Welcome and introductions

2

@Peter Williams

@Rob Hausam

Recording, notes & attendance.

Check questions in Zulip and SNOMED International #snomed-hl7-fhir

Previous Meetings

2

@Peter Williams

@Rob Hausam

• 2024-11-12 - SNOMED on FHIR Meeting (TS & TB)

• 2024-10-29 - SNOMED on FHIR Meeting (TS & TB)

• 2024-10-20 - SNOMED on FHIR Meeting (TS & TB)

Other Meetings

10

@Peter Williams

@Rob Hausam

Recent Events:

European XpanDH project closing remarks link future direction link.

Belgium Connectathon 25 - 26 November - Hospitals on FHIR, Connectathon, FHIR User Day.

See LinkedIn Update from Rein Wertheim: https://www.linkedin.com/pulse/4-takeaways-from-hoh-user-days-2024-rien-wertheim-nv3ne

See also OMOP + FHIR eg https://build.fhir.org/ig/HL7/fhir-omop-ig/

FHIR Camp - Portugal 28 Nov https://www.health-samurai.io/events/fhir-camp-2024

NZ December 2 - 3 SNOMED & FHIR workshops  https://hinz.eventsair.com/dhwnz24/pre-conference-workshops

Future Events:

HL7 January 2025 27 - 30 EST  Virtual.  Connectathon 14 - 16 

HL7 Europe WGM - Lisbon - 10-14 Feb 2025

Dev Days 3 - 6 June 2025 Amsterdam

HL7 May 10 - 16 Madrid + 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 

HL7 Cross group projects : CDA & FHIR Translation Meetings (weekly): @Jay Lyle Going well, there is an implementation guide and the group is doing connectathons. May in Dallas and Minneapolis.  See https://confluence.hl7.org/display/CGP/C-CDA+to+FHIR+and+from+US+Core+Mapping

Hospitals on FHIR, Connectathon, FHIR User Day

Round Table Introductions & Updates

All

Previous Notes from April 2024 Business Meeting

Australia  - Sparked FHIR Accelerator Programme - https://confluence.csiro.au/x/39Fqa

• AU CDI – including binding to ValueSets

• AU Core

• eRequesting

• Ontoserver now registered for use with HL7 IG Tooling allowing ECL-based ValueSets in IGs

@Reuben Daniels (Unlicensed) NB National Extensions of SNOMED need to transition to be self hosted rather than Tx.fhir.org end of 2024.

CSIRO @michael lawley now representing the NRC.  Terminology Registry - Ontoserver registered as authoritative.  Increasing interest from affiliates.   Interest in LLM for mapping and matching/ranking in VS.  

Austria / Germany - Annotating clinical narratives.  GemTex - annotated corpus (annotation using SNOMED CT) relations mapped to SNOMED patterns.   Looking for more support for mapping FHIR ValueSets to SNOMED concepts - particularly when involving negation, would like to collaborate on common solution.  2024-10-20 @stefan.schulz (Unlicensed) Focus on translation and indexing (interfacing terminology), analysis of clinical narratives AIDVA - Guide for Annotating Clinical Narratives (SCT + FHIR)

B2i @Márk Czotter (Unlicensed) SnowOwl Terminology Server supporting multiple versions of FHIR at same time.

Belgium  - National architecture level work being done, to consolidate regional structures.  Datasets for Allergies and Vaccinations,  Care Plan.  Now working on Medication Resources well in progress, Obs and Conditions followed by Procedures.

Canada  - 

Estonia  - Medication schema nearly complete.  2024-10-20 Ministry of Social Affairs, switching from CDI3 to FHIR.  Issues with document signing on a per Resource basis (possibly enough to prove use of secure gateway), plus issues with data access & permissions.

Europe - Laboratory Guide published 2024-10-20 (@Daniel Karlsson (Unlicensed) ) Ballot to make Lab work a global standard, some issues representing panels.  IG for European Patient Summary. Group working on VS for Lab.  HL7 Europe projects. "My Health EU" supporting cross border and lab result sharing for 2025 using HL7 Europe IG.   "Extended EHR" - logical models to support prioritised areas eg lab, patient summary and IGs based on same.

France -  

Hong Kong  - 

HL7 May Ballot for comment

India  - 

Netherlands  -  Looking for dissemination of patient friendly descriptions + longer explanations.   Using FHIR Supplement.  May look at converting more code systems to FHIR eg IDMP (see @Rutt Lindström) 2024-10-20 National Terminology Server (Ontoserver) Patient friendly synonyms and explanations. 

