Page tree
Skip to end of metadata
Go to start of metadata

Attendees

  • Brain Plew
  • Vitor Pamplona
  • Paul Knowles
  • Rebecca Distler
  • Paul Murdock
  • Nicolas Douliez
  • Drummond Reed
  • Sandeep Jain
  • Wendy Henry
  • Tony Little
  • Zsombor Szabo
  • David Janes
  • Rajesh Pillai
  • Steven Milstein
  • Kaliya Young
  • Brian Martin
  • Jim StClair

Agenda Items

TimeItemWho
2 minWelcome & Antitrust Policy NoticeRebecca & Brian
15 minIntroductionsAll
40 min

Overview of Current Draft & Discussion

Paul Knowles

3 minWrap upChair 

Presentations

Recording

Topic: Good Health Pass - Standard Data Models and Elements
Start Time : Apr 12, 2021 11:59 AM

Meeting Recording:
https://zoom.us/rec/share/dAcJ3yLZFElkChZFp7f8Kz8ZFnOHxUSZWplM3WYtRDyMiHh9hz8w7lCfSzTVpHdk.eTjN_F5kKCvHlrmZ

Notes

1. Welcome and Linux Foundation antitrust policy

2. Introductions

3. Overview of Current Draft & Discussion  

  • Design requirements = 1) Defining profile elements (forms and schemas); 2) Data capture (semantic harmonization - largely been ignored in decentralized space. Within an org, can do transposing, but beyond that into a decentralized space, you want to harmonize the semantics so that data is always machine readable and can contribute to better analytics); 3) Data exchange (common data model)
  • Rebuild semantics → rebuild in data models (schema.org, cdisc models, etc.)
  • FHIR is a complicated model; group working on OCA/FHIR at ToIP
    • Discussion on FHIR as a data model
  • Specific problems to solve:
    • Fake data, disjointed data, siloed data
    • Want authentic, harmonized, and distributed data for governance
    • Want to know how we can use JSON-LD 
  • FHIR is not a data model and can be hefty
    • Consider the OMOP Common Data Model (extensive traction in US research community)
    • FHIR4 not really operational yet
  • Large organizations to “announce support” within 30 days, implementation in 90 days / 180 days
    • Operational/business people to add suggested timelines for adoption (Brian Plew)
  • Need to deliver spec now - a really key piece for businesses to understand effort
    • Recommendations of this group are specific data elements and schemas (need to talk about whether they are JSON-LD or other formats)
    • Test, vax, recovery - diagram uses certificate (need to align with WHO)
  • Minimum viable schema specifications and forms
    • Machine readable spreadsheet - automatically generates JSON-LD code (once we’re happy with spec, can give people code)
    • Paul to turn into an annex?
  • Use "credential" to (1) not overlap with other existing concepts and efforts (2) lights up a number of non-medical-specific trust issuance use cases (3) avoid some negative connotations (although "credential" has some as well)
  • Business / operations perspective: Want to understand what’s coming, data elements, (doesn’t need layers) - overlay with attached elements (PDF/embedded excel - something really tactical that can be handed to engineering)
    • Important to give context on why the fields are there - identity management, DOB
    • Closest to something that can be imported and deployed; upload a schema to schema.org and make suggestions of when to use them (e.g., multiple use cases, direction on why you would choose them)
  • Product perspective, would want a crosswalk to JSON
    • OCA doesn’t care about what common data model is used
    • If we can use FHIR, we can use anything
    • Can we get sample FHIR covid vaccination records?
    • However you format the data - it needs to be consumable by the Rules Engine.
  • Could you have a pass with no PII (if cryptographic link back to source)?
  • Health credentials data standard models and data elements
    • Identity binding data elements - identity binding needs to articulate what they need for that purpose
    • Taxonomy between credentials and passes makes that even clearer - credential contains the full menu of what you may have in terms of health and identity binding data that’s available to produce a pass
    • The ultimate pass is binary - yes / no for whatever context (showing the pass in person)
    • Situations when getting across the border
  • Links between test records - stack records 
  • Take hash of schema base - keeps context
  • FHIR covid vaccination records (looks like W3C records) - https://github.com/digitalbazaar/vaccination-certificate-test-suite/tree/main/certificates
  • https://github.com/digitalbazaar/vaccination-certificate-test-suite/tree/main/certificates
  • This group needs to figure out what is in a testing credential (use documentation from EU)

4. Next Steps

       

Action Items

  1. All to review draft document and provide inputs on sections they could draft
  2. Paul K to connect with Brian M on potential data models to use
  3. Rebecca to look at framing document