Stream: australia
Topic: HL7 Australia Publication Proposal
Grahame Grieve (Jun 08 2016 at 01:25):
@Brett Esler @Andy Bond @Angus Millar I'd like to discuss HL7 Australia FHIR development and publication policy
Grahame Grieve (Jun 08 2016 at 01:26):
As for today, we now have tooling for publishing Implementation Guides that no longer requires the content to be part of the FHIR subversion, or to be published as part of the HL7 FHIR publication
Grahame Grieve (Jun 08 2016 at 01:26):
And we have content that NEHTA is working on that is heading into pilot implementation. So it needs a new home.
Grahame Grieve (Jun 08 2016 at 01:29):
I'd like to propose that HL7 Australia as an organization:
- creates a hl7au account on GitHub
- creates a project "fhir-impl" in there for fhir implementation guides
- creates a space under hl7.org.au/fhir for publishing the content found in the fhir-impl project
- agrees that the content NEHTA is working with RCPA on will be moved to this new home (still licensed under creative commons public domain), worked up by NEHTA/ADHA and RCPA together, and that there will be a migration process towards being managed by HL7 Au in association with all the stakeholders
Grahame Grieve (Jun 08 2016 at 01:30):
fyi @Tim Blake @Reuben Daniels @Michael Lawley @David Hay as well
Grahame Grieve (Jun 10 2016 at 10:04):
ok, the master for the RCPA implementation guides has now been moved to the HL7 Au github repository at https://github.com/hl7au/fhir-impl
Grahame Grieve (Jun 10 2016 at 10:04):
in rcpa
Grahame Grieve (Jun 10 2016 at 10:04):
@David McKillop all yours to commit the rest now
Brett Esler (Jun 10 2016 at 10:08):
@David McKillop anybody that needs push access let me know your info
David McKillop (Jun 11 2016 at 10:10):
@Grahame Grieve thanks for that.
OK @Brett Esler I'll be in touch.
Stephen Royce (Jun 12 2016 at 23:43):
@Brett Esler What details do you need to give us push access?
Stephen Royce (Jun 12 2016 at 23:45):
@Grahame Grieve I assume you have the HL7 Australia's Board approval for this?!? :stuck_out_tongue:
Grahame Grieve (Jun 13 2016 at 00:35):
Brett is a board member, and he invited me.
Brett Esler (Jun 13 2016 at 02:20):
Have had general approval before; will need to chat at next board meeting about it; but think should push to HL7 AU standards TSC on how to manage... I have some technical hurdles with the hl7.org.au web hosting as we apparently don't have full access to root folder under current setup... am following up on getting a static folder put in with access
Brett Esler (Jun 13 2016 at 02:20):
@Stephen Royce github account name or email
Stephen Royce (Jun 13 2016 at 02:27):
So I was not really being serious, but, if you want to be, I have heard board members and TSC members express an opinion that the "general approval before" may have, or indeed has, lapsed. 
Stephen Royce (Jun 13 2016 at 02:35):
My GitHub username is sroyce (stephen.royce@nehta.gov.au).
Brett Esler (Jun 13 2016 at 02:36):
@Stephen Royce invitiation in the mail
Stephen Royce (Jun 13 2016 at 02:37):
@Grahame Grieve Incidentally, what's the timeline for the logical model => profile transformation engine?
Stephen Royce (Jun 13 2016 at 02:37):
Suite! Got it, thanks, @Brett Esler . :smiley:
Grahame Grieve (Jun 13 2016 at 03:28):
draft in the next few weeks
Stephen Royce (Jun 13 2016 at 04:53):
Michael Lawley (Jul 01 2016 at 23:16):
Wow - lots of questions here:
1. What's with all the code systems?
2. Is this legacy stuff?
3. Are there plans to also publish ConceptMaps that link to standard CodeSystems?
4. The CodeSystem valueSet URIs look strange - CodeSystem is "RCPA" http://hl7.org.au/fhir/rcpa/CodeSystem/stain-type but the corresponding ValueSet is (no longer existing) "NEHTA" http://hl7.org/fhir/ValueSet/nehta-stain-type
Grahame Grieve (Jul 01 2016 at 23:55):
ok. see a draft here: http://fhirup.com.au/fhir/rcpa/index.html
Grahame Grieve (Jul 01 2016 at 23:55):
this will become http://fhir.hl7.org.au/fhir/rcpa/index.html, but the name delegation isn't done yet
Grahame Grieve (Jul 01 2016 at 23:57):
to Michael's questions: This is new stuff, but a representation of a joint CAP/RCPA project that has described how to do reporting properly. They list the dsplays that you can pick from, but did not concern themselves with terminology (it's just cap, after all, what do they know about tx?)
Grahame Grieve (Jul 01 2016 at 23:58):
we have to pick code systems; if you think there are code systems David could have picked rather than make up new ones, please let us know. Otherwise, our working process is to define the code systems ourselves and then request the concepts to be added to SCT or LOINC, which we will then adopt later.
Grahame Grieve (Jul 01 2016 at 23:59):
we won't be publishing ConceptMaps - if we could map, why not just use the base codes (only if licensing issues prevent, but then we probably couldn't publish maps either)
Grahame Grieve (Jul 02 2016 at 00:00):
this is still coming together, so I'm not surprised if there's broken links - I changed the canonical base, but didn't hunt everything down
Michael Lawley (Jul 02 2016 at 00:45):
[note the "no longer existing" comment was about NEHTA becoming the Agency, not about the link being resolvable :)]
Grahame Grieve (Jul 02 2016 at 00:46):
well, we can remove references to NEHTA now that we're publishing through HL7 Australia with appropriate agreements. ADHA happens to be providing editorial support now, rather than publishing outright
Michael Osborne (Jul 02 2016 at 02:50):
Further to @Michael Lawley 's comments, the code systems in the xml files on GitHub look to be a mixture of local codes, e.g. Tumour Stage (codesystem-crc-tumour-stage-n.xml) has nx,n1a and others like codesystem-stain-type.xml have more meaningful codes like "normal_staining". Is there going to be some structure to the codesets? I'd rather see ADHA NCTS develop new SCTAU content for these - at least they will be queryable and managed in the one terminology.
