I thought Gunnar Hellekson’s post on the recent RFP from the US Veteran’s Administration (VA) to open source their VistA electronic healthcare record (EHR) system was a great overview of the challenges facing massive government agencies as they attempt to adopt an open source approach to their mission. I definitely agree with the list of challenges and issues that Gunnar has laid out for the proposal. I further agree with Ben Mehling of Medsphere that the exciting thing is that “…it’s no longer a question of IF, but rather WHEN…” the VA is going to open source VistA.
As a tax payer, citizen and frequent user of healthcare services, I am terribly excited about the idea of a truly open source community creating a free platform for EHR for use around the world. Certainly we could use something like this here in Canada. The ability of open source to provide a worldwide, vendor-neutral software platform that an ecosystem can thrive, compete and innovate upon is unparalleled. Given the dire straits that virtually all modern healthcare systems find themselves in, the potential upside implicit in what the VA is proposing is an incredible opportunity. You can see a good review of the potential in this paper from Carnegie Mellon funded by the VA as it did its due diligence.
The RFP’s main objective is the creation of a “Custodial Agent” to act as the governing body for a VistA community. The interesting challenge ahead for the VA as they create such a body will be to balance their impulse to control, relative to the need for community. The notion of “Control versus Community” is well documented in Matt Aslett’s work from the 451 Group, a nice summary of which has been provided by Henrik Ingo. And frankly, I am very worried that there are a long list of impediments which could cause this initiative to fail.
We all know that governments are by their very nature conservative, and primarily serviced by firms which specialize in knowing how to meet the specific requirements of their complex procurement and management processes. And despite the best of intentions, I struggle to believe that the people and institutions behind this initiative will be successful in establishing a truly vendor-neutral, open, transparent and meritocratic community which can deliver on the vision. Not because of lack of desire, but because of lack of knowledge and the specific skillsets of experienced community development people. You know: the folks who’ve been there, done that and “get it”. I worry that what will come out of this process will be what Matt Aslett called “Stage 3 – Vendor-dominated open source development and distribution project”, except that the “Vendor” here will be the VA and a prime contractor. If that happens, I highly doubt that this initiative will be the success it deserves to be. Or perhaps I’m overly pessimistic, and this is simply a learning stage that the VA and its ecosystem will need to go through. After all, lots of very smart companies made the same errors early in their march towards openness.
To me, the logical solution to VA’s search for a governing body would be to take VistA to a community which already exists, and which already has a clear community-based governance model in place. Reading the RFP, I could not help but think that at least three quarters of the required responses could simply be links to already existing governance, process and licensing documents already here at the Eclipse Foundation. And most of the remaining pieces could be implemented as an industry working group. Too bad VistA is implemented using MUMPS, which is not exactly a mainstream Eclipse development platform.
Certainly our good friends and former colleagues over at OpenHealthTools.org would be a logical fit as well. They have been working in the healthcare domain with a view towards creating an Eclipse-style community for years. And no one knows ecosystem development better than their Executive Director, Skip McGaughey. Surely one of the vendors interested in responding to this will do the sensible thing and enlist their support in their submission?
So this will be a very interesting space to watch. It is incredibly daring for an organization the size of the VA to embrace open source as a strategy. This is a long overdue initiative, as the current mess of software systems and vendors has created the very definition of a gridlock economy in the application of IT to healthcare. I’m impressed with the process that led to the RFP. My fingers are crossed that the stars align to deliver a truly open platform for innovation in healthcare.
