HL7 Working group meeting San Diego

September 10-15, 2017

September 10, 2017

Name / Organization / Q3 / Q4
Ted Klein / TKCLLC / X / X
Joel Schneider / NMDP / X / X
OyvindAassve / Director of Health, Norway / X
Rob McClure / MD Partners / X / X
Jim Case / IMO / X / X
Rob Hausam / Hausam Consulting / X / X
Greg Gustafson / PenRad / X / X
Carmela Couderc / IMO / X

Vocab Q3 – Agenda planning

Agenda updated with chairs, topics and scribes

UTG PSS – Ted

Ted reviewed the work that he and Grahame have been doing the last cycle

-Workflow has been developed as a Jira project – UTG Proposal Creation Project

-Draft PSS has been developed

-Ted discussed an overall scope of the UTG project and a proposal to build tooling to support the “new” harmonization process. Presented to CTO.

-Demonstration project for the prototype of the tooling can be created by Now Orleans provided funding is available.

  • This is why the PSS was developed, to support the prototype development

-Ted asked for a vote based on the initial draft (attached)

-Discussion –

  • Q: The PSS is assuming that there is a terminology store that can be updated by the harmonization process. However, the PSS states that the project will create the terminology store. A: The prototype currently maintains the separate terminology silos, but these would be integrated under the UTG tooling project.
  • The Project scope was edited to reflect the integration of the separate stores into an single source. Concern is that some people may not want to move to a single source for terminology, but this has been presented as part of the CTO report for the last three WGMs so this should not be an issue, recognizing that there will be pushback by individuals who do not believe that this is an appropriate direction.
  • The scope will be to replace the current harmonization processes with a more focused integrated process.

-Discussion review and voting will take place next quarter.

A question arose as to whether the value set maintenance process being discussed by Structured Documents should be a part of the scope of this project? A new process has been proposed that CDA value sets are published in the ballots and also restructured and put into VSAC.

Lantana has published the results of a project to update the value sets in VSAC to correct the current errata for CCDA and to review the value sets to make sure that the descriptions of the value sets are correct; and to provide guidance about how to maintain value sets by workgroups in the future. There is also a project to align the value sets as represented in VSAC and some in Trifolia. Currently it is assumed that the source of truth for value sets for CCDA is VSAC.

The value set maintenance process document from Lantana has been published, but it is currently just sitting. No action has been taken on it, although a request for comments has been made.

The scope of the UTG right now is internal vocabularies, thus inclusion of external code systems and externally authored value sets are out of scope for the UTG process.

ACTION: The Lantana document should be read by the vocab co-chairs prior to the discussion at structured documents. A request will be sent to Don Lloyd by Rob McClure to get a link to the document.

Ted reviewed the changes to the UTG PSS.

Motion: Accept the edited PSS as presented. Moved Rob H./Rob M. In favor/Opposed/Abstain – 6/0/1

Workgroup Project Review

Action Item Review

Madrid Items

-FMG discussion on terminology services capabilities – to be done in San Diego

-Send a reminder to Heather and an action from today:

  • Heather will take the MetaReq 11179 related document and send it to the Vocabulary WG for comment along with the comment form. Send comments back to Heather who will collate the comments and submit to ISO.

-V2 versioning (Ted) – will be discussed in joint session with InM this week.

San Antonio Items

-Draft policy among vocab, conformance and InM for best practice re: v2 code content. – Will be discussed during our joint here in San Diego.

-FHIR TermInfo PSS (Rob H.) – Work will not be started until after this WGM.

-Examples of v3 value sets in FHIR specification format (Rob H.) -

  • ACTION: Schedule a discussion on the use of multiple identifiers and versioning and how to represent the source of truth for value sets. Rob M. (Suggested Thursday Q0)

Conference call actions

-Review Harmonization training (Ted) - may be unnecessary in the face of UTG progress and plan.

