Thursday, December 01, 2011

And now even CDA is troubled

From:  [] On Behalf Of Bob Dolin
Sent: 28 November 2011 19:07
To: Structured Documents WG
Subject: RE: CDA or greenCDA


Who is asking to have greenCDA sent over the wire, and what is their rationale?

I don’t agree with your learning curve statement. I would expect that most recipients would use a general purpose CDA parsing algorithm. In fact, if you deal with many types of CDAs, it can be harder to receive  greenCDA, because you must insert an intervening transform step.

As for the cover about 80% of the requiremen this is because greenCDA is a simplification, that removes some CDA bells and whistles. Is not that greenCDA cant express anything that is in CDA, but that we make design decisions when simplifying.

As for Robert Words proposed rule Every Green CDA instance should contain the URL of a transform to take it to the full CDA it is derived from I dont think that is sufficient. I think ll need to ballot  greenCDA schemas (e.g. a greenConsolidation) if we are to use it as a wire format.


Bob Dolin, MD, FACP
President and Chief Medical Officer
Lantana Consulting Group
t: 714.532.1130
c: 949.466.4035

From: [] On Behalf Of robert worden
Sent: Tuesday, November 29, 2011 3:33 AM
To: 'Bob Dolin'; 'Structured Documents WG'
Subject:  RE: CDA or greenCDA
Hi Bob  -

I think HL7 has two distinct groups of customers:

A.     Those who understand the RIM and CDA; have several different use cases for CDA; and might have invested in a general-purpose CDA parsing algorithm
B.      Those who don’t understand the RIM; have just one or two use cases to address; and want to do it as simply as possible.

We need to listen to all our customers, in both groups.  I think those in group B greatly outnumber those in group A. In the UK I know they do, for a fact. Most healthcare providers and most IT suppliers simply dot want to know about the RIM; they just want to build the interfaces they need, with as little fuss as possible.

But it is the group A customers who have the time to invest in HL7 to come along to WGMs, attend weekly calls, follow threads like this, define the standards and take part in ballots. It is group A who have overwhelmingly defined the direction of HL7; whereas group B are the majority of our customers. I think this systematic bias in our process has misled us over the years.

Green CDA suits group B just fine. They look at a Green CDA instance and say Yeah, I can do that.; whereas a full CDA instance induces silence and shaking of heads.  So recommending Green CDA, without even saying: full CDA is the real thing; yd better send that over the wire is what suits group B.

But the great thing about Green CDA is this: provided we have reliable transforms in both directions (full <=> Green) we don’t have to choose between our customers, because both groups A and B can work the way they want to. The issue of what goes down the wire should become as trivial as to zip or not to zip; you just apply a transform to get the form you want. Also, Green CDA is a handy migration path for group B users who want to move to group A.

On your last point (transforms alone are not enough) I fully agree. The proposed rule was an additional requirement if Green CDA is to go on the wire, not the only requirement.  I had assumed without saying that to publish any Green CDA, you should at least publish a schema to define its structure; and also (more important) publish its semantics in a RIM-independent for so group B can know what it means.

Best wishes


mobile: 07970 197968
landline: 01353 777668

From:  [] On Behalf Of Pratt, Douglas (H USA)
Sent: Tuesday, November 29, 2011 9:52 AM
To: robert worden; Bob Dolin; Structured Documents WG
Subject: RE: CDA or greenCDA

In my experience helping implementers, the most frequent request I had was for examples of "how do you represent..." that may or may not be helped by Green CDA (as there currently is no Green CDA to evaluate). 

Some common ones:
- How do I represent no known food allergies?
- How do I represent that the patient has positively asserted no known allergies vs. no information (not asked, etc.)?
- How do I represent that this allergy had an onset of "adolescence", when CDA requires a date?
- How do I send legacy codes along with their translation to industry code(s)? 
- What data pertains to visits, and which across visits (CCD)?

I found that if I could provide an example, they could work with it. The markup was not really an issue in that case.

- Doug

From: Rishel,Wes []
Sent: Tuesday, November 29, 2011 10:33 AM
To: Pratt, Douglas (H USA); robert worden; Bob Dolin; Structured Documents WG
Subject: RE: CDA or greenCDA

Are all the how do you represent questions you listed explicitly spelled out now in the CCDA?

I would note that the only guidance on the CCDA on sending blood pressures is a textual note under vital signs that says blood pressures should be sent as observations.

Wes Rishel

From: [] On Behalf Of  Bob Dolin
Sent: Tuesday, November 29, 2011 12:55 PM
To: Rishel,Wes; Pratt, Douglas (H USA); robert worden; Structured Documents WG
Subject: RE: CDA or greenCDA

Hi Wes,

What is CCDA?


