accuracy?

view this post on Zulip Isaac Vetter (Nov 30 2021 at 20:36):

I've been calling it "specimen collection supervision status"

view this post on Zulip Vitor Pamplona (Nov 30 2021 at 20:38):

FYI, for Urine collection, people generally use the term "Monitored Collection" to mean supervised but not directly visual, generally using audio cues, as opposed to "Direct Observation" which means visually seeing it (either by video or in person)

view this post on Zulip Vitor Pamplona (Nov 30 2021 at 20:45):

"collection reliability" could also be an interesting name.

view this post on Zulip Max Masnick (Dec 01 2021 at 02:06):

Isaac Vetter: I've been calling it "specimen collection supervision status"

I'm going to go with this for now unless anyone strongly objects

view this post on Zulip Max Masnick (Dec 01 2021 at 02:35):

Started working on the implementation of this: https://github.com/HL7/fhir-shc-vaccination-ig/pull/191

view this post on Zulip Rob Hausam (Dec 01 2021 at 23:11):

Since the value set really does seem to be pretty much all about supervision (more so than the "who" aspect), I think that "specimen collection supervision status" should work.

view this post on Zulip Max Masnick (Dec 02 2021 at 14:55):

Thanks Rob. It sounds like that’s the winner then.

I believe we are going to attempt to get SNOMED codes rather than coming up with our own bespoke codes.


Last updated: Apr 12 2022 at 19:14 UTC