New Zealand -  Cancer control agency building FHIR / SNOMED platform for sharing cancer data (multiple domains) with query based FHIR Valuesets.  Interested in hearing about: standardizing language reference sets, access to MRCM (see item 20 "MRCM in FHIR" demo being progressed), Concept Maps - logic for dependsOn (see Concept Map in R5) & how to select ValueSets based on context,  Syndication (see https://github.com/IHTSDO/termserver-syndication  but this is not FHIR. FHIR uses more of an NPM approach), Query Specification (Query based reference sets their expansion).  @Peter Jordan (Unlicensed) Affiliate Engagement Group.  Notes that 30% of (personal) primary health record codes did not pass validation.  Recommends LinkedIn FHIR group, see https://www.linkedin.com/in/darrendevitt/ Notes that Tx.fhir.org never supported ECL hence need for National TS to come into play.   See also HealthPoint use of SNOMED CT presentation.

CCA @Linda Bird (Unlicensed) 2024-10-20 NZHTS National TS (Ontoserver) using CanShare language reference set for patient friendly terms.

Norway - 

Sweden - Swedish national FHIR profile 2024-10-20 HL7 Sweden discussing TS for building Swedish IGs.   Existing TS is MVP so procurement underway with growing requirements.  Alerting work collaborating with Norway eg question of what is alert worthy

Switzerland - 

Snomed International - internal work to reduce configuration using archive hierarchy with cached metadata.  Request for MLDS badges.

Spain  - (SS update) RESQ+ Harmonizing data acquisition forms for stroke, looking to standardize form across several languages.  Difficult to work with clinicians directly in SCT, more success using SCT in FHIR context

United Kingdom - UK Core based on R4, developing IGs and Profiles.  Medicines, GP Connect Profile, National Terminology Service used for mappings and ECL expansions (looking to use for primary publication point).   Pathology team developed IGs, challenge moving from legacy arrangements.  OpenEHR being revisited.   Interested in SNOMED Extended maps in FHIR (see Concept Map in R5 ) and making it more automatically executable.  2024-10-20 @Stuart Abbott (Unlicensed) Improved dose syntax

Daedelus support services for multiple countries using Ontoserver.  Running "Learnathons" with InterOpen

United States - US CDI draft of v5.  Cancer on FHIR - connectathon at HIMMS

Regenstrief @Robert McClure (Unlicensed) Gravity - Social determinants of care.

Magic Codes Zulip Discussion

@Peter Williams 

@Daniel Karlsson (Unlicensed) 

@Rob Hausam 

Magic Codes and Observable Entities as Questions and Clinical Findings as Answers

https://chat.fhir.org/#narrow/channel/179166-implementers/topic/SNOMED.20Magic.20Codes.20.28RESTful.20API.20Observation.20Queries.29

Support for Clinicians in using Information Models

@Kusal Wijayaweera (Unlicensed) 

Suggested resources:

Introduction - Clinical Reference Groups

https://confluence.ihtsdotools.org/display/CP

SNOMED CT Clinical Implementation Guide for Allergy, Hypersensitivity and Intolerance

FHIR Observation Observation VS for Vital Signs

@Rob Hausam 

HL7 Orders and Observations Workgroup looking at Observation profiles for Vital Signs and Value Sets for what is considered to be a vital sign and need to consider SNOMED CT (possibly also NPU) for analogous value sets, started with LOINC.  DK We have looked at this - suggests using the superset of this material and allowing users to work with a cut down set.  Concerns that this might need to be context specific.  SA, UK has also looked at this. 

SH Could take forever to agree a set of vital signs, better to specify archetypes of data required.

Post Coordination

@Feikje Hielkema (Unlicensed) 

Information models as expression of post coordination → classifying to determine equivalence with pre-coordinated content.

Would like to associate a FHIR Resource with a SNOMED Template 

@Linda Bird (Unlicensed) noted (from previous work done) that the solution involves first defining the template and then working out the transformation to comply with the concept model.   DK notes that this equivalence with the SCT model does not always correspond.  SS notes problem with representing (specifically in modeling/classifying) negation in SCT.

