Question on Ethnicity and Race-based Value Sets

Hi Finnie, thanks for your support to create the race extension, and to point out the Canadian edition SNOMED codes in the CRDM toolkit for the race categories.

Hi Francis,

I support creating an extension in CA Core/Baseline for race/racialized group. With regards to SNOMED CT codes for the answer list/concepts, please see the CIHI Reference Data Model (CRDM) document www.cihi.ca/sites/default/files/document/cihi-reference-data-model-toolkit-en.pdf on page 92. If you are searching in SNOMED CT browser, please be sure to search in the CA extension. To mention, we do not use LOINC codes for the race/racialized group answer list/concepts.

Finnie

Hi Sheridan.

We have not created a value set yet that FHIR can use. So you are welcome to create one using the concepts from the Racialized Group in CRDM.

Thanks

Hi thanks for continuing this discussion. Here are my additional comments about ethnicity, race and CA baseline:

1. CIHI's proposed race-based data standard is consistent with the race question-responses used in Pinto's SPARK study currently underway on the collection and use of SDOH data to address inequities -see upstreamlab.org/spark-study/
2. I believe the reason CIHI's race categories have a mixture of race, ethnicity and geographic values is a pragmatic one to reflect the changing mix of the Canadian populations. Some people have difficulty distinguishing the difference between race and ethnicity, and identifying which categories they belong tom especially if they are of mixed heritage.
3. We did some initial mapping of CIHI's race categories to SNOMED codes and LOINC-answers, and found not all categories are covered. So some new values and/or value set(s) will likely be needed.
4. One added benefit of using SNOMED to encode race categories is the ability to include subtypes for each category if finer distinction is required.
5. Race is just one of the data elements in the social determinants of health (SDOH) standards that are still evolving. As such, I wonder if it is possible to create a race extension in the Canadian FHIR baseline, along with an initial value set drawing on existing and extension SNOMED CT concepts and LOINC-answer lists as needed. This will allow us to pilot the use of the race value set(s) in specific settings.
6. BTW, Dr. Andrew Pinto will be doing a virtual presentation of their SPARK study on May 19 as part of a mini-conference organized by UVic, CIHI and Infoway. Join us if you wish to learn more about SDOH. Here is the link www.uvic.ca/hsd/hinf/home/news/current/hts-virtual-miniconference.php

Thanks -Francis

We're still porting over the issues from our old CA-Core (now CA Baseline) repo into Simplifier, but the issues and discussion around inclusion/exclusion of Race and Ethnicity extensions that Shamil is referring to can be found here. simplifier.net/CanadianFHIRBaselineProfilesCA-Core/ExtensionEthnicity/~issues

Essentially, we did not originally include race or ethnicity extensions in the CA Baseline because we were waiting for confirmation that it was actually be collected and used by Canadian systems (issue #5).

Later on, our members in BC identified that ethnicity is in fact collected and used in Panorama and requested that ethnicity extension(s) be added to the Baseline (Issue #64).

I believe Panorama uses a mixture of CDC and local codes, but as Shamil pointed out, the CDC codes that Panorama starts with (~6 codes) are far from completely representing the span of 900+ ethnicity codes we see in the full value set that the US Core references in its "detailed" ethnicity slice.

Igor suggested we point to an established Canadian value set that draws from SNOMED CT & SNOMED CT CA codes that is more fulsome (216 codes) :https://tgateway.infoway-inforoute.ca/singlesubset.html?id=2.16.840.1.113883.2.20.3.208&versionid=20190930.

3/6 CDC codes map 1:1 to terms in the Canadian ethnicity value set, the other three do not (Aboriginal, Latino, Other Ethnicity). My suggestion is that the keepers of the Ethnicity VS determine whether or not those CDC codes should be extended into the ethnicity VS or whether they are more appropriately aligned to a racialized group value set. Changes should happen at the source, rather than at the Baseline to prevent proliferation of value sets that only further confuse implementers.

Finnie - will there be a source FHIR value set that the Baseline can point to for racialized groups or are you suggesting the Baseline will need to create the value set using the permissible values outlined in the CIHI reference model: www.cihi.ca/sites/default/files/document/cihi-reference-data-model-toolkit-en.pdf ?

Hi Lloyd,

Please see the other thread for further info on what CIHI has been doing with regards to this. infocentral.infoway-inforoute.ca/en/forum/254-health-terminologies/3834-ethnicity-and-race-related-concepts-terms-and-value-sets#6773

Thanks