2024-08-20 SHIELD Call Meeting Notes

Owned by riki.merrick
Last updated: Aug 20, 2024
6 min read

Attendees:

Topic:

Notes:

Reviewed the Standards WG report:

Race/ethnicity - issue with March 29, 2024 OMB rule is:

  • takes effect March 28, 2029 - but per this rule the prior approach expires in January 1, 2026, so would that mean this rule requires use of the new OMB rule before it is effective?

  • combines both elements into a single construct (without necessarily a new name / definition)

    • HL7 has not modeled this new approach (it is on the to do list)

    • other data models in research do not yet suport this

    • research data models do not support repeating race concepts, so loss of meaning will occur, unless they get updated (but that is the case now, too)

    • combining legacy data with new data will be more problematic

  • changes the level of some codes in the code system, which changes the meaning for some of them (example “White as higher level code now has different meaning, when “MENA” was elevated to the same level)

    • CDC has not yet provided updated vocabulary - 1997 version expires January 1, 2026, 2021 version still listed, but that would not cover the new hierarchy

    • need to check which versions LOI and LRI are calling out (if they are)

interpretation codes (no explicitly called out in HTI-2)

  • lots of discussion what is best to use for qualitative results - leave OBX-5 blank and always send in OBX-8, or only send in OBX-8, whe numeric result in OBX-5, like for AST (vs sending 2 OBX segments with different LOINCs)

  • many LIS don't currently have SNOMED CT support for OBX-8 - using HL70078 instead

  • recommend sending both HL70078 AND SCT codes for now

    • LRI base only supports IS datatype, so no alternate code systems can be sent - in LRI_PH_Component it is a CWE with support for 3 code systems => make comment on LRI to change to use CWE like LRI_PH_Component

    • means EHR-s and LIS need to allow for suport of more than 1 code system here

Relevant CLinical Information (no explicitly called out in HTI-2)

  • eDOS AOE document has list of most common AOEs http://www.hl7.org/login/index.cfm?next=/implement/standards/product_brief.cfm?product_id=151

  • AOE is not the same as precondition (those are more assertions of the condition, so represented as clinical finding in SNOMED CT rather than observables) - depends on where the precondition is recorded - in the question or the answer

  • traditionaly this was a string field, but now it is coded; but the label of the field does not support better categorization, hence the need for clinical findings value set

  • this needs a lot more work on how to best represent that!

Device Information:

  • LRI suports device identificaiton

    • profile component describes use of PRT - one PRT per device

    • alternatively can use

      • OBX-17, which repeats, but that is a coded element, but UDI device Identifiers are not really codes, so some folks in OO didn’t thikn they should be sent there - created a pseudo-code ystem for COVID reporting, but need a long term approach

      • OBX-18 for instance IDs

        • IHE LAW uses this field for device identifier of the UDI

      • This needs more discussion in OO, unless we can push all to use PRT

  • results with more than one device, specifically some that include manual steps, like cultures

    • would you need culture medium, plus and biochem reagents used for assessment and ognaism id?

  • No model in LIS and EHR-s to capture this currently

  • also need to look at the workflow to get this information into the LIS for each test

    • at the device ID level maybe during set up based on LIVD content

    • at the instance level would have to have an easy way to integrate into the workflow

    • get from instrument (for those that support IHE LAW)

  • Model for data exchange is UDI Pattern Document, as cross-paradigm IG it provides rules for ALL HL7 product families http://www.hl7.org/implement/standards/product_brief.cfm?product_id=487

Code system versions

  • Would be good to ask for clarificaiton which version would overwrite the minimum version for certification - the one called out in the vocabulary standards section (710.270) or the version called out in USCDI V4, which is also called out in the rule, but specifies a different version (170.213)

Why is LRI listed in CPOE

  • Would be good to get clarification around the meaning of listing LRI as an expected standard here

    • does it mean any CERHIT that sends CPOE using LOI must be capable of receiving LRI results back?

    • since this is bidirectional, should make sure each actor fucntionality and the corresponding standard are explicitly called out for each use case

    • does this cover LIS ordering from reference labs / public health labs (that would be send orders rather than receive orders for the Hospital LIS) and it would mean receiving results rather than sending them

    • Certification for PH systems is called out, but not specifically for LIS, though these apply for any HIT that wants to be certified

    • the implementation incentive comes from elsewhere - in case of Promoting Interoperability from CMS, which covers providers, but not other types of labs currently

    • for public health these incenstives could come from CDC

    • modular certificatin for specific functionality is what we might need here, as not all criteria apply to the PHS or LIS

 

Next call will be ALL SHIELD update, but we will use the remaining time for HTI-2 feedback

Please email specific topics for discussion to Riki and Andrea, so we can prep.