-Add the item for a needed policy statement on SCT allowed binding strengths in FHIR to the HTA agenda for clarification that Rob H and OO feel is sufficiently clear to address their needs (Heather)

  • ACTION: Ask Sandy about this in San Diego.

-Update the Glossary spreadsheet as per the 8/1 call and redistribute for the 8/14 call (Heather) – no update

-Reach out to Michelle W and contacts at NCHS to get more information to resolve FHIR tracker item #12783 on URIs for ICD-9-CM for further discussion in San Diego (Rob H, Carol M)

  • ACTION: Rob H will follow up in San Diego.

-There is some problem with the glossary spreadsheet on Google docs, it will not open and needs to be updated. Please send updated spreadsheet to Ted so it can be distributed and discussed on the 8/31 call (Heather) – no action

Project Review

All of the action items for which vocabulary is the sponsor were reviewed quickly.

Project 1247 – has not progressed, but needs to develop a formal project for approval. Will add a call slot on the next vocab conference call to figure out how to progress it.

Project 1146 – Need to ask Conformance to close this project (Ted)

Project 1022 – Project is complete and the STU has expired. Need to see why it is still active on the list. Russ will follow up.

Project 874 – Ted needs to talk with Publishing in San Diego

Project 630 – Need to update the project metrics (Rob M)

Monday, September 11th, 2017

Q3 Monday Project 1047 Characteristics of a formal ‘value set definition’

Chair: Rob McClure

Scribe: Rob Hausam

Attendees

  • Ted Klein
  • Carol Macumber
  • Susan Barber
  • Carmela Couderc
  • Sandy Stuart
  • Dr. James Case
  • Dr. Robert McClure
  • Russ Hamm
  • Joel Schneider
  • OgvindAassue

Project 1047 Characteristics of a formal ‘value set definition’ – Ballot Comment Reconciliation

  • Thomas Beal’s comments
  • Motion (Carmela): To accept the dispositions outlined in the spreadsheet to address the comments from Thomas Beale (attached): Second(Sandy). Vote: 9-0-0
  • Comment #1014: Will add additional text fields so there will be license, IP, and Copyright text, each with a slightly different focus and clarify how each would be used differently.
  • Motion(Russ)Moves to support the persuasive items documented in the spreadsheet on adding additional text fields to support be license, IP, and Copyright text. Second: Carol. Vote: 9-0-0
  • Block votes on: Comment #1081, 1082, 1083, 1085, 1086,1087, 1088,1089 comments and dispositions in attached spreadsheet:
  • Motion (Jim): Second (Ted): Vote: 9-0-0
  • Motion (Jim): Take the VSD Normative in the next ballot cycle upon completing the changes. Second (Ted): Vote: 9-0-0.

Q4 Monday – Value Set Expansion and TermInfo Update

Chair: Ted Klein

Scribe: Rob Hausam

Attendees

  • Ted Klein
  • Carol Macumber
  • Susan Barber
  • Carmela Couderc
  • Dr. James Case
  • Dr. Robert McClure
  • Russ Hamm
  • Peter Jordan
  • R Daniels

Value Set Expansion Discussion

  • Lively discussion and background on VSE project indicating that there is a need to define parameters of a value set definition that effect the expansion that are not in the definition. For display strings, sometimes multiple different display strings are allowed. FHIR does not specify or differentiate what display should be returned in an expansion. This is left up to the implementer, and it should probably remain this way.

-List additional capabilities needed to control the value set expansion

-Expansion profile is a constraint on an existing value set

-No motions or action items.

