Wednesday, June 30, 2010

Democracy In Action

I listened today to the HIT Standards Committee discussion on Electronic Document Standards for Discharge Summary & Other Encounter Summaries presented by Dr. Jamie Ferguson, leader of the Clinical Operations Workgroup.

The slides Jamie used are embedded below.  You can find the official deck on the HIT Standards FACA website when they get past their technical problems.

As Jamie and these slides point out (see slide 2), users of CCD and CCR also need to provide other kinds of documentation, for example, Discharge Summaries (required in meaningful use).  What the workgroup found (no surprises here, especially if you read my second ever blog post), is that CCD and CCR don't support the kinds of information needed for other kinds of documents.

As John Halamka stated in the meeting:  "The last thing we wanted was a CCD that has been modified for every hospital, because it is a standard document."   Nancy Orvus pointed out later that the CCD was never intended to be the future of all clinical documents.  She's absolutely correct, and I should know, having been one of the editors and a cochair of the workgroup that created it.

Jamie Ferguson points out that there is a pretty well-recognized group of content templates that can be used.  I'll add that HITSP already did a lot of review of those templates over the last four years, and you can find a pretty good collection of them in the HITSP C83 CDA Content Modules specification.  HITSP Care Management and Health Records modified Capability 119 Communicate Structured Document to support the use case that addressed this very same problem.  The following quote is from that document:

HITSP/CAP119 Communicate Structured Document addresses interoperability requirements that support the communication of structured health data related to a patient in a context determined by the author of the document. This Capability supports the exchange of a broad range of CDA documents related to clinical patient care. The following are examples of the type of CDA structured data that are supported:

  • Continuity of Care Document (CCD)
  • Emergency Department Encounter Summary
  • Discharge Summary (In-patient encounter and/or episodes of care)
  • Referral Summary Ambulatory encounter and/or episodes of care
  • Consultation Notes
  • History and Physical
  • Personal Health Device Monitoring Document
  • Healthcare Associated Infection (HAI) Report Document
  • Labor and Delivery Record
  • Antepartum Record
  • Operative Notes
  • Procedure Notes
Senders and recievers of clinical documents using the HITSP Capability 119 were REQUIRED to use sections and entries defined in the HITSP C83 Content Modules specification.  This document collected templates from more than half a dozen clinical document implementation guides, and harmonized the section and entry definitions so that problems, medications, allergies, et cetera, would appear IN THE SAME WAY in every clinical document used.

One of the problems identified by one of the HIT Standards Committee members (I believe it was Janet Corrigan) is the potentially infinite number of possible combinations of data and thus documents that could be produced.  I have to agree, if you focus on "documents", the number of different types of documents being used in the world is indeed quite variable. 

However, the number of different kinds of document sections is not, and there is a large body of work already done in this area.  Jamie Ferguson speaks later on in the meeting to this point, and indicates that the work in the area seems to be tailing off.  As someone whose been engaged in HL7, HITSP and IHE activities in this area, I have to agree.  The last few CDA implementation guides that I've worked on have required fewer and fewer new templates, and at least one was built wholey from existing work.

One of the interesting things that's happened to the IHE PCC Technical Framework and the CCD Implemention Guide is that it isn't the documents described by those publications that are being reused globally, it is the sections and entries.  These sections and entries appear in international specifications published in the French DMP Project, in Europe through the European Patient Smart Open Services (EPSOS) project, and as a work in progress through a group led by the Ministry of Health in China.  These are some of the very same sections and entries found in the HITSP C83 specification.

So, the question isn't really about how many different types of documents we need, but the number of different sections.  One of the HIT Standards Committee members put it forth in a way that was very succinct for me, but probably not for others.  His thought was that the best way to describe a complex space is with a basis of reusable modules.  He's clearly had some mathematical training (a basis is a way to represet a "vector space"), and that just tickled my funny bone.  That's the whole point of C83 and the templates that have been subsequently developed by HITSP, IHE and HL7.

Wes Rishel points out that there needs to be a way to continue rather than update existing standards work.  I would happen to agree.  HITSP delivered results on this request, and while HITSP is no longer active, I strongly reccommend that work to whatever organization takes this specific problem forward.  I'm also hoping that we learn soon what the direction will be for the new Harmonization framework, and how these pieces all fit together.

These were the "working" recommendations of the Clinical Operations Workgroup:
  • The process should standardize templated CDA sections to build upon and extend what was done in CCR and CCD
  • They note that this is consistent with the NIST direction for testing (I would also note that NIST already has tests for many of the C83 sections already).
  • We must enable more documents and reuse of existing work.
  • We may also recommend this direction for attachments (as I suggested in January)
There's more discussion to be had, since they were running out of time for this meeting.  This topic was suggested to be continued in the next HIT Standards Committee meeting.

As an observer of the FACA process, I am happy to find that the Clinical Operations Workgroup has come to the same conclusions that HITSP and others have.  I am a little annoyed with the rehashing of these topics, but as one speaker pointed out at the start of the afternoon: this is democracy in action, and "democracy is the worst form of government ... except for all the others."

As long as the results of this democracy produce what is best for patients, and remains in action, rather than inaction, I continue to live in hope.


  1. Your observation about the reusable modules is likely too elegant for the HIT-**** community. Fortunately they have surprised me that by the third time an elegant solution is presented enough members catch on and convince the others that it is indeed an elegant solution.

    As to HITSP, I think it is very much time that an organization that uses standards like processes be started to focus on and document the US Realm. It is unfortunate HITSP lost out, but their process was flawed and funding short. We need an IHE-US-Realm.

  2. Keith, thanks for the post. But you may want to fix the formatting of the quote from CAP 119, because it appears to contradict the main point that you're making. The lack of a bullet in front of Continuity of Care document makes CCD look like an "umbrella" over all the documents types that follow, rather than what it should be, just another bullet in parallel to the rest.

    "The following are examples of the type of CDA structured data that are supported:

    Continuity of Care Document (CCD)

    * Emergency Department Encounter Summary
    * etc."



  3. Thanks for reporting on this with clarity and insight.