Sex-Gender Stakeholder Consultation Session 2 - Follow-up

I agree completely with your comments. The example from Australia was only to show how data elements are defined using the International Standards Organization standards for metadata (ISO 11179), not to recommend this particular option set. In fact, your comments are exactly why we need a way to describe the rationale, usage, references, limitations etc. for a set of data element options, including how to show when they have been superseded by other definitions.
- Gillian

Hi everyone,

I haven't looked at the documents in detail but i think before we adopt anything, we need to really look at the language and ensure it reflects the language and and perspectives of the gender minority community in Canada. We can't simply take a document from another country and have it mean the same here. For example, looking at the Australian document it defines sex/gender as:

"the identification of a person as either male, female or not exclusively male or female, as represented by a code."

I have significant issues with this definition. Firstly, they make no distinction between sex and gender which are two different terms with different meanings. The are conflating the two in this definition and use. Further, this language identifies Female / Male as the "default" and anyone else neither male or female. This is problematic. It's like saying "white people and everyone else" with a similar impact. The document then goes on to say:

CODE X Other

Other refers to any person who does not exclusively identify as either male or female. Terms such as 'indeterminate', 'intersex', 'non-binary', and 'unspecified' are variously used to describe the 'Other' category of sex.

This literally "others" anyone who is not male or female. We have heard loud and clear the impact of "othering". People who identify as non-binary, or intersex, or gender queer are not "other". We need to begin to desegregate data. We avoid using the term "other" in any data referencing identities of people (race, colour. language, gender). Further, as noted above, these terms relate to gender, not sex.

These are my first few thoughts just looking at the first page. There is a lot of great work already happening in Canada to help set these standards which has been informed by the members of the community. I would strongly recommend we continue down that road.

One comment on FHIR and the use of DataElement resource. That resource is not included in the current FHIR R4 specification. It is replaced by crafting a StructureDefinition as a logical model. I'm not exactly clear on how this works but this sort of thing can not be "exchanged" as a resource. It is more of a pattern, such as FHIR definition pattern . I suspect we could use this approach for defining best practice patterns for sex and gender in FHIR. Worth getting more information on this approach.

FHIR can handle data elements defined according to ISO 11179:
hl7.org/fhir/STU3/dataelement.html#:~:text=Data%20elements%20might%20define%20the,outside%20FHIR%20entirely%20in%20a

"DataElement differs from Observation in that it describes what kind of observations can occur, while Observation focuses on a specific observation of a specific subject at a particular time that has occurred.”

The Australian metadata registry is a great example. Here is one of their data element definitions for gender:
meteor.aihw.gov.au/content/index.phtml/itemId/639275

Note that it gives the value domains and explanations of where and how to use and measure this data element.

Hi Gillian,

Thanks for your suggestion to use the ISO11179 metadata standard when defining sex and gender related concepts.. I think this is an excellent idea as we move toward implementation of the standard and EHR systems. At this time I am only aware of a few Canadian organizations that have adopted ISO11179. I believe CIHI is one such organization that has started to document its data holdings using this standard? Ontario Health Digital Services may be another example? I would appreciate hearing from folks who know more about this topic so we can all learn from this dialogue.

For reference, the Australian Institute of Health and Welfare has adopted ISO11179 for their national metadata online repository on health, housing and community services statistics and information called METeOR, which provides a wealth of information about health data definitions and standards used in that country -see meteor.aihw.gov.au/content/index.phtml/itemId/181162 .

Thanks -Francis

Hi, that was a great session (as usual).
Regarding my suggestion that sex and gender metadata be coded according to ISO 11179 metadata standards, here's a document that covers one component of the standards:
metadata-standards.org/Document-library/Documents-by-number/WG2-N1801-N1850/WG2N1845_WD_11179-6_Edition_3.pdf
Specifically, I suggest that variables be coded according to their registration status:

• incomplete
• Candidate
• Recorded
• Qualified
• Standard (e.g., the good classifications that you don't recommend but that are alternative standards, like CIHI's F,M,Gender Diverse)
• Preferred Standard (e.g., the classifications that you end up recommending at the end of this consultation process)
• Retired
• Superseded
• Application
• Historical

This would enable metadata registrars and developers to manage terminology as it emerges and becomes out of date.
- Gillian