FHR Lack of documentation that explains 'this is how to approach the problem' 

Question on de-labelling allergies

@Kusal Wijayaweera (Unlicensed) 

PWI Suggested Verification Status = Refuted here: https://build.fhir.org/allergyintolerance.html (DK See previous work done mapping to SCT Free SNOMED CT set for FHIR eg 723511001 | Refuted (qualifier value) | / 410594000 |Definitely NOT present (qualifier value)|)

See also SNOMED CT Clinical Implementation Guide for Allergy, Hypersensitivity and Intolerance

And https://ihtsdo.github.io/sct-implementation-demonstrator/#/allergies

HL7 and OpenEHR

@Maryam Razavi 

@Alejandro Lopez Osornio 

https://www.pulseit.news/australian-digital-health/hl7-and-openehr-considering-closer-ties-and-standards-alignment/

https://www.youtube.com/watch?v=qbq-kuWyTcg

Use of Codeable Concept to transmit multiple codes

@Kath Priest 

@Daniel Karlsson (Unlicensed) 

EU requirements for transmitting closest SNOMED International codes along with relevant local code systems is a great use case for the new ECL Top and Bottom operators.   Having this requirement come from the Member Forum would assist in ensuring that the implementation of this feature into Snowstorm is prioritised.

@Daniel Karlsson (Unlicensed) 

Note that Top and Bottom operators simplify access, which can be replicated in earlier implementations of ECL with a more complex query:  6.12 Top and Bottom

ALO See also Interoperability Demo (last tab) at https://ihtsdo.github.io/iid-ips/ which offers the closest concept in the IPS Set, along with the ECL used to obtain it.

Translation in Code System Supplements

@Daniel Karlsson (Unlicensed) 

Translation in Code System Supplements.   See Zulip discussion 

https://chat.fhir.org/#narrow/stream/179202-terminology/topic/CodeSystem.20supplement.20use.20cases

See also https://build.fhir.org/ig/FHIR/ig-guidance/languages.html#creating-a-language-pack

Terminology Binding in the Condition FHIR resource

@Katrien Scheerlinck (Unlicensed) 

There is currently a discussion going on in Belgium regarding the desirable terminology binding in the Condition FHIR resource. We are working together with a group of clinicians/domain experts who almost all indicate that there are a whole series of finding concepts in SNOMED CT, which are required in the use case of diagnosis registration in the patient.  As well, the information website (https://build.fhir.org/condition.html) on this FHIR resource indicates that, in case of persistent symptoms or in case there is no official diagnosis yet, only symptom identification as a preliminary diagnosis, concepts from the finding hierarchy may be used:  “…alt

hough in some cases a persistent symptom, e.g. fever, headache may be captured as a condition before a definitive diagnosis can be discerned by a clinician.”

What would be your opinion on this? Hoping this helps to give a direction to the discussion including good, substantiated arguments.

ML See https://build.fhir.org/ig/hl7au/au-fhir-core/StructureDefinition-au-core-condition.html

LB Has anyone looked at the alignment of the verification status and explicit context concepts?

PWI This group looked at this as part of the proposed "Free Set" in 2019 - see FHIR SNOMED CT Free Set - Questions for HL7 (Amber mappings).   Consider the likes of << 41769001 |Disease suspected (situation)| and << 373573001 |Clinical finding present (situation)|

ML Existing issues: “probably not present” is a child of “Known absent” where the term “probably not” indicates that it is NOT a child of “Known absent”

Resources for Allergy Intolerance / Adverse Events

@Kusal Wijayaweera (Unlicensed) 

Question on Allergy Intolerance resources.  Group responded with:

SNOMED CT Clinical Implementation Guide for Allergy, Hypersensitivity and Intolerance

https://build.fhir.org/ig/HL7/fhir-ips/ValueSet-allergy-intolerance-snomed-ct-ips-free-set.html

https://ihtsdo.github.io/sct-implementation-demonstrator/#/allergies

March 1, 2024 Allergy CRG meeting agenda

Previous discussions listed in : https://snomed.atlassian.net/wiki/display/FHIR/Discussions

Allergy Intolerance vs Adverse Events

TODO: Archive

How to handle ad-hoc post coordination in medical records

@Kusal Wijayaweera (Unlicensed) 

Question about EHRs containing unconnected concepts - like post coordination but not using SNOMED Compositional Grammar.   Like 12345005, 5678004 (which is syntactically invalid according to SCG) or 185324002 |Failed encounter| + 281399006 |Did not attend (finding)| (UK) + 275694009 |Patient defaulted from follow-up (finding)| (this would actually be valid syntax as these are considered to be parents, but would need to come from same hierarchy).  See also 410543007 |Did not attend (qualifier value)|

ML Well expressed post coordinated expressions are allowed in FHIR, but the challenge is what might this mean to some client system.  AP Suppliers are particularly resistant to receiving even the simplest post-coordination and would prefer to have atomic concepts split out to make use of the information model (might even get stored as free text if supplied but unexpected).

