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, June 17, 2011

Competition in Standards

If you read "Go ahead, I dare you" the other day, you might be wondering why I wrote it.

I spoke to a number of different people over the past couple of weeks, and several other conversations were reported to me.  Many of these statements I've heard before.  Most of these were second or third hand reports, but many I've heard from multiple sources on the same topic.  I'm sure lots of you reading this can fill in the blanks appropriately in the following statements (and there are multiple correct answers).  My point is not to call out anyone making the statements, but to illustrate the problem:

  1. What do you mean there's (an SDO/Profiling organization) stuff in (project).  It was supposed to be the (the previously mentioned SDO/Profiling organization) killer.
  2. (organization and/or person) wants to (various words meaning kill) the (an SDO/Profiling organization)
  3. (organization or person) doesn't like (SDO and/or Standard) so we can't talk/think about them/it for this use case. 

Few of these statements are anything that  the people attributed to them would ever say publicly.  Some because it would be bad politically, others because it would be outright suicidal to their careers, and/or contrary to existing government policy on the use of standards, which strongly promotes the use of consensus standards.  None of them help us move healthcare where it needs to be.

Some of the complaints about existing SDO/profiling organizations are that:

  1. Their work is too complex (e.g., CDA and/or XDS).
  2. Their work doesn't meet the needs of a particular stakeholder (e.g., small providers and/or vendors).
  3. Their work is driven by ___ (e.g., Big vendors).
  4. Their work is not freely available to all (e.g., HL7).
  5. Their work is not universally adopted or implemented.
  6. We tried this standard once and it didn't work (e.g., C32 in VLER).
The solution to these problems is not to oppose or replace these organizations, but to change their behaviors by participating and contributing.  None of them are exclusionary.  Many have membership requirements and all have costs associated with them (time and travel at the very least).  If the particular issues are reported back to the appropriate organizations, many will be willing to (and some are already) addressing them.

I've been there.  The "Kill" strategy on standards is my least favorite, most intensive, and most dangerous strategy to execute upon.  I'd much rather work from within.  I don't get everything I want, and it isn't always speedy, but I find it to be a much better approach.

I have a problem with all these internecine battles.  It results in emergence of so many projects and work groups that I cannot pay attention to all that is going on (and this is my full-time job, unlike many other of the volunteers in the standards space).  SI Framework just spent several months addressing one of these issues because there are at least four different laboratory reporting implementation guides that could be used, and what are they going to do?  Probably build another one to obtain consensus.  These projects are all subject to Rishel's law:  Change the consensus group and you change the consensus.  

While true, the consensus of many of these projects not often dramatically changed.  And from the ONC perspective, what they've repeatedly asked for is "good enough".  If nothing major is changing, then why are we spinning our wheels?  What we had already was probably good enough.  (Note: I'd love to see is a good study of Inter-rater reliability across the different standards projects).

So, the point of my post is that we need a place where we can do this once, instead of over and over and over again.  SI Framework, IHE, HL7, eHI, CfH, CHCF, et cetera, aren't managing it alone.  We need to work together and come to consensus ONCE, and then implement and deploy it.

A majority of the participants will be the same set of people we see at all of the various meetings and involved in the projects.  But it will be broader too. But instead of having to have three or four different projects in three or four different organizations to solve these US problems, maybe we could just have one.

Anyway, that's enough on this topic for now.  I need to get back to the real work.

1 comment:

  1. One of my pet peeves is persistent politics within consensus standards development and profiling groups. It is lamentable that some people and organizations hold their own work products, opinions, or needs for influence in such high esteem that they thwart the interests of other stakeholders. Collegiality is undervalued.

    My larger issue, though, is the lack of effective representation of some stakeholder groups. For example, health care consumers are poorly represented in two aspects: raw numbers of representatives versus other stakeholders, and misrepresentation by special-interests that do not represent a consensus viewpoint. The first responder community is not well represented, either. Clinical standards are made worse when there is a lack of consensus in nursing informatics. Standardization needs for clinical and financial communities in health care are met with minimal coordination. And so on...

    It is unreasonable to expect all of the issues will be fixed simultaneously. What we desperately need is prioritizing. I think this is best done with the help and sponsorship of government through facilitating communications, selectively funding projects, and establishing accountability for results.