Convert your FHIR JSON -> XML and back here. The CDA Book is sometimes listed for Kindle here and it is also SHIPPING from Amazon! See here for Errata.

Tuesday, October 30, 2012

Automating the Transition Towards CCDA in IHE PCC

One of the discussions in the IHE PCC Planning meeting this week (being held mostly virtually due to travel challenges created by hurricane Sandy) is about how to converge towards C-CDA instead of basing the profiles on CCD.

There are two challenges:
  1. C-CDA made some non-backwards compatible changes in the new templates, necessitating a transition strategy.  
  2. C-CDA is focused on US requirements, and the IHE PCC-TF is used Internationally.  There are some US-specific vocabularies that simply won't fly internationally.  Even SNOMED-CT will be challenging for some international users.
I think the solution is to introduce versioning into CDA templates, as I've suggested previously.  Then we can create new versions of the IHE templates that are more in harmony with the C-CDA templates, applying the C-CDA rules where feasible internationally.

My hope is to use MDHT to capture the existing PCC Technical Framework rules (actually, to verify the transcriptions already available here), compare the rules to C-CDA, and apply some logic within Eclipse across the models to create new templates that move us towards C-CDA.  When the results of the logic produce templates that are identical, we'd just adopt the C-CDA templates.  When they didn't, we'd create new versions of IHE templates that adopted as many of the C-CDA constraints as possible.

With luck, I'd be able to output all the PCC Templates from MDHT.

In that way, a C-CDA document could become the US National Extensions to IHE.  Perhaps we could even ballot the intermediate result as a DSTU through HL7, as an international realm version of C-CDA templates.

Fortunately, this work won't begin until (I hope), I've made more substantial progress finishing up Query Health and ABBI.  Otherwise, you might as well just shoot me now.

I expect some will want me to use Trifolia for this effort.  But there's a lot more I could do with MDHT, and the benefit is that MDHT is Open Source, whereas Trifolia isn't.  And if I do it right, and what comes out of MDHT goes back in, I've already built a reference implementation and testing tool.  I cannot do that with Trifolia today.