Tuesday, September 12, 2017

Name / Organization / Q1 / Q2 / Q3 / Q4
Jim Case / SNOMED International / X / X / X
Peter Jordan / HL7 NZ / X
Rob McClure / MD Partners / X / X / X
Susan Barber / TN Dept of Health / X / X
Carol Macumber / Apelon / X / X / X
Carmela Couderc / IMO / X / X / X
Reuben McDaniels / HL7 AU / X / X
OyvindAassve / Dir eHealth Norway / X
Joel Schneider / NMDP / X / X / X
Ted Klein / KCILLC / X / X / X
Rob Hausam / Hausam Consulting / X / X / X
Russ Hamm / IMO / X / X / X
Rick Geiner / Lantana / X
Shelley Spiro / Pharmacy HIT collaborative / X
Bill Omerocl / Cerner Corp. / X
May Terry / Flatiron Health / X / X
Marty Madera / AJCC / X
Laura Vega / AJCC / X / X
Matt Rahn / ONC / X
JoginderMadra / Madra Consulting / X
Swapna Abhyankar / RegenstriefInsitute / X
Matt Elrod / APTA / X
Lisa Nelson / LOTS LLC / X
Steve Hufnagel / CIMI / X / X
Genny Luensman / CDC/NIOSH / X
Lori Reed-Fourquet / eHealthSign / X
Andrew Statler / Cerner / X
GeorgioCangioli / HL7 Italy / X
Abdul-Malik Shakir / HI3 / X
Yvonne Cole / IPO / X
Russell Ott / Deloitte / X
Brian Reinhold / LNI/PCHA / X
Calvin Beebe / Mayo Clinic / X
Benjamin Flessner / Redox / X
Gay Dolin / IMO / X
Michael Clifton / Epic / X
Steve Fine / Cerner / X
Ray Mehra / Cerner / X
Julie James / Blue Wave / X
Yunwei Wang / IMO / X
Grahame Grieve / HL7 / X

Q1 – FHIR Terminology Items; Chair: Ted, Scribe: Jim

Terminology services capability statement

The existing capability statement is located at:

It is currently way out of date. There are no trackers for the terminology capability statement. There was a meeting earlier this year that this is needed and we need to describe the way forward to make this happen. The process may be the workgroup makes a proposal to the FHIR resource process and try to progress it.

ACTION: Create the initial proposal and when draft is ready an agenda item will be added to the conference call agenda(Rob/Reuben/Peter). Goal is to have this ready by the next vocabulary call.

Preparation for discussion on "Minimum requirements for a terminology services" (Wed Q1)

The questions are 1) what are the minimum requirements for someone to call themselves a terminology service and 2) what do we expect from a proposed resource or operation to move from one maturity level to the next?

What does it mean to be “Adequately exercised”? How is it documented?

Current capabilities are located at:

Many of the tests that have been identified have not been created due to the labor involved. Agrred that it must be automated. Question is whether there are tests that can verify whether implementations conform to the current minimum statements? Is the goal here to determine what the minimum requirements are or how to test whether implementations actually conform to the statement? There are three aspects:

-Functioning terminology service delivered by an implementer

-Description of the capabilities that a terminology service must meet to be called a FHOR terminology service

-How to determine whether the resources that use terminology services move from one capability maturity level to the next

It is unclear whether these are documented adequately.

Mapping other sources into FHIR resources (VSAC, MIF)

Ongoing discussion between Ted and Grahame to be able to represent all of the MIF functions. It has been demonstrated in the tooling proof of concept. Now working with Rob M. to represent the VSAC requirements. It will be reviewed by Rob M. to make sure all of the essential requirements have been met. There are still identifier issues to be worked out.

Version 2 code tables have multiple identifier representations that need to be resolved. It is unknown how SNOMED and LOINC are represented in the FHIR resources. Need to determine what content needs to be populated in the terminology server (i.e. all of the identifier types in SNOMED CT, LOINC, etc.).

Discussion ensued about the use cases that need to be supported around maintenance of both code systems and value sets. There is a need to have a better understanding of the code systems that are needed to be represented within the terminology server.

