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.

Friday, August 31, 2012

Standards as Innovation

Over on G+, +Margalit Gur-Arie (AKA @margalitgurarie on Twitter) writes on last night's post:
That "lost the distinction between innovation, and standards" is a bit disturbing, don't you think? Particularly when the standards created this way are enforced by governmental regulation before anything was even tried out in practice.
She's got a great point.  The pace of Meaningful Use is frenetic, and the deadlines are insane.  Those deadlines also happen to be law.

But the lost distinction between standards and innovation in healthcare wasn't caused by the HITECH Act.  Instead, it's been something that has been happening since I got my start in standards in the late '90s.

As I think about it:

HTML and XML were standards that were based on technologies that had been proven in the field, prior to being standardized.  XML Schema combines XML and XML Document Type Descriptions in what might perhaps be considered a novel way, but functionally, has all the same capabilities (and just a few more) that DTDs had, but in an XML representation.  The XML and HTML Document Object models standardized what vendors like Netscape and Microsoft had already implemented.  Even CSS harkens back to earlier technology (DSSSL).

In the ideal world with standards, when there are two (or more choices), you should pick the best one, and when there is no "best one", pick one way.  In a a battle between A and B, where there is no clear "best way", C is often better by virtue of the fact that it is new, and everybody will be equally disadvantaged.  Back in the era of the Browser wars, you'd often have two 800-lb Gorilla's having a dominance battle in a room full of Marmosets.  So, the marmosets introduce something new, and the problem is solved.


The pace accelerates.  Real years become Internet Years, and the need for innovation just to survive in this electronic world becomes a necessity.  And one shortcut becomes to make new innovations, the new standards, maybe even before they are ready.  Novelty becomes a way of solving these problems in standards, and avoiding giving away an advantage to one side or another.  This may in fact be one of the reasons why the LRI project developed something new, rather than picking between the existing choices.

On that front, LRI is new. But we've already had a few years experience doing very similar things with the two competing standards.  So, really, it isn't all that new.  It is pretty untested.  But what we do know is that industry is capable of doing it, because they've already done in twice in different but very similar ways.

You also run into situations like HL7 Version 2 experienced.  The standard, over time, got crufty.  There was a need to do something new.  But, because it was a standard, nobody wanted to innovate on their own. So V3 was born.  And because nobody was innovating on their own to help develop a body of experience in which we could base the standards, HL7 introduced quite a number of innovations of its own.

Of course, there were other innovative standards (OpenEHR), but then we get back to the Gorilla problem.  It's taken a while for option C to show up in that long-standing battle, but I think CIMI shows some promise.

Innovation in standards development has become a way of life.  But Margalit is right, what we need to do is ensure that these innovations are tested before they are deployed.  I was very encouraged by the completeness of the development framework that ONC created for the S&I Framework project.  It includes reference implementations and pilots.  I've never been pleased with the schedule.  We all know how rushing to hit a deadline affects quality.  Unfortunately, on the schedule, neither ONC nor CMS are in the driver's seat.  Congress was, and they put those deadlines in the HITECH Act.

It's up to us to meet them.  How do we do that?

There are parallel paths.  Tactics and Strategy.  Short term, and long-term goals.  Don't confuse your strategy for your tactics.  PCAST as short-term tactics sucked.  As a long-term strategy, it's got a couple of good ideas (and some bad ones).  S&I Framework is tactical.  It's about how to get something going quickly with the people and resources that we have on the ground today.


0 comments:

Post a Comment