From: Moehrke, John (GE Healthcare) []
Sent: Tuesday, November 29, 2011 11:53 AM
To: Bob Dolin; Rishel,Wes; Pratt, Douglas (H USA); robert worden; Structured Documents WG
Subject:  RE: CDA or greenCDA

I presume it is Consolidated CDA

Seems this is an effort to make this seem like not-CDA.

We need to be very careful at how all of this is perceived by the thousands of people who only get to see these efforts when they come up to breath. By bringing in yet-another-acronym, there will be more confusion. We really dont need or want confusion, right? We need to be working together to bring down the FUD, not make more of it.


From: [] On Behalf Of  Rishel,Wes
Sent: Tuesday, November 29, 2011 5:31 PM
To: Moehrke, John (GE Healthcare); Bob Dolin; Pratt, Douglas (H USA); robert worden; Structured Documents WG
Subject: significant concer re Moehke's response (was RE: CDA or greenCDA)

John: Seems this is an effort to make [Consolidated CDA] seem like not-CD

I take substantial umbrage that John implies that I have a deep and dark motive to somehow separate CCDA from CDA, particularly as my recent blog talked up the CCDA as a significant step forward for the CDA. This style of innuendo really does nothing to advance discussion on issues.  Many people properly discount ad hominem jabs like this, whereas for others it tends to deepen schisms rather than finding narrow points to build bridges.

A simple statement such as we must be careful that the acronym leads people to believe that the Consolidated CDA is something different than C would have raised the issue without appealing to conspiracy theory or implying that I had baser motives.

I make no bones that I have fear, uncertainty and doubt about the ability of the US to pull off a high degree of interoperability with the C32, how I think that CCDA is an improvement but that there is still more that needs to be done. However I do my best to address the issues, not the people, bring other experience into discussion and look for solutions. These are the same issues that I shared privately with Board when I was a member and held reasonably close until I had been off for awhile. After a while, though, I felt that the better thing to do was to express my issues on an HL7 list server rather than only in other venues.

Some would say that casting FUD is a bad thing. It is, when it is an attempt to obscure discussion on the issues. Just the same, decrying FUD is a bad thing if it is done for the same purpose.

I have actually learned a lot in the discussions over the last few weeks, some of which has been reflected in my blog. I would say, however, that if there are specific rules of decorum or a general sense on the blog that it should only be used for discussions that are pro the group sentiment, I would have to conform by going elsewhere.

This not the HL7 that I remember.

Wes Rishel

 From: "Boone, Keith W (GE Healthcare)" <>
Subject: RE: significant concer re Moehke's response (was RE: CDA or greenCDA)
Date: 30 November 2011 05:03:10 CET
To: "Rishel,Wes" <>, "Moehrke, John (GE Healthcare)" <>, "Bob Dolin" <>, "Pratt, Douglas (H USA)" <>, "robert worden" <>, "Structured Documents WG" <>
Reply-To: "Boone, Keith W (GE Healthcare)" <>

Wes, frankly, I read something very different in what John had to say, and I’m certain it wasn’t intended as a personal attack on you. 

I deal with CIOs and CMIOs frequently enough who still stumble over CDA, CCR, CCD and C32 and cannot tell the difference between them, yet are responsible for making sure they’ve got whatever ‘it’ is, implemented correctly for their facility or practice.  We just shifted the ground from underneath them with a new acronym.  After all, if Bob Dolin doesn’t know what CCDA is, how could they possibly be expected to know it.  You don’t set out to create FUD, there is no conspiracy, and nobody (at least in my reading) has accused you of starting one.

When even Bob Dolin has to ask you what CCDA is, it clearly delivers the point that this thing we’ve been calling the Consolidated CDA Guide needs a better name, and needs some marketing and customer education to explain to folks what it is.  Adding YAA (yet-another-acronym) to the alphabet soup isn’t going to help.  I’m already getting questions from the people who follow you at a policy level about what is CCDA, and even worse, there’s been other industry press describing this work that misquotes and misrepresents what it is.

If there’s a failure to communicate what the Consolidated Guide is, and there’s more FUD because of it, it isn’t your problem Wes, and you didn’t cause it.  It’s HL7’s problem, and we need to fix it.  That’s what I take away from John’s point.

As for names, there was some discussion of renaming for publication.  It might well be worth bringing that back up and thinking about how to market this effort professionally.  There’s a brand to be built here, a story to tell about what when into this work and the communities that built it over the last six+ years, and a community to educate about these guides and what they mean for their end users.  HL7 as an organization doesn’t have a great history at doing that sort of marketing, but somehow we need to figure it out.

Keith W. Boone
Standards Architect
GE Healthcare


No comments: