Skip to end of metadata
Go to start of metadata


Chair:  @Kathleen Connor

Scribe: @Joel Bales


Attendees


Name

Affiliation


Present

Name

Affiliation

John Moehrke , Co-Chair

By-Light


Alexander Mense, Co-Chair

HL7 Austria

Kathleen Connor , Co-Chair

VA (Book Zurman)

@Trish Williams, Co-ChairFlinders University

@Christopher Shawn, Co-ChairVA

@Mike DavisVA

@David Staggs


VA (Electrosoft)

SRS

@Francisco JaureguiVA (Electrosoft)

@Joe LamyAegis



@Matt BlackmonSequoia

HL7 FHIR

@Dave SilverVA (Electrosoft)

Kaiser

VA (Book Zurman)

@Ken SalyardsSAMHSA

@Peter VanLiesdonkPhillips

@Adam Wong adam.wong@hhs.govHHS

@Mohammad Jafari mohammad.jafari@bookzurman.comVA (Book Zurman)

@Ricky Sahu, @1up.health

1up Health

HL7
x
x


Craig NewmanAltarum
xJoel BalesDoD/VA Interagency Program Office



xCarlos PolkDoD/VA IPO
xGary DickinsonCentriHealth/UnitedHealth Group

Serafina Versaggi




Michael DonnellyEpic
xDidi DavisSequoia

Matthew Rahn




Agenda Topics

Agenda Item

Discussion

Decision Link(if not child)

Welcome

Recoding URL: https://fccdl.in/ChzNoDBehN

Basic Provenance Draft

Walk through guide and conformance statements:

Section 3.1 – basic provenance practices. Review of conformance statements. Key items of interest:

  1. The author organization shall be the health system or clinic responsible for the care when the author creates their update to data. Key point: organization where the clinician is working is the author organization.
  2. Data transformed through intermediaries . . . with no changes to content shall retain the author, author organization, author time stamp of the source system. Discussion: Transformations introduce risk, which must be addressed. Example: v2 lab messages from ELR ingested to EHR, then published as a CDA. Source system provenance, transformation provenance, and original information should be retained for end-to-end fidelity. At the previous WGM, Epic and Cerner interoperability architects agreed that capturing and tracking a full chain of provenance exceeded basic provenance requirements. Chain of provenance should be able to be constructed in a forensic manner, if needed, as long as provenance is retained for the prior hop. Each hop involves an algorithmic transformation, and the information must go through a validation phase.

Gary Dickinson, Brett Marquard and the group discussed the level of detail to include in basic provenance and future roadmaps for advanced provenance:

Carequality exchanges 36 million CDAs monthly without including the view seen by clinicians prior to the information being transformed into a CDA document. To maintain quality assurance, Carequality is updating implementation guides to perform monthly ad hoc testing. Must identify minimum requirements for vendors to implement within the next 18 months, and then identify future improvements.

Clinicians require sufficient provenance for clinical decisions (one layer back). Data must carry provenance information useful for the downstream system.

This Basic Provenance project, focused on clinician needs, is scoped similarly to existing systems (DoD-VA Joint Legacy Viewer, Epic Care Everywhere platform, Cerner) which contain time stamp, and author organization. Author may or may not be displayed.

The group discussed the importance of scoping Basic Provenance to move industry forward while setting a roadmap for future enhancements. The guide references previous provenance work in the appendix, which can feed into the roadmap for future industry directions. For example, consider capturing additional data if available in a source system, and clarify that the three basic provenance items are a start, and more may be required in the future. Basic provenance is built on previous work, which helps guide implementation. It defines a minimum testable set of data, but does not limit systems from including more information in their provenance (building on previous work listed in appendices).

Brett and Russ Ott would appreciate assistance for advanced provenance language, because they did not lead previous provenance efforts. Gary Dickinson can assist. Provenance is expected to have additional future requirements. Brett will add a section to the appendix listing use cases (e.g., legal discover) and references to address those use cases (he requested Gary’s assistance).

Brett will add language in the introduction to clarify that Basic Provenance is a first step to solving provenance, but it not solved. He will add disclaimers that this is basic provenance is a minimum, and that more robust data will make provenance stronger. Requested that Gary provide information on which artifacts to include. Need help to add the language for the appendix. These changes can be addressed through ballot comments.

Will present to Security tomorrow to request approval to go to ballot.


Argonaut Provenance Profile

Enhance existing invariant to apply to device also

Updates to Argonaut guide:

Draft guide:

  1. No new CDA templates in the draft guide. The CDA we helped design is in the companion guide that went to ballot in June, and is being worked to reconciliation. No comments specific to the template.
  2. Profile for FHIR: this group gave input and reviewed work by the Argonaut community. We will continue to provide feedback and not create a competing companion guide. Argonaut will propose to U.S. Core. Guidance here will complement that new profile.

Next Agenda

Next meeting at the WGM in Atlanta


Call ended3:45 PM eastern time
  • No labels