Welcome and introductions

2

@Peter Williams

@Rob Hausam

Recording, notes & attendance.

Previous Meetings

0

@Peter Williams

@Rob Hausam

• 2022-09-25 - SNOMED on FHIR Meeting (TS & TB - Hybrid Meeting)

• 2022-09-20 - SNOMED on FHIR Meeting (TS & TB)

Other Meetings

5

@Peter Williams

@Rob Hausam

Recent events:

Upcoming events:

FHIR North - October 26 & 27

Australia Connectathon

Other Regular Meetings:

HL7 Group TSMG (meeting Wed PM ET every other week) - Terminology Service Management Group (HTA Thursday AM is now a subgroup of the TSMG).   2022-05-17 RH Joint session with Vocab at last business meeting.  2022-06-14 Group has approved minimum capabilities for terminology servers.  Now looking at bigger/better HL7 Terminology Server 

R5 changes suggested around specifying language using BCP-47 

15

@Rob Hausam 

@Daniel Karlsson (Unlicensed) 

See Tuesday meeting item #6

Changes for R5  and check on use of dialect in x form in ECL Specification Appendix C:  Appendix C - Dialect Aliases

Note that where a language reference set pulls from more than one language, the BCP-47  tag could potentially missing out the <lang> element and start directly with "x-".  Options for referring to multiple languages: could be done either with a * (see https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Accept-Language#syntax ) or with a comma separated list to indicate a priority order (although this isn't supported by the FHIR parameter (unless the server went out of it's way to allow for this), or use the accept-language header).  Use case is in the displayLanguage input parameter to a ValueSet $expand operation

2022-08-31 ML CSIRO attempting to get R5 compliant Ontoserver ready in time for HL7 Connectathon.

2022-10-04 With this page moving to terminology.hl7.org it is no longer subject to the ballot process.  Instead a "TSMG" ticket should be created.  @Daniel Karlsson (Unlicensed) and @Rob Hausam will progress this.

Also tracking R5 changes to concept map - see http://build.fhir.org/valueset-concept-map-relationship.html.   @Peter Williams to check SI intention for R5 compliance.

Immunization resource

@Marie-Alexandra Lambot 

How and where in the immunization resource would you represent the information that the vaccine is given because the subject was exposed to a risk like exposure to monkeypox? Would you use Immunization.reasonCode element? And if so, would you populate it with SNOMED CT event concepts like 1290711000000105 |Exposure to monkeypox virus (event)| or would you use 409516001 |Post-exposure prophylaxis (procedure)| to convey the notion that the vaccine is given because of pre-exposition and use the Immunization.reasonReference element to further specify this? Now the second option brings me to ask if an exposure event concept can be used as a Problem in a problem resource. I'd say no when I look at the resource but an event can be the reason for admission and in the case of suspected exposure to a risk, you won't get any other diagnosis then that at the end of the consultation so how would you record this? How do you record and where a patient coming because he/she was exposed to a risk, that has currently no complaints or signs, and that gets a vaccine as prophylaxis against this risk transforming into a disease?

The interest of 409516001 |Post-exposure prophylaxis (procedure)| is that you also have 1137457009 |Pre-exposure prophylaxis (procedure)| which is applicable to all other vaccines given to prevent a risk one has never been exposed to. The exact nature of the risk prevented is actually known through the name of the vaccine concept itself and it might be redundant to try to precoordinate more the immunization reason, especially is one links a full report in Immunization.reasonReference.

2022-10-04 DK/PGW Four places where this information could be recorded - vaccineCode (mandatory), reasonCode, basedOn (linked to an ImmunizationRecommendation where the target disease is specified).  Also protocol applied specifies the target disease.

Investigating the reason code element, the suggested ValueSet offers 281657000 |Travel vaccinations| which is now inactive with PossEquivTo 14747002 |Elective immunisation for international travel|.  Also suggest that they add 409516001 |Post-exposure prophylaxis, 1137457009 |Pre-exposure prophylaxis| or the parent of both: 710170000 |Administration of prophylactic treatment|

Including Map Advice in $translate

20

@Kai Kewley 

FHIR Terminology Servers which include Map Advice from an extended map eg ICD-10?

The additional advice is essential to choosing the appropriate code to map to.

