Page History

...

Attendee type: M - TSMG Member O -Observer / Guest	Attendance: P - present A - Apologies X - (Apologies with Proxy)	Name	Affiliation
M (Co-Chair)	A	Reuben Daniels	Saludax
M	P	Jess Bota	Apelon, Inc.
M	P	Lisa Nelson	MaxMD
M	P	Michael Faughn	National Institute of Standards and Technology (NIST)
M	P	Davera Gabriel	Johns Hopkins University Institute for Clinical and Translational Research
M (CSDO)		Dan Vreeman	HL7 International
M	P	Rob McClure	MD Partners, Inc.
O		Ted Klein	Klein Consulting
O	P	Marc Duteau	HL7 International
O		Grahame Grieve	Health Intersections
O		Joan Harper	Canada Health Infoway
O		John Snyder	National Library of Medicine
O		Rob Hausam	MITRE
O		Alex Kontur	ONC
O		Ravi Kafle	Washington State Department of Health
O		Abdullah Rafiqi	ICF
O		Hope Gray	UAB - Birmingham, HL7 Intern
O		Martha Jones	George Mason, HL7 Intern

...

Agenda Outline

Time allocation

Lead

Agenda Item

Welcome and scene setting

5

JB

Note that meetings are recorded
Welcome to TSMG members and observers
Rollcall
Today's objectives:
- Continue to go through each element type where differences exist across duplicate resources
Goal is to get a list of resources we consider duplicates or easily categorizable fixes (only have non-consequential differences) and then a list of duplicate resources that need to be consolidated/resolved
From our last discussion
1. Resolve code systems and value sets anchored in THO but only in the FHIR spec (Grahame)
  1. Noting that this includes Gender Harmony value sets
2. Resolve code systems and value sets anchored in THO but exist in THO and FHIR core (Grahame)
  1. Modify build to warn the committees that own them via the build itself
  2. Will list CS and VS that need reconciliation by WG here: http://build.fhir.org/resourcelist.html
3. Resolve code systems and value sets with differing anchors, but believed to be very close to the same content (THO/UTG Subcommittee)
  1. Michael will rerun his analysis based on the FHIR CI build NPM ci build NPM package: build.fhir.org/hl7.fhir.r5.core.tgz
  2. Resolution will be based on CLD vs metadata changes for value sets, and content changes for code systems
  3. THO subcommittee will look into these differences further

Handling THO anchored content in the FHIR spec

JB

Handling THO anchored content in the FHIR spec FHIR-37440 - Many CodeSystems use terminology.hl7.org urls when they should be hl7.org/fhir TRIAGED
- Michael has completed an initial review
  - Assumption is that resources will share <id> (we know this isn't a rule but we need something to link the resources)
  - There are 42 code systems anchored in THO that are only in the FHIR core specification
    - Need to do same analysis for value sets
    - Assume they need to be moved but would like to finish analysis
  - There are 72 code systems and 190 value sets that are potential duplicates across THO/FHIR
    - Continue to discuss how to address changes across the following code system elements:
      - meta.lastUpdated (all) - not consequential
      - date (all) - not consequential
      - url - once moved, should all be anchored in tho
      - experimental (19) - we assume that all code systems may have been set to false upon moving to THO but this requires verification (V2 and V3 initially set to false, but some FHIR content set to true initially), consequential
      - title (23) - need to improve titles, they should have spaces, consequential
      - valueSet (70) - value set should be in the same place as the code system (for all codes value sets), this will get cleaned up with Grahame's cleanup and he will bring it back to us if something else needs to be done, non consequential
        If value set is more specific (enumerate or have cld)
        However, we are looking at CodeSystem.valueSet so this does not apply
      - meta.profile (62) - not consequential
      - description (22) - need to be reviewed, consequential
      - publisher (18) - not consequential
      - name (8) - consequential, involve WGs if needed
      - contact (18) - there were somethings done algorithmically, things in THO had a default value of Health Level 7, this should be made consistent as a bigger effort so non consequestial
        Need to consider adding email, url, and other subelements
      - status (17) - things brought in from R4 and R4B were not changed, if updated as part of R5. Keep anything populated and if both are populated then the R5 value should be used, consequential
      - oid (25) - consequential and require cleanup, need to understand how this is happening from Grahame as there are duplicate OIDs across and within THO and FHIR (urn:oid:2.16.840.1.113883.4.642.1.0)
      - property (1) - this one change was added to THO so THO property needs to persist (in measure-type), but should remove the property definition header since we are using the standard, consequential
      - hierarchyMeaning (1) - if blank, it is considered is-a, non consequential
      - extensions
        telecom - nonconsequential
        standards status - looks like lots of errors in R5 where 'cds' was incorrectly added, keep anything populated and if both are populated then the R5 value should be used, consequential
        workgroup - use r5 value if present, consequential
        fmm - looks like lots of errors in R5 where 'cds' was incorrectly added, keep anything populated and if both are populated then the R5 value should be used, anything with fmm 5 is likely an error that should be removed, FHIR value rules except for cds ones, consequential
      - concepts (18) - consequential, involve WGs if needed
    - Discuss how to address changes across the following value set system elements:
      - name (29) - consequential, involve WGs if needed
      - oid (5) - consequential and require cleanup
      - status (19) standards-status (10)- Needs manual review, there is no consistency (things going from active in FHIR and draft in THO, and vis versa, consequential
      - title (63) - need to improve titles, they should have spaces, consequential
      - description (25) normative-version (1)- need to be reviewed, consequential
      - experimental (159) - we assume that all code systems may have been set to false upon moving to THO but this requires verification (V2 and V3 initially set to false, but some FHIR content set to true initially), consequential
      - immutable (123) - everything is already set to true, non consequential
      - compose.include.system (19) - all should be fixed once urls are resolved, need to verify testscript-operation-codes (only this one is consequential)
      - date (180) - not consequential
      - extensions
        fmm (25) - looks like lots of errors in R5 where 'cds' was incorrectly added, keep anything populated and if both are populated then the R5 value should be used, anything with fmm 5 is likely an error that should be removed, FHIR value rules except for cds ones, consequential
        standards-status (10) - non consequential, all values should be removed at the approval of TSMG
        normative-version (1) - non consequential, remove the one value, it is an error
Opportunities for improvement of THO content
- Determine a standard publisher and perhaps contact and telecom
- Vocab should recommend to TSMG that standards-status should not be used in THO and FMM should be used instead
- Pay special attention to StatisticAttributeEstimateType and StatisticStatisticType as there are strange inconsistencies with title, fmm, etc. and may need steward to investigate
Marc and Jess to split up the code systems to start and work on consolidation
- Goal is to have one XML for any resource considered a true duplicate
- Will bring back to discuss any challenges, special cases, etc.
Michael to extract list of stewards using the following extensions:
```
http://hl7.org/fhir/StructureDefinition/structuredefinition-wg
```

Adjournment

0

JB

Will repurpose We are not sure if we will have someone to run this call again next week!.

Call adjourned at X4:XX 51 PM ET.

Page tree

Versions Compared

Old Version 1

New Version Current

Key

Action items