Stream: Orders and Observation WG
Topic: Specimen Type
Grahame Grieve (Nov 23 2021 at 22:12):
I'm looking at putting lab test results for Covid in smart health cards. a typical test result is LOINC 94309-2, value = SCT 260373001 | 260415000, and sample type = Nasopharyngeal swab, or sputum, or occasionally something else.
Grahame Grieve (Nov 23 2021 at 22:12):
there's two ways to show this :
Grahame Grieve (Nov 23 2021 at 22:13):
{
"contained" : [{
"id " : "sp1",
"resourceType" : "Specimen",
"type" : { "coding" : [{
"system" : "http://snomed.info/sct",
"code" : "2175005",
"display" : "Pharyngeal cavity
}] }
}],
"specimen" : { "reference" : "#sp1" }
}
{
"bodySite" : {
"coding" : [{
"system" : "http://snomed.info/sct",
"code" : "2175005",
"display" : "Pharyngeal cavity
}]
}
}
Grahame Grieve (Nov 23 2021 at 22:14):
the SHC implementers prefer the second, since it's more concise (really matters in SHC context), but a sample type isn't quite the same as a bodysite. But it's not quite a different thing either - the point of the sample type in a covid lab test is to get at a bodysite. So how does OO think about these two options?
Grahame Grieve (Nov 23 2021 at 22:14):
@Max Masnick
Max Masnick (Nov 29 2021 at 18:43):
Hi OO -- we are getting ready to implement this in the SMART Health Card IG, so any guidance would be much appreciated!
Rob Hausam (Nov 30 2021 at 05:01):
Maybe this is something that we should discuss in OO on FHIR tomorrow? @John David Larkin Nolen @Max Masnick
Max Masnick (Nov 30 2021 at 11:20):
That would be great if there's room on the agenda.
To put all the context in one place:
We need to represent the sample type for COVID tests for SMART Health Cards.
Our profile for COVID lab tests: http://hl7.org/fhir/uv/shc-vaccination/2021Sep/StructureDefinition-shc-covid19-laboratory-result-observation-dm.html. This is where the sample type would go.
Our preferred option (1) is:
{
"bodySite" : {
"coding" : [{
"system" : "http://snomed.info/sct",
"code" : "2175005",
// "display" : "Pharyngeal cavity" // SHOULD be omitted for data minimization reasons
}]
}
}
We could also do this (2):
{
"contained" : [{
"id" : "s1",
"resourceType" : "Specimen",
"type" : {
"coding" : [{
"system" : "http://snomed.info/sct",
"code" : "258500001",
// "display" : "Nasopharyngeal swab" // SHOULD be omitted for data minimization reasons
}]
}
}],
"specimen" : { "reference" : "#s1" }
}
(1) is less verbose, which is really important for our use case -- as this is for a SMART Health Card, we really need to be able to fit into a single QR code, so the number of bytes of FHIR that can be included is limited.
(2) is larger and more confusing for implementers (who don't necessarily have a lot of FHIR experience).
If there is a 3rd option you would like us to consider, in addition to the payload size constraint, we need to use a CodeableConcept (rather than free text for example) to ensure interoperability.
Our question: Does (1) work from your perspective?
While we're here, we also need to represent how the sample for the lab test was collected.
Our plan for this is to define a code system consisting of the following codes:
lab-collectindividual-collect-unobservedindividual-collect-virtual-observedindividual-collect-in-person-observed
And then bind a value set of these codes to Observation.method in our COVID lab test profile. (There are not currently SNOMED or LOINC codes that represent these concepts.)
Our question: is this approach appropriate?
An alternative is to use an extension, which we want to avoid because it adds to the payload size and adds complexity for implementers.
Max Masnick (Nov 30 2021 at 19:52):
FYI discussion of our 2nd question is here: https://chat.fhir.org/#narrow/stream/284830-smart.2Fhealth-cards/topic/Verification.20of.20specimen.20source
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Dec 23 2021 at 04:26):
In V2 SPM 4 is specimen type (i.e. swab, mapped to SCT Specimen hierarchy code indicating what was collected), SPM 5 specimen type modifier (if needed), SPM 8 is specimen source (mapped to SCT anatomic body site indicating WHERE on the body the specimen was collected (i.e. nasopharynx) , SPM 9 is the source qualifier (I.e. left). Would model in FHIR with Specimen Resource fields linked to diagnostic report with COVID observation mapped to LOINC and SCT code for qualitative value to SCT qualifier as indicated.
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Dec 23 2021 at 04:30):
Both the specimen/type of swab and the source of collection impact the sensitivity and specificity of results/values and typically want to be known for decision making by health professionals. Will health professionals be viewing these results (i.e. for access)?
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Dec 23 2021 at 04:32):
How would you represent a parent/caregiver, etc collecting a specimen and/or performing the test on a child/person they are caring for/elderly parent, etc.?
Last updated: Apr 12 2022 at 19:14 UTC
