Wednesday, October 23, 2013

Template Versions ONE LAST TIME (I hope)

We're still stuck on the template identifier versioning problem in HL7 as we try to finish up the CCDA Release 2.0 specification.

The challenge is simply this:  In order to clean up after ourselves, we have two choices:

Change just about every template identifier we are currently using in the guide, and replace it with a new (but likely similar identifier).

So:
<templateId root="X"/>

Becomes in most cases:
<templateId root="X.2" />

But that doesn't happen the same way every place because some of those values for X come from a different namespace and have to be totally replaced, and some of those X.2 values may already have existed as well.

OR do something like this:
So:
<templateId root="X"/>

Becomes in all cases:
<templateId root="X" extension="V2.0"/>

And of course, we could always use V2.1 or V2.2 for minor version changes.  This latter course is compatible with the current solution that the Templates workgroup has proposed, and also the course that IHE went with (working with HL7 Templates) to upgrade its templates to work with CCDA.

Thursday (tommorrow) is when we make big decision in SDWG.  There are two camps still, and they seem to be divided not really over the solution, but rather over the timing of it.  For the most part, I believe we are all agreed that when Templates makes the final decision, SDWG will follow it for future release.  But in this particular case, there are some outside pressures pushing SDWG to have CCDA Release 2.0 out the door before templates can make its decision.  That pressure seems to be coming from the usual source: ONC/CMS.

Now, I have to ask, one more time.  Do we really need to make this kind of trouble for ourselves? Or can we do the right thing?  Frankly, the right thing is to let templates make the decision, and follow it.  Given the recent interruptions in service from the government, I'd say that several bets about deadlines are changing.  I know the HIT Standards Committee was already rethinking its deadlines before the shutdown, and now are almost certainly not going to reach their goals. I really would like to see ONC/CMS back off on any deadline pressure that they have, and let HL7 governance be able to address the specification development the right way.

When I count the voices I've heard, it seems overwhelming that the negatives are coming from a very few sources.

If you care about this issue, and want to be heard, I ask for one thing: Show up on the Structured Documents call tomorrow (See the Upcoming Calls Section of the page).  We'll be covering the topic in the first hour (10:00-11:00 Eastern).


0 comments:

Post a Comment