PWI Specific example allows for general rules to be determined eg Failed Encounter is redundant due to more specific concept in list, or that we could particularly flag up concepts which come from separate hierarchies as needing intervention.

IPS ValueSet Question raised

15

@Alejandro Lopez Osornio 

https://jira.hl7.org/browse/FHIR-44904.  "Resolve SNOMED Version challenge with value set"

Canada specify version details for each filter in a valueset specification, which then trips validation when the version being expanded against does not match.

DK: If you use extension content then you need to specify the version to know where the concept exists.  Although the lookup will work across multiple versions.   ML:  Note that compose/include/filter descendent of that specifies the core edition, and so it would not include Canadian concepts.   Also, a version - if included - should include the effective time.

Discussion about sending both the local code plus the IPS / International Code using a Codeable Concept (in that case the version element of each code would indicate which one is International.  Or a test could be made for subsumption) which could be determined using (ML supplied) ECL:

See also https://ihtsdo.github.io/iid-ips/ and be aware of new ECL top and bottom operators which could be used to simplify this.

MRCM in FHIR

@Linda Bird (Unlicensed) 

Exposing the MRCM in FHIR

KK Exposing it directly would be difficult because we can return lists of concepts, but we cannot return a string eg ECL

ML Thought it could be exposed via the CodeSystem as another property - metadata about each domain's range.

LB Use case is driving an expression or ECL editor (KK Remember content type).   KK choice between (simple) server statement vs (more work) implicit expansions.

General Discussion

2024-12-18 - SNOMED on FHIR Meeting (TS & TB)

All

Misc Discussion

Internal discussion on languages in general and specifically latin words.  MAL suggested looking at https://iate.europa.eu/home which includes latin.  See Latin Terms - Discussion

Snowstorm-lite revisited (see 2024-01-09 - SNOMED on FHIR Meeting (TS & TB) )

URIs and Identifying Modules - "Edition identifier" - this is the full URI, not just that module. "Most dependant module" - yes but fails in the International Edition.  "Identifying Module". "Root module" - Not always the top/bottom of the hierarchy.

Discussion on adding / removing modules to / from packages.   Check on license conditions allowing for automatic promotion to ensure interoperability.

Discussion on expansion related to 5.2.1.7 Query Specification Reference Set.  ML Could we try to expand the referenced component id, ie look in this refset if no refset members are initially encountered?  MLO Canada is already doing this and they have pre-expanded simple reference sets.  KK Snowstorm does not use this reference set.   LB An extended version of this refset has been drafted which allows the substrate to be specified.  

New Slimline Agenda 

5

@Peter Williams 

See discussion pages for links to archived discussion:  Discussions

Post Coordination

@Kai Kewley 

Moved in from "Potential Topics": 

API for FHIR Resource to SNOMED Expression

See also discussion on URIs for PCEs. 2024-03-05 - SNOMED on FHIR Meeting (TS & TB)

Additional information from ComplexMap $translate

@Daniel Karlsson (Unlicensed) 

ICD-10 Complex Map in ConceptMap$translate (see agenda item 18 2024-03-05 - SNOMED on FHIR Meeting (TS & TB) )

DK Suggests using match.product for each additional item with a URI for each column in refset as per 2.6 URIs for Properties

KK The R4 Products are flat and various ICD-10 rules are grouped.  Example here:  https://snowstorm.ihtsdotools.org/fhir/ConceptMap/$translate?code=421671002&system=http://snomed.info/sct&targetsystem=http://hl7.org/fhir/sid/icd-10

FHIR Resource cross field validation

10

@Marie-Alexandra Lambot

Question:  Can we do cross field validation of FHIR resources, to restrict valueset available for one element dependent on the value entered in some other field.

DK Do this using "Invariants".  See https://build.fhir.org/conformance-rules.html  and https://build.fhir.org/profiling.html#slicing (complex approach!)

ALO FHIR Questionnaires have more support for cross field rules.

LB - would specify this in the Implementation Guide with the specific rules.

@stefan.schulz (Unlicensed) 

ValueSets eg for procedures which recommend SNOMED CT codes that have been retired.

ML: ValueSets described via enumeration will need to be kept updated.  The party that specified the ValueSet is primarily responsible for maintenance.

RC: License requirements do imply that SNOMED CT versions should be kept up to date in systems. 

