Stream: social
Topic: CDA is ... ?
Jens Villadsen (Dec 10 2019 at 19:00):
Whats the take on new projects/greenfield areas involving HL7 standards (I'm trying to start war here)? Does HL7 itself have any recommendations towards the use of either HL7 2.x, HL7 CDA or HL7 FHIR?
David Pyke (Dec 10 2019 at 19:05):
You should ask that in Chat.hl7.org. It's likely to be less bias
Jens Villadsen (Dec 10 2019 at 19:05):
does anyone actually use that?
David Pyke (Dec 10 2019 at 19:06):
Yes, people do.
Jens Villadsen (Dec 10 2019 at 19:08):
well ... double post achievement to me then
Jens Villadsen (Dec 10 2019 at 19:09):
However, the HL7 Org itself ought to have the same opinion regardless of the forum ;)
Lloyd McKenzie (Dec 10 2019 at 19:11):
Not as active. I think the answer is driven by who you hope to communicate with and what infrastructure is in place. It'll also be influenced by how suitable the exchange is to documents. If documents are a good fit and your recipients support CDA but not FHIR, sticking with CDA may make sense.
Jens Villadsen (Dec 10 2019 at 19:12):
consider it greenfield
Jens Villadsen (Dec 10 2019 at 19:12):
no interacting parties besides my clients that I control
Jens Villadsen (Dec 10 2019 at 19:12):
initially
Lloyd McKenzie (Dec 10 2019 at 19:12):
HL7 will never officially pick one of their children as "favourite". They present the options and leave it up to the community to decide what thrives and what withers in a given space.
Lloyd McKenzie (Dec 10 2019 at 19:13):
If totally green field, no good reason I can think of to pick CDA over FHIR.
Jens Villadsen (Dec 10 2019 at 19:16):
lets say that there's a tiny system sitting over in the corner ... and that you at some point ... need to do an export to that system ... and that system is based on XDS and CDA ... would that change your mind ... ?
Jens Villadsen (Dec 10 2019 at 19:17):
would that make you consider changing the entire architecture and base everything on CDA?
David Pyke (Dec 10 2019 at 19:51):
In my mind, that would depend of if that little system over there can manage any other formats. A full system export should be something it doesn't have to convert. So I'd make the big system able to write in whatever format the little system uses internally
Peter Jordan (Dec 10 2019 at 20:39):
HL7 will never officially pick one of their children as "favourite". They present the options and leave it up to the community to decide what thrives and what withers in a given space.
That's not quite correct. One of HL7 International's declared, core strategic goals is to "Establish FHIR as the primary standard for global health data interoperability".
Grahame Grieve (Dec 10 2019 at 20:54):
that's true but I still think Lloyd is right
Jens Villadsen (Dec 10 2019 at 21:11):
https://www.hl7.org/Special/2018StrategicValues.pptx - yep ...
Establish FHIR as a primary standard for global health data interoperability
- it says so ... ;)
Jens Villadsen (Dec 10 2019 at 21:11):
sounds like a favourite to me
Lloyd McKenzie (Dec 11 2019 at 00:09):
That was a radically bold statement for the organization to make, actually.
Lloyd McKenzie (Dec 11 2019 at 00:10):
If your little system in the corner is the only thing you're talking to, then CDA makes sense. If it's one of many, then use FHIR and transform a FHIR document into CDA for the system in the corner.
Brendan Keeler (Dec 11 2019 at 00:34):
Jens, I'd take a step back from the tools and first look at the job to be done. What is your workflow/use case? What data flow between systems accommodates that workflow? And finally, in your region, what standards are in use to facilitate that data flow?
René Spronk (Dec 11 2019 at 07:52):
Strategically CDA is a dead duck for new projects. Open source tools are slowly entering their zombie phase, I haven't presented a CDA training course [anywhere in Europe] in the past 6 years or so (interest in training courses is like the canary in the coal mine).
XDS is certainly not a reason to go CDA, as one could easily exchange either FHIR Documents or FHIR Bundles via XDS as well. Preferrably on-demand FHIR Bundles or FHIR Documents.
Most CDA IGs are mappable to FHIR, so that's the solution for your 'tiny' system in the corner.
The only hesitation I have right now is the fact that most of the 'clinical' resources in FHIR have a low maturity level. If I have to pick one reason for using CDA in a new project over the use of FHIR, it'd be maturity of the clinical models. It'll take some time before FHIR achieves the same level of maturity as the CDA clinical statement model. So if your use case calls for highly complex clinical data, CDA may just still be your best bet. But architect for change, FHIR Documents will overtake CDA in terms of maturity within a couple of years.
Marc de Graauw (Dec 11 2019 at 09:47):
(CDA) Open source tools are slowly entering their zombie phase
That would be partly about ART-DECOR? I'd say "zombie phase" is a slight overstatement. Though I'd agree with the gist of what you say, CDA maintenance will be a huge job for years and years to come, and we still need (and have) the tools to do that. In the Netherlands we still roll out CDA next to FHIR for large new projects such as BGZ. And so does for instance the IPS.
René Spronk (Dec 11 2019 at 09:52):
No. That wasn't about ART DECOR, which is very much alive, supported and also used for projects other than CDA. I'm referring to dozens of other CDA tools that we used to have (at the height of its popularity) and that are slowly dwindling away.
René Spronk (Dec 11 2019 at 09:54):
As long as it aint broken, don't fix it. So CDA will be around for a while. The question was however related to "new projects".
Jens Villadsen (Dec 11 2019 at 09:54):
yep - new greenfield projects - ... where an export at some point in the work flow is needed .... to legacy systems
René Spronk (Dec 11 2019 at 09:57):
We had a community talk about that: "FHIR to Fax gateway" ;-) .. (actually the project was the other way around: from Fax to FHIR).
Jens Villadsen (Dec 11 2019 at 10:21):
HL7 will never officially pick one of their children as "favourite". They present the options and leave it up to the community to decide what thrives and what withers in a given space.
That's not quite correct. One of HL7 International's declared, core strategic goals is to "Establish FHIR as the primary standard for global health data interoperability".
this statement becomes more and more bold as I hear the community is not strictly aligned with it
Grahame Grieve (Dec 11 2019 at 11:05):
volunteer communities are never that well aligned. The question is whether they are aligned enough
Lloyd McKenzie (Dec 11 2019 at 13:19):
I think the community is relative aligned around the statement, but we're trying not to be overt FHIR cheer-leaders. CDA will have a place for quite some time, but it's definitely in the 'decline' phase rather than 'growth' phase.
John Moehrke (Dec 11 2019 at 13:27):
people will be happy to give opinion on what is the best protocol to choose, consultants even happer. Picking a best-protocol is what Implementation Guides do for specific use-cases. Find your use-case in the list of Implementation Guides, and you will find some consensus body decision on what is the best solution (at the time that IG was written). But to expect a preference to come from HL7 at this time, when FHIR has such few parts as normative, is not going to find a single recommendation.
Jens Villadsen (Dec 11 2019 at 13:48):
well - apparently HL7 did that ...
John Moehrke (Dec 11 2019 at 14:01):
I read the text you found as a 'mission', not as a recommendation. I expect any organization has a mission to promote their newest developments. Right?
Jens Villadsen (Dec 11 2019 at 14:03):
Your are right. They are listed as
Core Strategic Goals
John Moehrke (Dec 11 2019 at 14:03):
Im sure that you would find similar statements from any organization about their newest product. If an organization doesn't have a mission to promote their newest product then one should question why that organization is working on that newest product... so, seems natural
Last updated: Apr 12 2022 at 19:14 UTC
