This wiki has undergone a migration to Confluence found Here

Difference between revisions of "AID 201705 Meeting in Madrid"

From HL7Wiki
Jump to navigation Jump to search
Line 6: Line 6:
==Monday Q3==
==Monday Q3==
*Rene Spronk, chair/scribe
*James Agnew

Revision as of 12:35, 8 May 2017

Return to: WGM Minutes > 2017 > May Madrid

This is the agenda of the AID WG for the WGM in Madrid ES, May 2017.

Monday Q3


  • Rene Spronk, chair/scribe
  • James Agnew


  1. Administrative
  2. Project #1178 - Implementation Packages
    • Rene: the concept has been put forward to Publishing in the past (they like it, but don't feel they can make a decision), and has been informally communicated with the TSC and the CTO. In April 2017 I asked Wayne, the current CTO, for his suggestions as to how we can move this initiative forward. His suggestion is to a) bring the idea to the ArB so they can advice whether or not it is sufficiently fleshed out for further consideration (if not they'll have suggestions), b) go the TSC (they'll ultimately have to decide), and c) ask O&O (as the core WG which will form the HL7 v2 product management group) for comments on the exact content of the HL7v2 implementation package.

Wednesday Q1


  1. Project and issue review
    1. Project #550 - Whitepaper review: project details, List of AID Whitepapers
    2. Issue List Category:AID_Issue

Sunday Q3/Q4 (joint with FHIR)


  • Rene Spronk, co-chair & scribe
  • 38 other particpants, see scanned attendee list below

20170507 AID attendees1.jpg 20170507 AID attendees2.jpg


  1. Rene calls to order at 14:45
  2. FHIR and legacy databases
    1. Architecting a FHIR frontend using a legacy database backend, (Rene Spronk, Ringholm, NL)
      • High level introductory discussion of implementation issues
    2. Building FHIR Servers on existing Relational Databases (Robert Worden, , UK)
      • To make a FHIR Server from an existing healthcare relational database, you have to develop transforms between FHIR resources and the native database formats - in both directions (for GET and PUT). One way to do this is to use the FHIR Transform Engine. This gives bi-directional transforms from declarative mappings - saving coding effort, and increasing reliability. Transforms integrate with the HAPI server engine. An example implementation is described.
      • There is an important side-benefit to this approach. RESTFul FHIR searches are converted to efficient SQL queries, automatically - supporting a wide range of searches without coding.
    3. Building a FHIR facade for an existing GP system (Christiaan Knaap, Furore, NL)
      • We have used Vonk to implement a FHIR façade to an existing GP system to open up the data to patients. We had to bridge the gap between the system’s database schema and FHIR resources, map search parameters and apply authorization based on the OAuth2 token for the patient. I will discuss the setup of the solution, what Vonk offers out of the box and what had to be coded specifically for this use case, along with difficulties we encountered.
  3. Other FHIR related presentations
    1. Implementing databases for FHIR - 3 years experience report (Nikolai Ryzhikov, HealthSamurai, US/RU, see for slide deck)
      • With the upcoming SQL standard 2017 with JSON - the FHIRBase approach will be suitable for Oracle, MSSQL and other compliant databases.
    2. FHIR Analytics using Apache Drill (Chris Grenz, ,US, see for slides)
      • Agenda: A Clinical Analytics Use Case; Using Profiles for semantic definition; SQL for FHIR Data; Apache Drill case study
    3. FHIRForms: generating clinical forms from FHIR profile definitions (Robert Worden, , UK)
      • Generate a form design from any FHIR profile or bundle; refine the design for specific use cases
      • Form can be used to view or edit FHIR resources and bundles; or for clinical validation of profiles
    4. Dynamically Configurable Scalable FHIR Repository (Tuncay Namli, SRDC onFHIR, Turkey)
  4. Adjournement at 16:45