|10:10||HL7 Unified Terminology Project||Ted|
- This project began 3 years ago and an initial architecture has been developed. Proof of concept has been based upon FHIR tooling and JIRA. Trial prototype bagan last October. The objective is to have a continuously updatable, crowd source process for terminology informaiton information and governance. Processes include:
- submission of a proosal for change. the HL7 community provides consensus review and then the content moves to an internal process for internal implementation and availability.
- FHIR Based Tools - content includes: V2, V3 coremif, CCDA (VSAC) used for proof of ocncept. Extensions were required to resources of model V3 and V2 to support governance.
- Tools to operate on the FHIR Base include - terminology server instance (or you can use an external terminology service), Edit/Expand Value Sets, Edit Code Systems. Each of these is downloadable and open, they are prototype tools but the are operational.
- Current State - calls ever fortnight. the workflow is documented in JIRA. Workflow includes - Drafting a Proposal, Submission of a Proposal, Review, Oversight Review (manual), Discussion occurs in Confluence, The submission might be
- Approved in which case it goes through the process for inclusion.
- Rejected the submitter is notified (they may submit again).
- Withdrawn for additional work (if the proposer decides to make changes - in which case it reverts to draft).
- finalise persistence model and move to the HL7 server environment - refresh content with current release
- integrate hand-off between JIRA and Confluence for consensus descussions
- Design up-front operations and screens to support them.
- Complete JIRA screen for proposals
- Modify JIRA template for issue tpes to match our operatoins (not sure if this is possible). The Create and Issue process in JIRA is the likely functionality to be used.
Where to find more information: Unified Terminology Governance Project (UTG) Page
Should the system push or pull input on consensus operations.
When will we be operational? Cologne meeting (April 2017), but this is dependent upon may things. there is an itnent to use tracker capacity in JIRA which will hopefully be able to be leveraged. There is a document on how voting could/should work. Oversight groups have been determined. Currently it is not known how you will sign into a notification pool, a similar model to the process for notification of potential changes to a specific terminology object or group of objects. The sign up could be for specific objects or by HL7 list members. There will be no more cycles of harmonisation, they will be ongoing. It is likely that we will be considering whether to go fully live with this around the end of 2018.
Usability issues for general users have not yet been considered. It was suggested that additional requirements can be documented in Confluence and the project will see where and what can be done.
Intention to move ahead to content unification between products. This means a single code system for multiple product family members. There are a broad range of tasks listed in the requirements for things to be changed includes things like making display names consistent.
Access to content - there is a persistent expectation that use of content require license agreement (for VSAC content). There may be issues associated with NLM IP. HTA should consider how to advise our community on the requirements for NLM.
FHIR identifiers of value sets does not work (as evidenced at the meeting Connectathon). This needs to be reviewed.