Tuesday, June 15, 2010

Black smoke on NHIN Direct again

The black smoke continues to billow over NHIN Direct.  On today's Implementation Workgroup call, two basic proposals were made:
  1. Continue to develop 2-3 implementation approaches by forming communities of interest around each one (as suggested by Wes Rishel)
    • Communities will likely form around SMTP and IHE/SOAP
  2. Seek common ground and Unify the Group with a single approach
    • SOAP with an e-mail bridge model
    • REST with SMTP
There was a lot of discussion, with some of it divisive and some of it trying to work towards consensus.

I believe in general that the group is moving towards trying to seek common ground, and that sentiment was strongly supported by many EHR vendors. 

One of the problems on today's call was that these one line descriptions of the proposals did not present enough information about what was included in them.  We did not come to a decision today, but Vassil Peychev proposed that the remaining proposals be written up and a call be held on Friday to review again. David McCallie and others are working on write-ups now. Given that suggestion occured near then end of the NHIN Direct call, Ariensuggested that be the go forward plan.

Let me briefly summarize how I think these proposals work out:
  1. The "go your own road" proposal around communities of interest seemed to be a non-starter.  Two people really supported it, but I heard strong support (including my own) on trying to come to consensus, even though it may be hard.
  2. There seem to be two "grand unification theories" going around:
    • Use the IHE XDR protocol as the backbone and provide a bridge to the SMTP model.
    • Use SMTP as a backbone, with REST as an interface layer, and IHE XDM as recommended content.  This would enable step up/down to/from current NHIN and IHE protocols (XDR and XDS). [I strongly support this theory].
Note that I'd be happy with XDR as the backbone, as the IHE model already demonstrated that capability, and I'm much more comfortable with the security of that model.  That really speaks though to my own comfort level with TLS and SOAP, and clearly, doesn't address the complexity issues that others have raised, nor will it, I believe, be able to move forward as a consensus position, given the fact that two parties are very strongly against that option.  Given that the XDR as a backbone position doesn't seem to be in a position to generate a consensus viewpoint, I very strongly support the next choice.

There was still a great deal of discussion / dissent over the lack of representation from the smaller providers who this is being built for.  The next time this process comes around, there ought to be some sort of provision to provide for them to be directly represented (EHR vendors do have these providers as customers, but representation by proxy is not seen as valid representation; at least in some of this group).

There are also a few who have expressed concerns that large vendors are still dominating the process.  However, I would also note that it seemed to be those same vendors strongly expressing desire to reach consensus, rather than divide up around preferred approaches.  

I also heard some concerns raised that a very small but strong minority opinion was blocking a majority viewpoint, and also that the majority was trying to work with them. 

This was another hot-button topic, and I think that the words minority and/or majority, and large/small vendors are keywords that should be used very carefully in any NHIN Direct call that wants to make progress (I'm just sayin').

There were a number of comments about scalabity of solution A vs. solution B.  In a community of experts however, an expert opinion carries little weight when two or more the experts disagree.  There was one attempt to shift grounds from "technical scalability" to "social scalability" of the two alternatives.  It was an interesting discussion, but I didn't quite get it. 

All in all, it was a pretty intense and wide ranging (perhaps too wide ranging at some points), discussion.

We are all very frustrated, and that comes from a number of places.  The ability of a small group to hold up a larger group, the struggle by under-represented but important stakeholders to have their viewpoints heard, the desire by some (including me) to move things forward.  The newness of this process, our lack of experience with it, and our inability to reach some of our timelines contributes to all of this frustration.

Even with all the frustration, we've gone from four possible outsomes to somewhere around two in a matter of a week, and that is indeed progress.  I have to give Arien Malec a lot of credit (worry not, I'll give him grief like I did a few weeks ago when it is deserved).


  1. Keith -- thank you for this very helpful summary, and feel free to keep poking at me when deserved.

    You are right that there is a lot of attribution of bad intentions going on that's slowing things down a bit. I've been talking to everyone, and all I hear is a strong desire to do the right thing for multiple constituents:

    Small practices
    Large practices
    Complex health systems
    Different technology models

    The attribution of good intentions is critical to reaching consensus, and I very much appreciate your tone here.

  2. Keith, I've been reading your blog for a year now and have heard you speak at multiple healthcare conferences (including the recent Governor's conference in Boston). Your summaries are always well informed, have a positive spin, and do a great job of keeping me up to date on the latest happenings in a very busy landscape.

    Thanks a bunch and keep up the good work!

  3. Keith, good post, thanks! As far as representing smaller organizations directly rather than only by proxy (though I see nothing wrong with proxies as well), I hope that ways can be found. Web conferencing is a low-cost equalizer (vs. requiring travel) that helps. But there still is the issue of the time commitment. There was one year on CCHIT where we were blessed to have a SOLO practice doctor join the Interop WG, who was a valuable contributor, but could only afford (time-wise) that single year. But even that's the exception rather than the rule for most SDOs, projects like NHIN Direct, and even the public participation in FACA meetings like HIT PC and HIT SC (open to everyone, but who has the time?). I would like there to be a solution to broaden involvement (in addition to proxies), but am a bit stumped what that would be.