Sep 07 2013
It was good meeting you at the Second Annual OSEHRA Summit meeting yesterday. It looks like you really have your hands full with your work on planning the evolution of VistA. I thought your question, to the effect, “How do we decouple our information architecture from the organization chart?” was right on target. I addressed this my recommendations to Chuck Hagel in a previous open letter:
Decouple the IT architecture from the Organization Chart. The designs that I’ve seen coming from the DoD are enterprise-focused, “baking in” all of the stovepipes, organizational turf wars, and protecting rice-bowls of the many political, economic, and professional constituencies hoping to influence the architecture. Instead of patching together an “integrated system” of point-to-point connections, we need to move to a broader vision of creating a common information space. Note the words of Tim Berners-Lee in his design of the World Wide Web:
What was often difficult for people to understand about the design of the web was that there was nothing else beyond URLs, HTTP, and HTML. There was no central computer “controlling” the web, no single network on which these protocols worked, not even an organization anywhere that “ran” the Web. The web was not a physical “thing” that existed in a certain “place.” It was a “space” in which information could exist.”
I was pitching an idea called HealthSpace, a way of creating a “space for health information” akin to the way that Tim Berners-Lee created the Web as a “space within information could exist.”
For example, a web user can drag a book’s URL from Amazon to Twitter, press send, and just assume that anyone, anywhere, and on any web-enabled device would have “interoperable” access to it. We don’t need an interoperability agreement between Amazon and Twitter, and if I want to pass the information through Facebook or Gmail, that’s easily done. I don’t have to re-engineer the whole system if I want to use a different routing, nor do I have to wait for some standards committee, government agency, or vendor to come up with the perfect standard for defining book information exchange.
The web created a large-scale, fine-grained network that used surprisingly few “moving parts” to do an amazingly large amount of information processing. I’d like to do the same for health care. This would also solve many of the political problems facing VA-DoD sharing. The same information could be shared as a “flat” information space, but different agencies could superimpose their hierarchies or constraints on it as they see fit. The agencies are not giving away the “family jewels” but are rather being given greater control over their information.
For example, blood pressure measurement may seem like a fairly benign piece of information. It might come from a VA clinic, a WalMart convenience clinic, or a home smart phone gadget. However, if it is a Navy Seal located in some remote mountain village, this puts the information in an entirely different context. The metadata about the blood pressure measurement – the time, location, etc. is hugely different than the WalMart reading on a Vet.
The information space model would allow the Navy to place restrictions on this information – from compartmentalizing it entirely, to applying whatever protocol they choose for that class of information.
Part of Tim’s design genius in creating the web was allowing it to be broken – the “404 not found error.” Prior efforts (such as Doug Engelbart’s Hypertext system) required bidirectional referential integrity: If A pointed to B, then B must also point to A. Of course, in the best of all worlds, this would be preferable. But in the real world of a dynamic, constantly changing world wide web, the 404 error was a key design decision to allow robustness in the design of the web.
As an aside, Tim played an interesting role in the creation of My Health eVet system. In the very early days of the web (1996?) , I had arranged a meeting with Rob Kolodner, Clayton Curtis, and others from VA to meet with Tim, Peter Solovitz from MIT, and Zak Kohane from Harvard. Rob Kolodner credits this meeting as the initial stimulus for the Health eVet program.
I think that your question about decoupling data from the organization is a very timely and important one – which could lead to a breakthrough in VA/DoD sharing efforts. I would be delighted to help you explore the issue.