Example (not always available): https://dev-is-browser.ihtsdotools.org/fhir/ConceptMap/$translate?&url=http://snomed.info/sct?fhir_cm=447562003&code=254153009&system=http://snomed.info/sct&targetsystem=http://hl7.org/fhir/sid/icd-10

ICD-10 CM Example:  https://dev-is-browser.ihtsdotools.org/fhir/ConceptMap/$translate?code=254153009&system=http://snomed.info/sct&targetsystem=http://hl7.org/fhir/sid/icd-10-cm&version=http://snomed.info/sct/731000124108

Output saved to Gist:  https://gist.github.com/pgwilliams/981405102964c3d8ea74a3f3e7b4c206

PWI Noted opportunity for Snowstorm to specify the correlation as well (which we have):  http://build.fhir.org/valueset-concept-map-relationship.html

2022-8-23 KK Not currently a requested feature, archive.

SNOMED URI Specification

@Peter Williams 

Discussion around URIs for module composition.

Also URI for SNOMED languages eg Post Coordinated Expressions.   Discussed and agreed but not yet officially published.

MRCM in FHIR

Machine Readable Concept Model (applies to both pre and post coordinated content)

30

@Kai Kewley 

Going forward:  suggestion that PJ and RH expose both the MRCM and the capabilities of ECL 2.0 with the wider HL7 and FHIR Communities.  See slides on ECL 2.0 here.  ML more generally this is a refinement axis which is represented as two things - the attribute and the range of values that that attribute can take.

Support for other CodeSystems in Snowstorm.

10

@Kai Kewley

Confirmation that SI will be adding some support for other CodeSystems based on leveraging the existing HAPI library support for such things.  Will be limited to the FHIR API (not the "native" Snowstorm API, except perhaps augmenting our return for ICD-10 maps to include the term associated with the code).  This Will allow us to host the UK Extension due to the requirements of their extensive maps.   

Targeting August release of Snowstorm.  Plan is to contribute code back where enhancement is required.

2022-08-23 See https://github.com/IHTSDO/snowstorm/releases/tag/8.0.0-beta

Cardinality on Designation.use

@Peter Jordan (Unlicensed)

@Rob Hausam

https://jira.hl7.org/browse/FHIR-29821 "Updated cardinality of designation.use"

Expected to be part of R5.

Could we offer Use of BCP-47 + designation.use for review by the HL7 Vocab Group @Rob Hausam? Targeting 9 - 13 meetings.

2022-10-04 Designation.Use extension still required for specific use cases eg consumer definitions.

Post-coordination Expressions on FHIR

@Kai Kewley

2022-02-22:  Presentation: https://docs.google.com/presentation/d/1Kz-8N6UQsGZ8C1xsC8TZT6LEbeyr6elrgMt2FI-whjU

• Supplements might not be able to hold new concepts: https://www.hl7.org/fhir/codesystem.html#supplements bullet no 4.

• (Ontoserver seems to do this: https://ontoserver.csiro.au/docs/6/postcoordination.html#Expression_Libraries )

Proposal might include using side-effects which some might be sensitive to.

Runtime maintenance is required. Is it possible to reuse the existing API in order to standardize the approach?

A Code System Supplement may be used to hold a set of postcoordinated expressions as FHIR concepts

• This approach used by Ontoserver so that they are "known" to the Terminology Server so that $expand can be used without considering all possible PCEs.  Not that some clients may not be able to process PCEs so the CS Supplement allows these to be kept separate.  @michael lawley will raise ticket to "Add parameter to ValueSet $expand to specify Code System Supplements to be included" (would need to be compatible with code system that the ValueSet is being expanded against)

• "No requirement for a designation"

CodeSystem/$validate-code operation could be used to validate the expression

• Yes (doesn't need previous knowledge of the PCE, similarly for $lookup

Could calling CodeSystem/$lookup automatically create expression concepts in the supplement, to be returned in ECL queries for analytics? Or use PATCH?

• Probably not, see use of Supplement above.   Yes use of PATCH seems reasonable (PUT a whole new version more common, but PATCH might allow for finer grained updates ie not having to re-submit the entire cohort of expressions).   ML: Custom operation might be useful here to add to the codeSystem and update the ConceptMap in one go.

ConceptMap/$translate could translate an expression to a single identifier to support more limited clinical record systems.  Could this work in real time, automatically creating entries as required?

• ML: Use a local CodeSystem for the expression identifiers and a concept map to link them together.

2022-05-17 KK Were tickets raised around clarifying questions?   2022-05-31 ML Taking an action to raise a ticket for expand so that you can specify one or more CodeSystem Supplements to be used in that expansion. "Expressions already exist as concepts in the given CodeSystem implicitly, Therefore they can be used in a CodeSystem Supplement.   The rules for CodeSystem Supplements state that they can only be used with Concepts that already exist in the CodeSystem"  "An expression is not adding a concept to a CodeSystem, it is just expressing something that is already there"

See https://jira.hl7.org/browse/FHIR-37529

2022-06-28 KK would like an agenda item added to next meeting (1 month time) to discuss a couple of strawmen proposals.

