Talk:Mobile access to Health Documents - Supplement revision 2

October 24, 2014

attendees:

• John Moehrke
• Bill Majurski
• Justin Fyfe
• Karen Witting
• Dragon
• Sungkee Lee
• Walco van Loon
• Ben Kraufmann
• Lynn Felhofer
• Alex Lippitt
• Dave Franken

September 12, 2014

attendees:

• John Moehrke
• Bill Majurski
• Justin Fyfe
• Karen Witting
• Dragon
• Sungkee Lee
• Walco van Loon
• Ben Kraufmann
• Lynn Felhofer

Notes

• Justin
  • will be able to provide a few hours a week under a new contract he has with OpenHIE
  • Will be at HL7 FHIR Hackathon
• John has gotten no progress
  • Will be at HL7 FHIR Hackathon
• Bill
  • has been working on a few of the key difficult attributes and how to translate (e.g. URI)
  • Working on the example. Need to validate that it fits spec, and that spec fits the example
    • We can rely on the HL7 FHIR code-base to do the XML to JSON format translations
    • concept of "code first". Try to have a reference system and test tooling, that will help drive spec accuracy
    • cleaning up the wiki page for attribute mapping
  • Will not be at Hackathon, but will be at HL7 meeting
• Walco
  • has been working on the Provide transaction
  • Not liking the Mailbox model, likely will end up using it
  • Concern around FHIR principle of everything being a referencable
    • Discussion: We can and should use the 'contained' concept of FHIR when we really need it (e.g. PatientSourceInfo)
  • Mailbox issues
    • it is a very underspecified thing
    • therefore conformance is really only testable with a profile
    • note we can add transactional behaviors to core mailbox behavior
    • What is the difference between Transaction and Mailbox?
      • Walco: Message exchange pattern (messaging, REST, mailbox)
    • John: I think there is a concept of "Transaction" that is being developed. It is not well exposed to the broader community.
    • John: Lets just focus on Mailbox only as a way to allow us to focus on the bundle content and encoding inside the transaction. We will likely use a totally newly invented URL for our Provide endpoint. It will have as many characteristics from FHIR as we can take, but will not likely be a FHIR endpoint.
  • Will work on the Provide trasaction. Focus first on the bundle content profiling. secondary on HTTP transactional characteristics.
• Gila has forwarded the shortened MHD supplement to Mary for publication on the external www.ihe.net web site.

February 28, 2014

attendees:

• John Moehrke
• Bill Majurski
• Eric Heflin
• Mauro Zanardini
• Rob Horn
• Gila Pyke

Topics:

• Reveal Wiki page and work through operating mechanism
• Attribute Mapping - Bill
  • Author cardionality mismatch
    • might be deeper solution
  • architecture difference
  • Bill to post link to google document he is using for tracking
  • analysis doing broad assessment first, starting with XDS focus
• Scope - John
  • have not done
• Provenance Resource - Rob
  • has looked at W3C provenance.
    • Primer http://www.w3.org/TR/prov-primer/
    • Model: http://www.w3.org/TR/prov-dm/
    • this looks usable by MHD
    • They use this for associations and lifecycle
    • Multi-Node edges is a concept in W3C provenance that could be useful in medicine
    • Not clear this is the same as FHIR
• Transport Pattern -- Walco
  • Noted FHIR has concept of many transports, but HTTP REST is only one clear
  • looking into other
  • Will focus on Publishing including multiple documents and submission sets.
  • should also look into defining this simply as a service that invokes FHIR resources
  • Note that a "Profiling" tool is being developed. Not clear if it is accepted inside of HL7 FHIR yet.
• FHIR vs IHE lifecycle
  • See W3C discussion

October 10, 2014

Attendee: John Moehrke, Bill Majurski, Elliot Silver, Jim McInnis, Karen Witting, Laura Bright, Lynn Felhofer, Peter Bernhardt, Walco van Loon, Derrick Evans,

Topics:

• FHIR DSTU2 is going to be delayed. So we should continue to focus on using DSTU1
  • This will allow IHE-HL7 workgroup to be created and better support us inside HL7
  • This will allow us to focus on DSTU1, while evolving DSTU2
  • We should focus on IHE profiling, thus not trying to us the FHIR Profile or Conformance resources
  • We will need to use FHIR extensions as needed for our work.

• Connectathon
  • Minimal support will be available for "New Directions" (aka Hackathon)
  • Bill is looking for a java implementation that he can standup for connectaton use
    • Try Josh Mandel / SMART platforms -- http://api.fhir.me

• Provide and Register
  • Walco has provided an example published through a google document link
  • We should spend more time on the bundle content definition
  • Given the current direction we should focus on creating a new service. This will not be usable with existing FHIR servers.
  • We should ask the FHIR core team to review our proposed service
  • Could leverage MHD original Provide transaction