Stream: smart/health-cards
Topic: Lab Results Value Set
Pascal Pfiffner (Aug 05 2021 at 23:08):
(I thought this was raised already but can't seem to find it)
Right now the value set for quantitative lab results contains overlapping codes from both SNOMED and LOINC. IIRC the LOINC codes were added after the SNOMED codes, but given the difficult SNOMED licensing agreements is there a reason we're not just cutting them and using LOINC only? This would simplify the value set, obviously, and make international use easier because of the much more permissive LOINC licensing.
Pascal Pfiffner (Aug 05 2021 at 23:10):
Even just clicking the SNOMED codes, going to ihtsdotools.org, makes you accept an agreement first.
Josh Mandel (Aug 05 2021 at 23:31):
I agree that listing multiple codes in our value set for the very same concept is going to leave implementers confused about which ones to pick
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Aug 06 2021 at 03:15):
So LOINC is required for the results and SCT for the results values at the point of origination for lab results for ELR, and to meet HHS requirements in the US. LOINC Answers (LA codes) are not implemented/nor functionality supported in any LIS/LIMS given SCT is required for result values. Highly recommend LOINC and SCT consistent with other requirements and not requiring different coding for the same item. SCT is being adopted globally too. It's true licensing can be an issue for non member countries. However, at last in person SHIELD meeting several years ago, IHTSDO indicated they can create a value set of result value SCT codes that could be used globally for free. I don't think work has been done on this yet, though.
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Aug 06 2021 at 03:20):
Here is the HHS COVID codes guidance: https://www.hhs.gov/sites/default/files/hhs-guidance-implementation.pdf
You'll note the pooled result values SCT codes I mentioned above that don't appear to be in your value set.
Pascal Pfiffner (Aug 06 2021 at 04:52):
I don't think the argument of "LIMS already send SCT codes in the US" is too convincing as to keep duplicative codes in the value set. As a counter-point, supporting SCT codes is a huge hurdle for us because of the licensing issue. We support SHC worldwide but will/would have to country-gate resolution of SCT codes, which comes with considerable engineering effort and prevents roll out for lab result SHCs for us (and I don't see a workable way forward). OTOH, mapping that small set of SCT result codes to LOINC codes seems low effort when an SHC implementation has to convert lab data anyway. I would really prefer only having one set of codes so it's clear what implementers should use, and that set should easily be usable internationally which makes me think LOINC is preferable. Happy to hear other arguments, though.
Pascal Pfiffner (Sep 14 2021 at 06:37):
https://jira.hl7.org/browse/FHIR-33652
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Feb 24 2022 at 14:34):
One point for lab results. CLIA lab regulations require the manufacturer/test kit package insert to be followed (deviations result in additional validation/regulatory requirements as it becomes a lab developed test). This also means manufacturer result terms must be used (i.e. detected, not detected versus pos/neg/indeterminate) with the specific assay "test".
As a result, multiple result codes may be in a value set for all the types of possible responses for a COVID result, but only the responses specified by a package insert can be used for reporting a particular manufacturer's assay by each user. I'll mention there are known gaps in both LOINC and SNOMED CT so if a code is unavailable (i.e. gray zone for some hepatitis result values), text may be the response. I'd expect COVID test kit apps or websites to provide their choices in a drop down or other choice to make it easier for consumers to report correctly.
Last updated: Apr 12 2022 at 19:14 UTC