2022-07-26 KK walking through the postcoordination workflow (link above).

CodeSystem supplement for expression library, ConceptMap to map between expressions and fixed-length generated identifiers. ConceptMaps need explicit ValueSets for source (expressions) and target (generated ids).

What should be the system of the generated identifiers, should these identifiers be SCTID(-like)? ML: Would use UUIDs rather than SCTIDs.

A POST on $expand would allow composing a value set with the dependent SNOMED CT edition and the post-coordination supplement and the be filtered with any relevant ECL (for example).

RH: we're not creating new concepts (new meaning) but rather a new identifier from something that is already SNOMED defined meaning.

Having the generated id as a property on the concepts in the supplement would allow:

• the alternate id would be returned for any $expand operation including the supplement, possibly filtered by ECL

• an expand on a ValueSet composed of the supplement and supplemented CodeSystem filtered by the alternate id property would return the expression.

Alternatively, an additional CodeSystem for the generated ids could be created where the expression could be provided as properties.

2022-08-23 We'll pick this up when @Former user (Deleted) is next available for a Wednesday 10:00UTC call.

2022-09-06 KL worked through SNOMED CT Postcoordination Workflow (Discussion)

ML Would a refset of PCEs represent a Code System Supplement?

PGW - Archive Above Discussion.

Upcoming R5 changes

@Rob Hausam

Discussion about proposed federation.  ML Doesn't work so well with SNOMED CT because there's not a single authoritative source due to country extensions (and multiple versions being released).  Use case for this is mainly in creation of IGs.  Agreement that what a conformant server should be able to state in Capability Statement is helpful.   ML Touchstone test script artefacts need investment.  See https://touchstone.aegis.net/touchstone/

See also https://inferno.healthit.gov/

16 Nov Update RH: Updates in for next ballot (not January now).

8 Feb RH: Next ballot confirmed May (deadline March 6, since extended 2 - 3 months)

8 March RH Community being polled on preferred approach to next release from three options: 1. R5 Release with additional Normative content (ie in 2023) 2. Hold R5 3. New content, but not new normative content (looking most popular option).   See https://onfhir.hl7.org/2022/02/28/soliciting-feedback-concerning-the-roadmap-for-fhir-r5/

22 March 2022 RH Expecting September Ballot (hoping for sooner, prior to the working group meeting? - RH)

2022-05-17 RH Ballot expected later. Content freeze Aug 21

2022-8-23 RH R5 Deadline now September 4  

New change proposals for $validate-code

@Peter Jordan (Unlicensed)

G Grieve have created new HL7 Jira tickets: https://chat.fhir.org/#narrow/stream/179202-terminology/topic/Change.20Proposals.20for.20.24validate-code.20post.20Connectathon

Grahame Grieve: https://jira.hl7.org/browse/FHIR-33941: $validate-code should return system | version | code when display is returned

Grahame Grieve: https://jira.hl7.org/browse/FHIR-33943: $validate-code should return an operation outcome with the full details of issues found, including paths

Grahame Grieve: https://jira.hl7.org/browse/FHIR-33944: $validate-code and $expand should have a parameter for related value sets and code systems

Grahame Grieve: https://jira.hl7.org/browse/FHIR-33946: $validate-code and $expand need a session parameter (as optimisation where ValueSet may need to be uploaded for a series of calls)

Grahame Grieve: https://jira.hl7.org/browse/FHIR-33947: Clarify the use of fragment  (this is a discovered issue from the same session, relating to the first of the tasks above)

ML - general principle that services should echo back the details they've been sent to process.

PJ - CodeableConcept (with multiple codings) will require multiple responses.  ML - call multiple times with single codes as a workaround.

2022-05-03 PJ Agreement to add "Operation Outcome" as a way of returning more detail.

Topics for Terminology Binding Stream

@Daniel Karlsson (Unlicensed)

@Jeremy Rogers (Unlicensed)

Collecting topics for TB stream here, with a view to having a call when there's sufficient material

FamilyMemberHistory - how is "no family history of X" best represented?

FreshDesk ticket question on Allergy substance cross field validation.  DK: Mappings to other information models, is that helpful?

Oct 5 2021 Request for input from nutrition CRG.

8 Feb 2022 MAL Patient Care group proposing the use of an extension (specified by Daniel) to represent Allergy Type to add non-allergic hypersensitivity.  MAL wondered about allowing a Codeable Concept there rather than a Code to allow for greater expressivity.   Resource not planning on being Norminative in R5.

