12:40am and here I am in Washington, DC after some great discussions at New York Presbyterian during the day and a treat of a ride on the Acela Express down the coast during the evening. I really should be in bed --- we have seven hours of NHIN Direct discussion scheduled for tomorrow -- but the guilt of leaving my blog untended for so long has finally reached un-ignorable proportions.
As it turns out, NHIN-D is really the primary reason for my radio silence. Between me and Umesh we are spending a ton of time trying to help push this initiative forward. So far at least, the time investment has been more than worth it - it really is a moment in time where things seem to be aligning for real progress. I am super-excited.
So what is NHIN-D anyways? The elevator pitch is easy: NHIN-D is secure email for docs and patients. Just give your address to your provider and they can send you your immunization report, recent visit summaries, whatever. A provider looking for a second opinion can safely send background information to a specialist. Lab systems can send results to patients and providers. Pretty simple, really.
Well --- simple once we get all the edge cases figured out. I seem to be finding myself in a bit of a middle position, trying hard to keep things as simple as possible, but not so simple that we miss critical requirements necessary to make the system adoptable, durable and truly useful. The good news is, there are a ton of great people committing real brainpower to working the issues through --- I am honored and lucky to be part of the team.
The last couple of weeks have been particularly fun because we've started to transition from "talk" mode to "do" mode --- time to let the code talk and see what really sings in the real world. And that means a new level of responsibility for participants; we're asking folks to contribute working code to a set of volunteer concrete implementation groups as part of a bake-off to help us make informed decisions about the next level of system detail.
Here at Microsoft we're putting our marbles behind an SMTP-based implementation, creating an agent that can sit within existing email pipelines (like this or this) and handle our still-emerging security and trust requirements. Our premise is that:
- email works pretty darn well for trillions of messages every day,
- there is already a ton of battle-tested code supporting it, and
- it scales easily from small-practice web clients to deep EMR integration.
We hope to write as little code as possible; Umesh told me a couple of hours ago that he is nearly done with a first cut of the agent implementation --- so when I get home we just have to start plugging it in to show an end-to-end working pipeline. WOOOOO HOO!
OK --- 1:14am now and really time to get some rest. Hope to see you around the NHIN-D Wiki --- we can use all the help we can to get this stuff running!