Friday, November 21, 2014

My First AMIA

Believe it or not (and most people there found it difficult to believe), prior to last week, I had never been to an AMIA meeting before.  I think part of the reason few believed I had never been is because I knew so many people there, whom I have met in HL7, IHE, HIMSS, HITSP, S&I Framework or other settings over the past decade.  And while I knew so many, even more knew me.

The scientific sessions were probably the most interesting.  Adam Wright had a great presentation on using Possion parameters to detect changes in alert activation levels.  What was cool about that was that you could apply the technique to any countable event, and I've got a ton of different applications for that.  Another presenter had shown (or though she had shown) that NOT using a system was better than using it badly, but in fact, as it turned out (at least in my analysis), she had NO data in the study to illustrate that at all, and in fact, in her control arm, where the system was not used, folks did not fair as well as when the system was used imperfectly.

One session I attended was about the formation of Healthcare Services Platform Consortium (HSPC), a panel discussion led by Stan Huff and other HSPC members.  Wow!  Just what I needed.  Another consortium to pay attention to.  Oh, and this one is run by physician centered organizations, so it is clear that it will be balanced.  I get a little tired of consortium creation and a job function.  After all, there's already not enough things going on, including IHE, HL7, CCC, HIT Standards Committee, CommonWell, S&I Framework (and before that HITSP), CIMI and ...  I honestly don't know how Stan does it, given that he's leading up four (now five) of these efforts, but I sure don't have the travel budget for this.  Don't get me wrong, I really don't have any complaints about direction, just lack of coordination.  We have plenty of that going on already.

There was quite a bit of buzz about FHIR going on throughout AMIA, and it played a major role in at least 5 sessions that I attended.  Healthcare Informatics called it a Smoking Hot Topic at the meeting, and Neil Versel talked about it as a Public API on his blog over at Healthcare IT News.  I think we've just recast Phoenix Rises as a FHIR-bird.

Probably my favorite activity was the poster session, where I got to meet people doing some very real implementation work, many using the standards that I know and love, and write about here.  There were easily half a dozen projects that I found interesting, and I really enjoyed spending time with the people investigating the use of these standards.  In several cases I was able to point them to some other things they might also look at to improve further.  For example, one person was showing a system he and others had built to merge CCD documents, so I pointed him to the IHE PCC Reconciliation profile for some advice on how to identify and display reconciled medication lists.

I also find it fun to walk down University row, and say "Sorry, I'm already in a program..." whenever anyone asks me if I'm interested.

Of course, while AMIA was going on, the rest of the world didn't stop.  I had double homework in one class (at least if I wanted extra credit), and a draft of my term paper due for another.  Which meant that I got precious little sleep.  I'll plan better for next year, because while this year is my first visit to AMIA, it won't be my last.



2 comments:

  1. Keith, I'm not certain, but I think you mean Poisson parameters, as opposed to to Possion...

    ReplyDelete
  2. I appreciated your Monday challenge to the HSPC adherents.

    In fairness, I think the distinguishing characteristic of the HSPC approach is its orientation with the software service model, and the composite EHR applications that this enables. Defining standard, application layer interfaces will be a nightmare, however.

    ReplyDelete