Publishing SNOMED codes in IGs and licencing conditions.

@michael lawley

Licensing issues for IGs referring to SNOMED codes. Is this written down anywhere with some sort of rigour?

Consider the licence statement that is presented when accessing the browser. Should something similar be mandated for inclusion in any document published? What if a patient's medical record were to be published?

Update 12 Jan - "One Page Policy" to be discussed internally.   HL7 agreement indicates other parties would require affiliate licence unless they restricted their usage to the Global Patient Set.

26 Jan Question from PJ about inclusion of COVID Vaccine concepts in the GPS (advised that it is only released yearly)  FYI @Rory Davidson See also MF discussion here.  SuzyR will follow up.  See also early preview page.

06 April: SR covid vaccines have been/will be incorporated into July release. They will be in GPS when ready. 

8 Feb 2022 This was discussed at a CSRM meeting - @Peter Williamsfollow up with @Suzy Roy .  DK Also do we specify boilerplate copyright statement?  This should also be specified in our IG (or referenced).

2022-04-19 Beta release of the IPS sub-ontology (which will include relationships but not axioms)  see https://www.snomed.org/snomed-ct/Other-SNOMED-products/sub-ontology-international-patient-summary due for Tues 3 May.    ML concerned that having the codes in a different CodeSystem breaks interoperability - further discussion required on this point internally @Kai Kewley @Alejandro Lopez Osornio @Suzy Roy

Language Reference Sets in FHIR

All

Mechanisms for working with Languages

Designation extension

24 August DK Added pull request to Snowstorm  here

Implementation by ML using existing fields and specifying LangRefset in displayLanguage parameter with "preferredForLanguage" code and wildcard  Eg https://r4.ontoserver.csiro.au/fhir/ValueSet/$expand?displayLanguage=*-x-sctlang-20581000-087109&_format=json&url=http://snomed.info/sct/20611000087101?fhir_vs=ecl/160303001&includeDesignations=true

Oct 5 Release of Ontoserver 6.5.0 which includes support for language reference sets as above.

Next Steps:   How to promote - could we present at a Vocab call?   (1530 ET Thursdays - may clash with SI Languages call)

8 Feb 2022 DK  added Designation Extension into the SNOMED on FHIR IG see http://build.fhir.org/ig/IHTSDO/snomed-ig/StructureDefinition-designation-use-context.html

SNOMED FHIR Implementation Guide

@Peter Williams

Implementation Guide for using SNOMED CT with FHIR.

IG Documentation: http://build.fhir.org/ig/FHIR/ig-guidance/index.html

Also look at the sample IG https://github.com/fhir/sample-ig see build http://build.fhir.org/ig/FHIR/sample-ig/

TODOS 

• Check build issues http://build.fhir.org/ig/IHTSDO/snomed-ig/qa.html    Update 2022-06-20 Tooling currently giving a null pointer, see https://build.fhir.org/ig/IHTSDO/snomed-ig/failure/build.log  asked question in Zulip  See https://chat.fhir.org/#narrow/stream/179252-IG-creation/topic/Null.20pointer.20in.20ValueSetRenderer   

• Note page to manually execute build through HL7 Tooling: https://fhir.github.io/auto-ig-builder/builds.html

• Rerun locally with latest tooling (also sushi)

• Run with local snowstorm provided as Tx

• See Mark Kramer presentation   https://www.youtube.com/watch?v=OmcSuAMf20s  Also https://fshschool.org/ 

• CONTENT! 

• Merge in DUC branch (Designation Use Context) with a view to implementing in a branch of Snowstorm and making available via connectathon.ihtsdotools.org

• Clean up other branches

• Check templates are up to date - should now be self updating, see https://github.com/FHIR/sample-ig (fsh may ask to update this structure).   Sushi sending out information about new format (PJ says leading whitespace is now significant).

• ReadMe should provide links for Publisher tooling

• Add discussion on Refset / IN (see above)

• Provide link to it in http://build.fhir.org/snomedct.html and discuss further socializing.

• TerminologyCapabilities caching is overly aggressive.

• Added 2022-08-23 Discuss ownership, hosting and copyright (joint copyright with HL7).  Source will be hosted in GitHub, pull requests will be welcome with this group acting as gatekeepers (reviewers) for that?   @Suzy Roy would the wording of our original collaboration be helpful here?

Expanding ValueSets across multiple Editions.

