Page tree

Summary

The CMAG is being asked to contribute to the discussion on additional description types, in particular what additional description types would be required by extensions. Links on the discussions to date and the planned discussion at the next Editorial Advisory Group meeting are provided below. 

Examples of potential extension use cases may include: 

  • Search terms - colloquial terms
  • "Patient-friendly" or consumer terminology
  • Abbreviations/truncation

Current description types:

  • Fully Specified Name
  • Synonym (includes the preferred term) 
  • Definition

 

Would you please add your comments below in the discussion section as well as the particular requirements for your country in the country response section. 

Relevant documents

No files shared here yet.

 

Links:

2019-10-28 Editorial Advisory Group Face-to-face Meeting - Kuala Lumpur -see item 9. Use of hypernyms as descriptions

2020-04-06 Editorial Advisory Group Conference call

Actions: 

Date

Requested action

Requester(s)

Response required by:

Comments

19 March 2020Input on additional description types
Please post your final responses in the Country response table below. Discussion comments can be made as comments below.
  •  

Country response 

CountryDateResponse
 Sweden2020-03-23 

We already have language reference sets, which is a powerful tool to specify description use contexts such as (potentially) patient friendly/common language terms, speciality terms. Linguistic variants (abbreviations, singular/plural forms, some derivations) we would (mostly) consider as acceptable synonyms. We are still in the process of finding out what constitutes over use of language reference sets.

"Search terms" we assume are generally not (true enough) synonyms and thus might require a new description type OR a new acceptability value and a slight(?) re-interpretation of what 900000000000013009 | Synonym (core metadata concept) | means.

Australia2020-03-24

Very similar approach to Sweden. Abbreviations are already supported and editorial guidance exists.

I could see a "patient friendly" synonym type being viable for when the same synonym will apply to a large number of specific concepts. And patient doesn't "need" to know that that much detail. "Patient-friendly" terms are inherently "lossy" - reduced detail. I'm not sure about the value of truncation? We create a lot of synonyms, specifically aimed at a consistent search experience.

As an aside, we've been informed by a clinician that we're in a (quote) "post mellitus" world, and "Diabetes" is an appropriate/preferrable synonym to "Diabetes mellitus". We're in looking into how we'll action this.

Norway2020-03-27
2020-
We have had input that there is a need for Patient-friendly terms. Personally, I am not sure of the consequences. It is important such at term is semantically equivalent to the FSN, so in cases where this cannot be achieved, a patient-friendly term cannot be assigned (at least not within the Scope of SNOMED CT). It is really up to SNOMED to define what a patient-friendly term should be defined as. Search terms / abbriviations are Language spesifici and should be solved outside of SNOMED CT. Search term lists are also shared With other terminologies, e.g. ICD.
WW
Canada 2020-04-07 For the time being we suggest:

•UI terms should be kept in local extensions to meet the local requirements
•Keep it simple and use the existing language refset mechanism
•Use a description refset for searching (same as we did for Apelon additional synonyms)
•Currently when we include terms that may “deviate” from the approved editorial guidelines, the rationale and rules are documented in the national guidelines.

But we do have questions related to the SNOMED Edition(s) Future Strategy:

•How would the proposed changes impact the future of the SNOMED International Community Edition, SNOMED CT International Edition and Clinical Core Edition?
•If more than one country is using the same « patient-friendly » terms will they be included in the International Edition?
•Will these additions be documented in a specific « patient-friendly » guide?
•How important is it to have a descriptionType of Abbreviation when the term is obviously an abbreviation?
•Suggest that the patient-friendly terms be created outside of the core or community files, so organizations can decide if they want to use them or not.

 US 2020-04-08

 Sounds like we are moving away from reference terminology towards interface terminology. This is not necessarily a bad thing, but I think we should clearly acknowledge the move and possibly bound it. Otherwise, there is a significant risk for scope creep.

Regardless of the mechanism, what is important is to be able to capture the type of variant explicitly (e.g., to select/filter out such variants in specific contexts).

 UK08-04-2020 We also occasionally get asked about 'patient-friendly' representations, particularly as more and more have access to their records online.  And certainly for e.g. Diagnostic Imaging, and more recently, pathology content, truncations and abbreviations have been something we have had to contend with, and finding ways of representing these within the current Editorial Guide principles can be problematic.  But as per the US comment, I would second the concerns regarding 'scope creep' and the move towards an interface terminology.    
   
   
   
   






Member countries without a CMAG rep  

 

CMAG response

DateCMAG ResponseNext steps
   
   
   

 

Final outcome: 

Date: 

 

  • No labels

3 Comments

  1. Hello,


    interesting discussion considering we are currently having the same discussion in Belgium regarding search terms.

    What was the final outcome here, if any?


    Kind regards, Katrien

  2. Is this functionality being considered in order to support the LOINC related names (i.e. search names), display names, and consumer friendly names for the LOINC/SNOMED cooperative? The related names would be considered "search terms" and are helpful in identifying content in a search that otherwise may not be returned. The display names would be considered clinician friendly names.

    The US does receive request where the proposed use case is to "increase searchability", however, there is always a question of synonymy or semantic equivalence to the FSN. 

    Also, could some additional clarification be provided about what is meant by "Abbreviations/truncation" as a description type? Would this open the door for having a description that is solely and abbreviation without following current editorial guidance for including the spelled-out abbreviation? (General Naming Conventions)

    1. Hi John,


      thank you for your reply. Exact this we see in Belgium ' the proposed use case is to "increase searchability", which opens the door to abbreviations (and descriptions in general) that are not in line with (inter)national editorial guidelines... 


      You need these terms on a national level, but you don't want them to be published as a 'correct' acceptable synonym.