FHIR in the new SNOMED Implementation course

@Marie-Alexandra Lambot

@Kath Priest  

The new implementation course could focus on SNOMED + FHIR, and transformations between information models, using maybe the Allergy demonstrator as an example. Maybe adding consideration on adding LOINC concepts, like the allergy test using LOINC (Observation in FHIR).

Suggestions: add onset to reactions (date), and use problems already recorded for reactions manifestations, and vice-versa.

Helpful Introduction Videos:

Introduction to SNOMED with FHIR YouTube recordings of Dev Days Presentations:

https://www.youtube.com/watch?v=40Lvv2t8OxU

https://www.youtube.com/watch?v=x6UqJtcJwmQ

https://www.youtube.com/watch?v=8cXRAtNoSvc

April 2024 Business Meeting General Discussion Topics 

@Ronald Cornet (Unlicensed) 

@michael lawley 

FHIR for Terminology vs FHIR for Patient Records - eg for cohort identification.  See https://www.hl7.org/fhir/search.html#modifierin can try out on https://hapi.fhir.org/ Also CSIRO Pathling server for analytics 

Navigational ValueSets - how can you express hierarchy and/or grouping in a ValueSet expansion?  Nested code can be used to show hierarchy (navigational concept could be represented as blank).  Question is more about how you initially input that structure such that it would take effect.  Note that valueset composition (valueset containing valuesets) is not sufficient because it does not infer any structure, only membership.   Note that in this context navigational groupers are not considered to be concepts - they don't have a CodeSystem identifier.   SH: users would prefer to search than to navigate their way through a hierarchy to the item they want.  "Picklists" can be considered distinct from ValueSets (adding priority and other properties).  ML Usage in questionnaires.   "The way to do this is in the expansion"  DK Solution also available using Concept map translate given  'source' context, what likely values would be selected in a subsequent field. RH See also ValueSet Concept Order Extension. PWI "Tagging" may be preferred as it allows for multiple dimensions to be represented at the same time ML and also avoids codes appearing multiple times if they belong to more than one group. SH Would like to see standard for wider sharing of resources like reusable UI elements, and avoid overloading FHIR (separation of concerns).  SS Much harder to add semantics to existing data, than it is to enforce the semantics at the point of data capture.

TODO Elicit use cases from @Rutt Lindström 

Condition vs Observation - care plan based on symptom.

Other CodeSystems which reference SNOMED CT

General discussion on work being done in EU

@Marie-Alexandra Lambot @Daniel Karlsson (Unlicensed) 

See https://build.fhir.org/ig/hl7-eu/laboratory/branches/master/index.html

https://github.com/hl7-eu/laboratory

https://confluence.hl7.org/display/HEU/Laboratory+Report+Implementation+Guide%2C+Edition+1

https://build.fhir.org/ig/hl7-eu/xpandh-hdr/

ML: https://csct.be/projects.html

https://nuva.mesvaccins.net/

2023-09-19 ML CSCT now has official "Not for Profit" status as an entity.

IG Tooling Compatibility

20

Requirements for making a Terminology Server compatible with IG Tooling

See Zulip discussion here:  https://chat.fhir.org/#narrow/stream/179202-terminology/topic/Terminology.20server.20requirements.20for.20IG.20builds.2E.2E.2E

2023-07-25

• @michael lawley is implementing the requirements in Ontoserver

• Graeme is preparing the Terminology Server registry, probably a JSON document, that would make available different terminology servers, listing the capabilities, including which code systems on each server

2023-08-22

• @michael lawley the scope has expanded, advancing on the implementation. The registry will document which terminology server meet requirements, it is a centralized resource, maintained by HL7. There is document that describes this. An implementation guide developer would use this information to choose a compatible terminology server as the backend of the IG valuesets. There is a test suite that servers can apply to demonstrate compatibility, Graeme will be the curator of the list, the submission process for a new server is under development.

2023-10-03 ML Ontoserver still a work in progress in this area.

2023-10-22 ML Ontoserver progressing to be trialed during NZ Connectathon in this context.

• Test suite is here: https://github.com/FHIR/fhir-test-cases.git

• They are executed via the valiadtion-cli, something like:
  java -jar org.hl7.fhir.validation.cli-6.1.3-SNAPSHOT.jar -txTests -tx https://r4.ontoserver.csiro.au/fhir

Observables model

@michael lawley 

Question on the Observables Model - was discussed in both the MAG and the EAG.   currently looking at proposals and next discussion will be at all staff meeting in June.