Monday, September 12, 2011

Improving SDO Inter-relationships

Sunday's Q4 session on HL7 relationships with other SDOs was a bit different this time around.  I promised Austin Kreisler that I would report on some of my thoughts that came up as we went through the meeting.  This is my report which I'm sharing with you.  The meeting started with the typical brief report from Liaisons, but we spend most of the times in a Panel Session that included:
Scott Robertson
Liaison to NCPDP and chair of the HL7 Organizational Relationship Committee
John Quinn
Liaison to the Standards Collaborative Organization
Tim Buxton
A participant in ISO initiatives
Christian Hay
Liaison to GS1
Lisa Spellman
Liaison to ISO TC-215 in her new role as ISO Secretariat
In my role as HL7 Liason to IHE
Each member of the panel briefly presented their role and issues they have in the Liason role with respect to the following three topics:
  • Perspectives on cross SDO cooperation strategic road map (i.e., what the strategic priorities ought to be)
  • Additional opportunities for cross SDO cooperation 
  • Ideas on how to streamline existing cross SDO activities
One of the interesting issues that came up, especially with respect to the Joint Initiative Council, was the amount of rework and reiteration that has to go on as ballot items go through respective working groups in parallel.  Decisions made in one work group get reviewed and revised in another, which results in a great deal of reworking, sometimes with small but important divergences along the way.

Now, I remember when HL7, IHE and HITSP were all working together (quite well I might add), on the use case of sharing laboratory reports between healthcare providers.  HITSP was assigned the use case and was working on a solution, IHE was developing the XDS-SD profile (implementation guide) for sharing laboratory reports, and HL7 members were advising on the best way to go about it.  Austin and I were both deeply involved in that work, and were part of the core group of five that followed it through at least two of three organizations (I was unique in physically being present at meetings in all three organizations, others did T-con in).

There were several significant factors that led to the success of that effort that differ apparently from the JIC process.
  1. All of us worked within the organizational governance of the meeting that we were present at, if HITSP, the rules of that organization, IHE or HL7 similarly.
  2. The rules of all three organizations were sufficiently flexible to allow us to pull in outside parties to engage as necessary: Fracois Macary from France in a HITSP meeting, Austin who was involved in HL7 and HITSP as a member into an IHE meeting.  In fact, just about every SDO activity I know allows expert participation with appropriate process.
  3. We reviewed progress at each meeting, but didn't reconsider progress just because we were in a different environment.
  4. We deferred issues to the next appropriate meeting when we needed input from the right organization (e.g., several issues in HITSP and IHE were deferred until we could gather appropriate HL7 expertise).
  5. There was enough of a core group to maintain both the momentum, and the collective corporate memory.
  6. Everybody was willing to cooperate, and everybody got something out of the process, even if only one organization was publishing a guide.
The joint work on XD-LAB is still, to me and to this day, the best example of collaboration across multiple organizations that I've seen.  It was truly a grass-roots effort, where the participants were all interested in each organization, and so they were willing to work out a process to keep things moving forward.  One important contribution to this was that through HITSP (and also IHE), this project had a very clear deadline.

So, my suggestion for JIC projects in the future is that:
  1. Each organization agrees to move the project forward serially through the organizations,
  2. Each organization contributes its expertise in the way that is most appropriate to the project, and the project team agrees on the scope of each organizations contribution, so that issues clearly playing to one organization strengths are addressed by that organization,
  3. Governance is handled through existing processes of each organization where possible, but with sufficient flexibility to allow "outside" contribution.
  4. That a core group be put together to manage the project through each organizations process.  That core group should be committed to attending as many meetings across organizations as possible, and the SDO's should enable that participation as much as possible.
  5. That these agreements be put into place across organizations is as formal a way as possible to enable 1-4.
I don't have the answer to the publication formatting issue that the querant who raised this issue also asked, and that still is a real problem.  However, as John Quinn suggested, I believe the ISO format is probably the best starting target for all involved, and I further believe that organizations should be agnostic as to how it gets into that format (if it starts as an HL7 V3 ballot, tooling should be used to massage it into the ISO format).  Let's not let how it gets there get into the way of getting it done.

1 comment:

  1. hi Keith

    Horses for courses. Some projects can be divided up, so that would be the way to go. Others are creating shared infrastructure for other JIC projects, and would just have to be shared properly. Serially moving them through the different organizations could never work for them.

    The publication issue - using the ISO format delayed ISO 21090 by 18 months, because the ISO secretariat wanted to "fix" things that they wouldn't have wanted to fix if it wasn't in ISO format. That has to be a big negative vote against the ISO format.