Use of 370159000 |Country of birth (finding)|and its subtypes

SNOMED Groups Content Managers AG
,
1

Dear CMAG members,

SNOMED International has received requests via CRS to update several child concepts of 370159000 |Country of birth (finding)|, as some concepts in this subhierarchy reflect outdated country names and appear to be duplicates. For example, concept 315563003 |Born in Zaire (finding)| is currently active; however, Zaire ceased to be an official country name as early as 1997, when the territory was renamed to the Democratic Republic of Congo (DRC), the focus of concept 705023006 |Born in Democratic Republic of Congo (finding)|.

Following internal discussions, inactivating duplicates has been considered, but there are also arguments against inactivating obsolete country names for historical accuracy and longitudinal analysis. For example, 315562008 |Born in Yugoslavia (finding)| is currently active, even though Yugoslavia ceased to exist in 1992 and was dissolved into several countries; therefore, these are not always straightforward terming changes, but rather the redrawing of geographic boundaries. For individuals born in those countries at that time, these terms still have significance and may be required for administrative purposes, further complicating inactivation.

There has been prior discussion about the potential inactivation of country concepts in SNOMED CT (e.g., < 223369002 |Country (geographic location)|) due to their lack of maintenance over time. Many countries instead use the country codes defined in ISO 3166.

SI are aware that the UK currently uses these concepts for GP registration, and would like to request input from all members: If you are currently using, or planning to use, 370159000 |Country of birth (finding)| and its subtypes in your implementations, please share your use cases. Specifically, we would appreciate details on how you use these concepts clinically or administratively, any reliance on historical country names, and whether you use alternative approaches such as ISO 3166 country codes. Also include any challenges you foresee if the content were to be deactivated, such as legacy/longitudinal analyses or administrative use cases.

Understanding international use, particularly regarding historical names, will help us manage these country-of-birth concepts within SNOMED CT. Your feedback on current or planned usage will be invaluable to our review.

1 Like
Tag for concepts considered for inactivation?
2

The short answer from Australia is leave it alone - There’s more important content problems that SI should be focusing on. Retiring the content risks impacting existing implementations, with negligible - if any - benefit. For anybody not using it, there’s no impact in the content existing (Noting translations might be an exception).

Longer response.

  • We do not support wholesale inactivation of these concepts. The content is not actively causing problems elsewhere in the terminology, and retiring it introduces implementation risk for limited practical benefit. We got burned with the ethnicity retirement and would like to avoid a repeat.
  • Where a country has simply been renamed with no change in territory, updating the PT/FSN to reflect the current name is reasonable, provided the former name is retained as a synonym (e.g., Zaire to Democratic Republic of Congo).
  • Concepts such as “Born in Yugoslavia” represent historically valid entities and should not be inactivated. These cannot be straightforwardly mapped to a single successor country, as the territory was divided across multiple modern states.
  • The countries did exist and people were born in them. This is an immutable fact.
  • For individuals born during the existence of these countries, the historical name may still carry personal or administrative significance.
  • If the content is not causing terminological problems and inactivation carries implementation risk with no clear benefit, the pragmatic course is to leave it in place. There are higher-priority areas of the terminology that would benefit from editorial attention.

Most NRCs don’t have full visibility of how these concepts are used in practice. Use case knowledge often sits with national information model custodians, jurisdictional health authorities, or individual implementers, not with the NRC itself. Asking NRCs might be all you can do, but it shouldn’t be considered a comprehensive insight to what is actually in use.

2 Likes
3

Hi Matt,

These are helpful comments. At the moment, we are just gathering requirements (if any) to inform how we will need to manage this content, along with the CRS requests we have received in a consistent manner.

Thanks for taking the time to respond.

4

Hi Michael,

In general, I agree with @mcordell.

That being said, I have had at least one discussion with Nathan McEachen from TerraFrame, Inc. regarding the potential to represent geolocation information in a SNOMED-like framework. Nathan has attended SNOMED Conferences in the past, but if this issue is of sufficient importance perhaps there is a partnership opportunity between SNOMED International and TerraFrame, Inc. to investigate the potential for appropriate representation of this information using a logical model that SNOMED could adopt or that could be spun off into the community space?

Just a thought.
Thanks
John

5

John,

While I appreciate the value in capturing geolocation associated with events (I used to teach a class on spatial epidemiology) I am not sure that is in scope for what we are talking about here. The question relates to the value of maintaining clinical findings of both current and past birth places, given that most countries now use ISO 3166 to capture country information (ISO 3166-3 includes obsolete country names). As this hierarchy has not been maintained for many years, what is the value of outdated country names. I think Matt outlines some good reasons and we have also received similar justification from the UK.

1 Like