Michael Osborne (Jul 02 2016 at 02:55):
In addition - there is work going on internationally with University Medical Center Nebraska and the iPALM SIG at IHTSDO creating SCT concepts for many CAP cancer protocols and the terminology is modelled using the new observables model. Jim Campbell and Scott Campbell from UNMC are also collaborating with the RCPA/RCP and ICCR via David Ellis and the working group on cancer reporting at RCPA. I would be keen to do this once and not multiple times.
Michael Lawley (Jul 02 2016 at 03:50):
+1
Brett Esler (Jul 02 2016 at 11:15):
ETA on http://fhir.hl7.org.au is on Monday sometime; it is expected in the longer term plan to also publish the same under http://hl7.org.au/fhir
Brett Esler (Jul 02 2016 at 11:59):
also draft of smart on fhir AU profile is now at http://fhirup.com.au/smart-on-fhir based on http://docs.smarthealthit.org/ and discussions documented at http://confluence.hl7australia.com/display/PA/Current+FHIR+Profile+Drafts
Brett Esler (Jul 02 2016 at 12:09):
repositories for /fhir/rcpa and /smart-on-fhir can be fouind on https://github.com/hl7au
Grahame Grieve (Jul 02 2016 at 21:22):
with regard to the terminology work, the project is being done in association with David Ellis. I don't think that anything said invalidates our approach: define local code systems until the content is properly defined elsewhere. I cannot comment on the process for defining the content properly. Perhaps someone from ADHA can comment further on that
Grahame Grieve (Jul 02 2016 at 21:23):
Brett - I don't think it makes sense to publish the same content at both fhir.hl7.org.au and hl7.org.au/fhir. Particularly if it's not the same actual server. We should pick just one.
Michael Lawley (Jul 02 2016 at 22:48):
I would venture that the approach skews towards unnecessary proliferation of new code systems because it removes any incentive to identify and reuse appropriate existing content
Grahame Grieve (Jul 02 2016 at 22:52):
so what would you suggest? that we don't do anything until SCT processes come to frution in some distant future?
David McKillop (Jul 04 2016 at 12:39):
Apologies for coming to this late - I was a bit side tracked today and didn't get to Zulip.
As Grahame has said this project is driven by David Ellis who is the President of the ICCR (http://www.iccr-cancer.org/) and Chair of the RCPA Structured Cancer Protocol work (https://www.rcpa.edu.au/Library/Practising-Pathology/Structured-Pathology-Reporting-of-Cancer/Cancer-Protocols). The development of the Colorectal and Prostate (radical prostatectomy) are trial developments using FHIR infrastrucutre for the PITUS 15-16 project (working group 5).
These FHIR products will be used with selected private and public pathology laboratories to a selected cancer registry, with the aim to see if :
a) the labs can produce something similar
b) the cancer registries can consume it
c) what are the gaps and there will be quite a few and what is the best way forward.
The College of American Pathologists (CAP) used their own codes, but we've tried using LOINC and SCT where codes are available and because the content of each cancer protocol is different ie the Pathologist say that the cancer staging etc is different from one cancer protocol to another and therefore there is very little commonality between protocols. Basically all questions (usually LOINC) and answers (usually SCT) are protocol specific and hence there are lots of gaps in the current available terminologies for each specific cancer protocol.
There are NEHTA/ADHA terminology specialists involved in mapping the answers eg procedure type, procedure site etc, for the colorectal and prostate cancers to SCT and there will be a number of codes available following the next SCT-AU release.
The made up codes are there just to provide content and it is envisaged that they will be replaced with appropriate SCT and LOINC codes in due course.
The ADHA/Agency (apparently we aren't allowed to use ADHA as an acronym as it is trademarked somewhere else) terminology staff know the Nebraska team and communicate regularly with them via an IHTSDO teleconference, hence I'm sure they'll be collaborating appropriately on this. I understand the Nebraska team would like to see the cancer specific terminology elevated to the international SCT edition rather than just SCT-AU. However, I'll leave that with the Agency terminology people to progress/discuss.
Yes - I agree the output to date is developing and it will be some time before the products and process will be fully defined, but progress is being made and we'll know a lot more following the trials with the selected pathology providers and the selected cancer registry.
I'm happy to hear any feedback as we are all looking to making the structured cancer protocols more usable.
Brett Esler (Jul 05 2016 at 00:11):
@Grahame Grieve better to go for hl7.org.au/fhir yes? fhir.hl7.org.au (actually fhir.hl7.org.au/fhir) was to be temporary and they would be the same server anyway. I am going to need to get a HL7 AU board approval and plan for changing the hl7.org.au hosting done will talk to them tomorrow. Not sure the turn around time on hl7.org.au/fhir availability; am chasing fhir.hl7.org.au for today....
Grahame Grieve (Jul 05 2016 at 00:20):
well, either not both. I guess I prefer hl7.org.au/fhir
Grahame Grieve (Jul 05 2016 at 00:20):
hosting should not be so hard.... time for a new provider...?
Grahame Grieve (Jul 05 2016 at 01:44):
http://fhir.hl7.org.au/fhir/rcpa/ works for me now
Brett Esler (Jul 05 2016 at 01:46):
yes new provider is the plan... will be fhir.hl7.org.au until sorted then hl7.org.au
Last updated: Apr 12 2022 at 19:14 UTC
