Web Meeting Info:
Join Zoom Meeting - https://zoom.us/j/7183806281?pwd=WHVnUUlkWWhhcnRaYk9sWWQyOEkvUT09 | Meeting ID: 718 380 6281 P: 370553
+1 646 558 8656 US (New York) | Find your local number: https://zoom.us/u/aciVC9RrJ6
- HTA and URLs for external code systems
|5 min||Administriva for this call||All|
|5 min||CTS2 Withdrawal||All|
NIB deadline is July 5. Carmela will either find the NIB or create one, send the link to Rob M. who will send the approval email to the Infrastructure Steering Committee listserv.
|40 min||HTA and external code system URLs||Rob M|
URLs may change for code systems and for value sets. HTA has been chartered to manage/address URL assignment.
The implementer community is rejecting the HTA process which may result in a change to a canonical url that has been used. The HTA process acknowledges the existence of a URL that may have been adopted by the implementer community that isn't necessarily what the code system owner/steward would define.
Part of the issue might be that there is not a technical solution to the issue of a canonical URL that changes.
RD: not all parties are in agreement on how the system works, are people ignoring the HTA policy, or has it not been socialized sufficiently?
The is no documentation on how the system works, so when there is disagreement, there is a Zulip chat that sometimes doesn't go anywhere. Policies are not documented, and are not signed off. Governing by Zulip has not been shown to be effective.
RH: Are there some edge cases that aren't clear?
RD: Do we accept the wishes of IP owners or not?
RH: What is the IP we are talking about? Is the url really part of the IP for a code system?
RD: Would SNOMED accept if implementers started using a different canonical URL?
Does it make a difference if we have reached out to the code system owner, or did they reach out to HL7? Have we been clear that one has been either suggested or is in use? (Yes)
CM: We need to draw a line between HTA and vocab. HTA does the outreach. Vocab owns the resources. It is vocab's job to decide. If implementers decide not to use the canonical URL, that is on the implementer. If FHIR decides not to implement tooling to support, that is FHIR's decision. (Note: not all implementers have rejected/raised alarm when a code system steward defined a canonical URL)
RM: the domain of the canonical URL hasn't been discussed. The domain chosen is owned by NUCC (provider taxonomy example). We could define a rule that within HL7, in defining a canonical URL for a code system, we cannot use another organization's domain in that canonical URL without getting their permission. This was done with good intent, Rob proposes that any URL defined by HL7 for an external code system, (because people can't wait), must use the domain = terminology.hl7.org. OR define a URN (which communicates that the value might change).
When a code system owner communicates they would prefer another URL, HL7 needs to support that request.
RM: We should distinguish between URLs are resolvable vs: those that are not. Should we define: the value in this element always goes to a page (with documentation for example), and the value in this element resolves to a resource - source of truth for a code system. We haven't built the resources to address a federated terminology system.
RM: Identifiers matter. Multiple unique identifiers?
Continue next week at the co-chairs. Carmela to create a Confluence page to start documenting.
|5 min||R4a vs R5||All|
Not discussed at the 6/20/2020 call.
We need a more formal statement around the plans for R5. Informally it has been left on Zulip as (continue on R5, 4b will be put out in between).
R4a - Official release of R4, execept a few resource additions
R4b - Changes that mean it is not R4 compatable, but the precursor to R5
Two resources, at least, are driving a R4b release (Medication Knowledge Resource, Subscription (notification when something changes))