 

From Chat:

Serge Jonnaert 12:19 PM
All the race/ethnicity value sets alre listed on the ONC ISA
Messages addressed to "Meeting Group Chat" will also appear in the meeting group chat in Team Chat

Serge Jonnaert 12:20 PM
OK. Sorry, on two calls at the same time

Andrea Pitkus 12:23 PM
Back to "interpretations", shall SHIELD make recommendation that SCT codes be sent with the HL7 078 table terms so SCT is available for both places labs may message (OBX 5, OBX 8)?

Xavier Gansel 12:26 PM
I think so, yes
Ray Zhang joined as a guest
Ray Zhang left

Serge Jonnaert 12:30 PM
Signing off. Have a great rest of week!
Serge Jonnaert left
SZARFMAN joined as a guest

John Snyder (NLM) 12:36 PM
Is represent agar as a device correct? Current all of the agar content is in the substance hierarchy as subtypes of << 421955000 |Culture medium (substance)|. Would you require concepts in the physical object hierarchy to represent the culture plate plus the culture medium as a precoordinated term?

Andrea Pitkus 12:37 PM
Good question. Many labs will purchase different agar plates (prefilled)
https://www.fishersci.com/shop/products/bd-bbl-prepared-plated-media-macconkey-ii-agar-4/B21270X?gclid=CjwKCAjw_ZC2BhAQEiwAXSgClj1D4eetQ7JIf5lHiOQOF02m1dz1Xtkdyqfk5xbhURSHVbflR6zT2hoCvHUQAvD_BwE&ef_id=CjwKCAjw_ZC2BhAQEiwAXSgClj1D4eetQ7JIf5lHiOQOF02m1dz1Xtkdyqfk5xbhURSHVbflR6zT2hoCvHUQAvD_BwE:G:s&ppc_id=PLA_goog_2086145683_81843405314_B21270X__386247001357_3125162398882597393&ev_chn=shop&s_kwcid=AL!4428!3!386247001357!!!g!825073101821!&gad_source=1
A link for one vendor's MacConkey plate. They list their formulation (ingredients as ): Pancreatic digest of Gelatin 17g/L, Pancreatic digest of Casein 1.5g/L, Peptic digest of Animal Tissue 1.5g/L, Lactose 10g/L, Bile Salts 1.5g/L, Sodium Chloride 5g/L, Neutral red 0.03g/L, Crystal Violet 0.001g/L, Agar 13.5g/L

John Snyder (NLM) 12:41 PM
Correct, but does it matter since this is US based, wouldn't we use GUDID for the device instead of SNOMED?

SZARFMAN 12:41 PM
Which incentives can be provided to identify which are best agar plates for identification of specific micro-organisms?

Andrea Pitkus 12:43 PM
@SZARFMAN This is determined by each laboratory in accord with CLIA per the test system they validate

SZARFMAN 12:44 PM
How can we use this information for harmonization of the results?

Andrea Pitkus 12:45 PM
@John Snyder (NLM) the agar plate example/question was does the UDI field support multiple UDIs in cases where they are used to produce a single result value. I wouldn't expect SCT to be used for this field.

Andrea Pitkus 12:43 PM
@SZARFMAN This is determined by each laboratory in accord with CLIA per the test system they validate

SZARFMAN 12:44 PM
How can we use this information for harmonization of the results?

Andrea Pitkus 12:47 PM
So the request is to use all the UDIs that would represent the "device" of the agar plate. Similar to how different additives (kind, amount, %, etc) are in containers, then I'd expect a similar approach taken for UDIs (listing all reagents for a test result). I'm assuming the UDI differs for those with different % of ingredients, etc.
So FDA or other receivers would need to look up all the UDIs to understand that Lab A uses plate X and Lab B uses plate Y and use it in their studies/assessments accordingly

Andrea Pitkus 12:50 PM
@John Snyder (NLM) per the other part of your question as to a precoordinated term, I don't think that is needed currently. The current proposal is to use UDIs.

Recording:

https://aphl.zoom.us/rec/share/xOqi78jVduvHfJlU9cUawAqG6n8OZufhN2hk1-iecG07MhaenGPNgxUQ-3yjC1hW.eUCISjYY4qnNpK3G
Passcode: h8B@9!Dj