Facilitator: Anthony Julian
Note Taker: Ulrike Merrick
|X||Elizabeth Newton||Kaiser Permanente|
|X||Frank Oemig||Deutsche Telekom Healthcare and Security Solutions GmbH, HL7 Germany|
|X||@Kasayuki Dambara||HL7 Japan|
Meeting Minutes from Discussion
|Decision Link(if not child)|
INM meeting minutes 8/29/2019 accepted by general consent
Not discussed - need to move to a call
|Methodology||Discussion with V2 Management|
HL7V2 Quality Criteria:
CWE refers to exemplar code system rather than just value set (since that is what FHIR is doing, even at the base resource)
In v2.9 we have at least a concept domain to bind an element to – give an example of terms for context
Condition Predicate: can we do SHALL there? – for the base standard that may not be possible to completely define C(a/b) – so conformance supports the undefined conditional = in the base the conditional needs to be worded clearly enough to understand the dependencies; IF a condition Predicate is written then it SHALL be computable from the message content
Example: RXA-7 – administered units – it is conditional on the administered amount code (which is RXA-5), but it cannot be easily computable
Looking at CWE:
CWE.3 is required if CWE.1 is valued, and CWE.14 is not valued
CWE.7 is currently only dependent on CWE.3 – need to update to include CWE.14
Similar for code system versions for the other 2 triplets – ACTION ITEM: submit a gForge tracker for V2+ = #24635
Change testable so that a computable statement can be derived from it.
In chapter 2 we have only reference to C(a/b) not to C
Initially C was interpreted only as C(R/X) and CE (not used in the base) was C(RE/X).
We have to separate:
Meaning of C and CE: editors should replace that with the C(a/b) constructs based on the new conformance document – that new document is overwriting the conformance related sections in ALL versions of v2.
There is a discussion of what requires errata vs technical correction – Rob will write up the requirements.
First segment requirement – was introduced in v2.5, so identify that as legacy.
Each segment group must have a unique name within the message structure and SHALL NOT contain any spaces – UPPERCASE_LETTERS_NUMBERS_AND_UNDERSCORES so that you can render valid XML element naming convention (can we add link to the section in the XML specification?)
We are going to adopt the information model for v2 and then the rules derive from that, but we want to spell them out
The usages of a data element must have same name, length, datatype (or one of the allowed flavors) as well as constrained values of the referenced tables
HL70203 is a categorized table – a superset of all the use cases – would be good to add metadata to this table to give guidance on which values to use for what use case in the messages to help folks
For base QA – add requirement of specifying a base version for V2+ will use the year e.g. V2+2021
For IG QA list refer to base list and then only add additional items
Homework to review the QA criteria and provide more feedback
Update the ballot deadline to May 2020 and downgrade to Draft for comment ballot
Keeping this joint quarter for Syndey – yes; it will be Wednesday Q1
|Adjournment||Adjourned at 10:30 AM EDT|
|harmonization proposal||Approve InM-03_HL70203|
Anthony Julian Schedule WGM meeting Wednesday Q1 in Sydney
- Anthony Julian Schedule Conference calls through next WGM - doodle poll to include Alerts crew
- Anthony Julian Put fix for https://confluence.hl7.org/display/INM/Version+2+Quality+Criteria+PSS on the agenda for the first Telcon after the WGM - see 2019-07-29 TSC : Approval items from last week's e-vote referred for discussion:
- Anthony Julian Finish consideration of duration.