Discussion on * version wildcard. (relates to a Zulip discussion on expanding a ValueSet against multiple editions to give a superset?)  DK How could we specify the latest version of multiple modules.   Answer http://snomed.info/sct/*   (ie not specifying the version) or http://snomed.info/sct/*/version/*  to include concepts from previous versions.

For use in the module / version URI.   Existing workaround is to use multiple include elements for each version.

7 Sept ML Sticking point - what do we do with incompatible properties eg active/inactive between two different versions - the most recent?

Use case: primarily being able to validate codes that might exist in different editions.   Secondarily, being able to validate code received in a historical record ie where codes were valid at time of data entry.

25 Jan 2022: Snowstorm will already look across all Editions loaded when a codesystem version is not specified, eg:  https://browser.ihtsdotools.org/fhir/CodeSystem/$validate-code?coding=http://snomed.info/sct|443971000124108 only exists in the US Edition.  (See also the History proposals for ECL)

2022-05-17 ML Ontoserver supports ValueSets that draw from more than one version, so you could use that to compute the difference between one version and another.

Long Term Home for Terminology Information

@Rob Hausam (primarily Vocab WG)

Long Term Home for Terminology Information that can be maintained independently of FHIR versions.

See https://chat.fhir.org/#narrow/stream/179202-terminology/topic/Why.20does.20the.20spec.20contain.20a.20SNOMED.20CodeSystem.20resource.3F

7 Sept Relates to problematic page https://www.hl7.org/fhir/r4/codesystem-snomedct.html   PWI discuss internally - done.

Update 21 September 2021 - General agreement that that page can be removed.  Is out of date and not serving useful purpose.

Update 25 Jan 2022 - HL7 intend to make The THO https://terminology.hl7.org/ a home for all terminologies and valuesets.

Next Time: Move to archived discussion, make note in IG

IPS / GPS progress

@Suzy Roy

@Alejandro Lopez Osornio

IPS Terminology Subontology - beta went out 2 weeks ago, now in feedback until June 30.  Publication for the end of the year.  ML expressed issues around not being conformant to RF2 spec but is missing module dependency file.   ModuleId is present but invented Concept.    ALO/SR suggested ML give that as feedback.     

GPS expected towards the end of September. Currently working with contributing organisations, which will then go to internal authoring.   ALO GPS may turn out to be less useful and focus could be on IPS in future.

Any Other Business

Quick Questions:  

Next time:  Need some way of identifying which language reference sets are "patient friendly" aka "consumer"  (eg make those specific reference sets have an additional parent that indicates their use CS "Explanatory Terms".   Alternative solutions: annotations/metadata )    2022-05-17 MAL Daniel presented at translation User Group - looking for definition of "what is a patient friendly / consumer.  See here. ) 2022-08-23 PWI If we go forward with annotations, would a flag for "consumer" be one of those values?  Alternatively (or additionally) do we want a common ancestor for these refsets.  To add to agenda for the Joint AG session on the Wednesday PM September Business Meetings.

2022-09-20 What was the outcome of the IPS Subontology feedback?

SNOMED Family of Languages

@Peter Jordan (Unlicensed)

@michael lawley

Impact of proposed changes (eg text searching in ECL) on FHIR. Questions around which language reference sets to use when there are multiple, especially partial/overriding context (referred to MAG for discussion)

8 Sept The FHIR specification does not specify a particular version of ECL, so we assume the latest.  Any enhancements added to ECL will be immediately relevant and available in FHIR.   Note that these latest additions while targeting descriptions are a concept filter, so display options (language etc) will affect the output of those concepts.   How about the filter parameter though, especially since ECL wo (ld allow multiple filters in multiple/different languages.

Specify CodeSystem in FSH

@Daniel Karlsson (Unlicensed)

FSH apparently has no way of specifying the version of a CodeSystem. Daniel checking the ANTLR spec. https://github.com/FHIR/sushi/issues/473

API for FHIR Resource ↔ Post coordinated expression mapping

@Peter Williams

API for FHIR Resource to SNOMED Expression

Looking up an SCTID in an unknown module

@Peter Williams

Problems when dependencies do not align. Multiple code system resources represent multiple editions / versions.

ML: See code parameter to code system search. Should return code systems (ie versions) where that code is defined. International concepts would appear in every edition known to the server.

eg /CodeSystem?system=htp://snomed.info/sct&code=12345678&_elements=version

GPS

See Discussion on Global Patient Set (GPS)

FHIR Shorthand

@michael lawley

https://github.com/HL7/fhir-shorthand/wiki

https://build.fhir.org/ig/HL7/fhir-shorthand/FSHQuickReference.pdf

Tooling: https://github.com/FHIR/sushi