Stream: genomics
Topic: FHIR R4b - Observation-genetics profiles
Joel Schneider (Sep 24 2020 at 20:45):
Per Grahame, the FMG has discussed what can be done with the Observation-genetics profiles in FHIR R4b, and has agreed it would be ok to add a note that refers the reader to our current work. Have we made a JIRA ticket for this yet?
Lloyd McKenzie (Sep 24 2020 at 21:13):
There's already a link. However, you could re-word the language around it. What you can't do is yank the profiles that are in R4. The most R4B can do is say something stronger than R4 does. E.g. "This content is deprecated and SHOULD NOT be used. Implementers are instead directed to the [Genomic Reporting IG] for guidance..." or something like that.
Joel Schneider (Sep 24 2020 at 21:32):
There's a link and note in the Observation-genetics landing page.
http://hl7.org/fhir/observation-genetic.html
However, there's no link or note in the individual profile extension pages. It may be best to repeat the note within each Observation-genetics profile extension ...
http://hl7.org/fhir/extension-observation-geneticsallele.html
http://hl7.org/fhir/extension-observation-geneticsaminoacidchange.html
http://hl7.org/fhir/extension-observation-geneticsancestry.html
http://hl7.org/fhir/extension-observation-geneticscopynumberevent.html
http://hl7.org/fhir/extension-observation-geneticsdnaregionname.html
http://hl7.org/fhir/extension-observation-geneticsgene.html
http://hl7.org/fhir/extension-observation-geneticsgenomicsourceclass.html
http://hl7.org/fhir/extension-observation-geneticsinterpretation.html
http://hl7.org/fhir/extension-observation-geneticsphaseset.html
http://hl7.org/fhir/extension-observation-geneticsvariant.html
Kevin Power (Sep 24 2020 at 22:14):
@Lloyd McKenzie - So we have a few considerations, I want to make sure we are clear on each:
1 - We would like to remove much of the 'guidance' in the Genomics Guidance page. We would keep the page, but greatly limit the content and instead much more clearly point at our IG.
2 - We have numerous examples that we would like to remove. Is that OK, or do we need just the strongly worded note somewhere in the example?
3 - Sounds like the decision is that we cannot remove profiles or extensions, so we would need to develop strong language to include in all of those pages, yes?
Lloyd McKenzie (Sep 24 2020 at 22:20):
You can't make any change to the language of the section without bringing it to ballot again, and you can only bring it to ballot again if there's an urgent regulatory requirement to do so. The most you can do is change the language at the top. You can't add anything, remove anything or change anything.
Lloyd McKenzie (Sep 24 2020 at 22:20):
You can't even change the ballot status of what's there.
Kevin Power (Sep 24 2020 at 22:28):
What about examples? Given that criteria, are we forced to simply add something to the narrative such that something will be rendered to say "DO NOT RELY ON THIS EXAMPLE" ?
Kevin Power (Sep 24 2020 at 23:05):
Can we break out the Dragon: https://www.hl7.org/fhir/search.html#date
(for anyone who hasn't seen it before, the FHIR spec has a few places where something is really important, and a fire breathing dragon is used to get your attention -- you might have to scroll down a bit to see it)
:smile:
Lloyd McKenzie (Sep 24 2020 at 23:10):
The dragon might be possible. Can't remove the examples, but should be able to flag them.
Kevin Power (Sep 24 2020 at 23:12):
I will have to think about the examples more - I am unclear how to flag them with something like this.
Kevin Power (Sep 24 2020 at 23:18):
Give the relatively limited nature of what we can do here, what do others think? I'll have to admit, I am not sure there is a ton of value in just adding a stronger message to the artifacts. However, calling out a few that have had stronger opinions about this in the past:
@Bret H @Liz Amos (no Clem on Zulip?) @Patrick Werner @Bob Dolin @May Terry
Of course, welcome input from anyone.
Joel Schneider (Sep 24 2020 at 23:50):
If there is agreement that "fail fast" is the best option for consumers of this deprecated content, I think making the notes/notices as strongly worded and prominent as possible is the best option.
Joel Schneider (Sep 24 2020 at 23:51):
Best if they don't waste time using it ...
Kevin Power (Sep 25 2020 at 12:56):
I would not argue it is the “best option” for this, but for R4B it is the only option.
Kevin Power (Sep 25 2020 at 12:59):
To be clear, we will move forward making updates for R5 as we discussed and voted on at the WGM. This work for R4B is something additional.
Liz Amos (Sep 25 2020 at 13:22):
My vote (for R4B) is to put a stronger warning (I think our Note now is not strong enough. We need lights and sirens). I like Lloyd's suggestion: E.g. "This content is deprecated and SHOULD NOT be used. Implementers are instead directed to the [Genomic Reporting IG] for guidance..." . A dragon would make it even more obvious (and 10x cooler).
Liz Amos (Sep 25 2020 at 13:23):
Sadly, Clem is not on Zulip! He barely checks his email during the day. :) Please @ me and say it's for Clem and I'll let him know.
Joel Schneider (Sep 25 2020 at 15:21):
This topic was briefly discussed during a joint meeting between FHIR-I, O&O, and others on Thursday (yesterday). At that time, Grahame indicated the result of FMG's discussion about this.
The Observation-genetics page and related extension pages are officially owned by the Orders and Observations workgroup. I think changes to those pages (i.e. new or revised notes) will need to be detailed in a JIRA ticket which will be assigned to the O&O workgroup.
Kevin Power (Sep 25 2020 at 15:56):
If someone can log the request to O&O, I would appreciate it. If we need to catalog the specific files that need the stronger warning, someone can review the following pull request where things were pulled out for R5:
https://github.com/HL7/fhir/pull/877
Joel Schneider (Sep 29 2020 at 05:18):
Last updated: Apr 12 2022 at 19:14 UTC