Q: Will someone working with UTG be creating value sets? If so, how will the source code systems be represented? A: For SNOMED, HL7 has a license for value set creation (development), but it is not allowed to distribute an alternative distribution format from RF2. There is no violation of the license if the alternative format is not used to distribute SNOMED CT content. The question remains as to how would SNOMED and code systems be represented in UTG?

If we are going to be creating value sets using UTG, the entire code system needs to be represented as a source for the value sets. The approach as to how to do this has not yet been determined. There are architectural decisions that still need to be made.

Do we want to make a list of requirements for UTG as a way to create value sets? This will be discussed Q3.

The simplest thing UTG can do is represent internal HL7 terminology artifacts. Harmonization has been primarily focused on that. Should that be the primary focus of UTG before trying to get down into the details of external code systems? Ted verified that this is the approach that is being taken. The issue around external content only arose a couple of meetings ago.

The issue revolves around VSAC content that includes both HL7 content (CCDA value sets) as well as other value sets derived from external vocabularies. So the question is whether the MIF and VSAC sources can be represented as FHIR resources? Ted stated that this is being looked at and has been completed for MIF, but not for VSAC.

ACTION: Rob M. will meet with Grahame to document the mapping between the proposed tooling and the VSAC.

Question to vocab is what aspects of external code systems are needed to be represented in the FHIR resource to create and maintain value sets? Ted discussed the incremental approach that has been taken to extend the resources to support the requirements need to maintain a specific value set from a specific code system. Carol asked where the set of requirements were that are needed to be supported? The requirements document are located on the Vocabulary wiki under the UTG project.

Discussion on how the FHIR CodeSystem resource is used and a number of specific use cases were described. Rob M. asked whether all of the aspects of a particular code systems is represented in FHIR to enable the creation and maintenance of values sets.

The demonstration of the capabilities is in the proof of concept projecton github.

FHIR Gforge tracker items

Postponed due to lack of time in the quarter.

Q2 – Joint with Structure Documents: Chair: Rob McClure; Scribe: Rob Hausam

Q3 – UTG: Chair: Russ; Scribe: Jim

Goals for session:

-Review of current state (20 minutes)

-Next steps (5 minutes)

-Review outstanding issues

Review of current state

Ted presented the history of the project. A proof of concept set of tooling was developed by Ted and Grahame using FHIR tooling and Jira. The project is on github at the address:

Ted has been talking with stewards for a number of terminology sources including CDA, CIMI, VSAC, MIF, V2 and V3. Not all have been implemented in the proof tooling. Some extensions to exsiting resources were needed to implement V2. Concept domains for V3 also required some changes to the resources.

The current proof of concept includes content and tooling. The tooling is on the FHIR Tools Release page. Ted did a brief demo of the tooling. The first tool was the value set editor. A mechanism for browsing the content needs to be discussed. Also included in the prototype is a code system editor.

The workflow based on Jira was also demonstrated. It is meant to follow the UTG requriements document. Q: Is it possible for Vocab members to review and comment on these? A; Yes. Requires a Jira login, which can be gotten from HQ.

Q: Are the business requirements for this defined? A: Yes, on the wiki page for the UTG project on a shared dropbox folder.

Next steps

Got approval from the FTSD to move forward on developing a full prototype that will lead to a production release. The beta is intended to be available for review by New Orleans. This will also involve moving the tooling and production environment to an HL7 managed environment.

Q: Is the plan to represent all code systems from all products to be represented in a single terminology model? A: Currently represented as FHIR resources. There are a number of requirements that involve FHIR resources plus extensions that are product family based.

Q: What is the practical use case for using these tools to create value sets to meet a particular project? A: This was initially developed to replace the harmonization process and to get all of the terminology artifacts into a common repository. This is essential due to the new requirement to add v2 terminology requests to the harmonization process. Structured documents has been encouraged to participate in a harmonization process as well. Streamlining the process and the management of value sets is the primary goal of this project. Also essential because the existing tooling is no longer maintainable. These tools will assist vocabulary facilitators in meeting the needs of their WGs.