Stream: Covid-19 Response
Topic: What version of FHIR should be used?
Nick Rupley (Apr 29 2020 at 18:50):
R4 is the latest "stable" version, is it expected that everyone should be using R4?
We could also use R5 Preview #1, but since we are using HAPI FHIR 4.2.0 we're limited to the current snapshot of R5 that HAPI FHIR 4.2.0 supports.
I don't think there's anything specific to R5 that would be needed right? Specifically talking about the Situation Awareness for Novel Epidemic Response track: https://confluence.hl7.org/display/FHIR/2020-05+Situation+Awareness+for+Novel+Epidemic+Response
It doesn't look like there are any differences in the MeasureReport resource between R4 and R5 Preview #1.
Grahame Grieve (Apr 29 2020 at 19:34):
it's all R4 based
Nick Rupley (Apr 29 2020 at 19:51):
Thank you Grahame!
Nick Rupley (Apr 30 2020 at 16:55):
Sorry, another follow up on this. That seems to be at odds with the Confluence page for the Connectathon: https://confluence.hl7.org/display/FHIR/2020-05+Connectathon+24
It says "Read the Connectathon version of the FHIR Specification", and that links to: http://build.fhir.org/index.html
And that appears to be FHIR version 4.2.0 which is R5 Preview #1, right?
Nick Rupley (Apr 30 2020 at 17:11):
I also just saw that there is a separate stream for "connectathon mgmt", is that meant to be used instead for Connectathon questions? In the kickoff webinar this "Covid-19 Response" stream was specifically linked to and the presenters said that this stream should be used for the Connectathon.
Nick Rupley (Apr 30 2020 at 17:32):
Update, I see on the track-specific page: https://confluence.hl7.org/display/FHIR/2020-05+Situation+Awareness+for+Novel+Epidemic+Response
This track will use what version of FHIR
The SANER draft implementation guide is based on FHIR version R4
So looks like for the Connectathon in general it's R5 Preview #1, but for this track specifically we're using R4 instead, since the SANER impl guide references that.
Abbie Watson (Apr 30 2020 at 19:08):
Just as an FYI... standards development proceeds at a different (slower) pace than software development. Unlike open-source or even for-profit software, standards require balloting (voting) and consensus building, rather than just the latest publication of a library.
So, the R5 Preview isn't exactly the development branch or pre-release that you might be accustomed to. Rather, that's where the report-out from after the connectathon lands. It's downstream in the standards development process, and what we are feeding into; not what people should be using during the connectathon itself.
Keep in mind that a lot of the FHIR servers in production at hospitals are still on DSTU2 in the United States, and on STU3 in Europe. For those production systems, the R4 release _is_ the latest version that they're trying to move towards. So stick with R4. The R5 Preview should be considered alpha/experimental, rather than beta/testing ready.
Abbie Watson (Apr 30 2020 at 19:11):
Also, if you're in the US, the R4 version is what was officially adopted by HHS and Medicare. It's the one specified in law, not R5 Preview. Err on the side of compliance rather than latest release or preview. We might be using similar terms, but the R5 Preview means something different in this context than much of the rest of the software world.
Keith Boone (May 01 2020 at 06:03):
R4 is what the SANER Track should be using, since that's what we are building the guide upon, and what we said in the track proposal.
John Moehrke (May 01 2020 at 14:16):
The only reason to use the R5 Preview is because your IG needs the changes that have happened since R4 (turtles...). There are some IG that have told the FHIR authorities that they need a R5 ish release. This is why it is created. Not because they want everyone to us it, they actually prefer no one needs a non formal release; but it is a fact of life.
Last updated: Apr 12 2022 at 19:14 UTC
