Healthcare Open Source and Community: Some Lessons from the Field by Paul Biondich
I’ve followed the evolution of Health 2.0 with much interest, often tracking the blogosphere chatter to
get a sense of how the community would realize the vision of “user-generated” health care. Perhaps my adventures in the developing world over the past few years might lend some perspectives to the ongoing discussion.
In early 2004, my good friend Burke Mamlin and I were asked to visit a HIV care system in western Kenya. This growing program’s leadership was interested in how we could apply our background in medical informatics towards evolving a scalable health information system in their resource-constrained environment. In case you didn’t know, there are a growing number of people like Burke and myself, doctors who both practice medicine and spend considerable time building and studying health information technologies (the Regenstrief Institute is an example of a medical informatics skunk works full of geeks like us). During that week in Eldoret, we realized two things: we felt compelled to do something, and their health information needs were vast. We were both experienced in developing large scale health IT, but knew that the amount of work required to realize their vision was beyond our means. So, instead of attempting to hire a bunch of developers and building an unsustainable effort, we reached out to others who cared for HIV patients, and encouraged a collaborative effort. We laid down our notions of the technical underpinnings as a straw man and encouraged others to critique it, so that it became a shared vision. We then built and implemented the software together as a community. This project became known as the Open Medical Record System (OpenMRS). We open-sourced all of the ideas and source code, and put real resources into mentoring others to take part, instead of doing the work ourselves. We wanted to facilitate “information independence” in these developing environments not by bringing fish, but by teaching how to fish. So far, so good: the software has over two dozen developers, and is being actively implemented in a dozen different countries.
During the last three to four years, we’ve been surprised as to how the project has evolved. Here are three lessons we’ve taken away from the experience that I think have great relevance to Health 2.0:
- Communities place great value in software cores as technical foundations.
There are no two implementations of OpenMRS that behave or look exactly the same, but they all share the same core architecture. In many way having that core software, even in an incomplete form, reduced the barriers to entry for others to participate and allowed each environment to build sharable plug-and-play customizations to meet specific needs. As I work in my day job, I often wonder if a similar approach in the rest of the world would empower increasing numbers to participate in the health information revolution. Many likely wonder how commercial entities would react to such an “open-sourced” approach. My gut instinct is that there’s a place in the ecosystem for a common, community-developed foundation which would allow commercial entities and garage-based startups alike to innovate at a higher level, both from a services and new feature perspective.
If you really get down to the pragmatics of “Web 2.0”, it’s really all about making it easy for people to do the right thing. In building foundational, core services for the web, real innovation evolved from end-users who have simply “mashed” these services together in fascinating ways. I submit that the same should occur as part of the Health 2.0 movement. Don’t be confused or misled by the rhetoric of the open source idealists who to date have focused mostly on the cause… I submit that open source is a very pragmatic and necessary foundation towards a Health 2.0 world.
- The crowd is wise when it comes to standards development.
In order for individuals to be empowered in their own care, medical school and residency training shouldn’t be a prerequisite. From my perspective, “user-generated” healthcare truly thrives once health information is represented and stored in a way that’s interpretable, computable, and compatible between differing systems. Information systems are well equipped to present individuals with an overview of what specific data means to their overall health through simple, low level processing, which is a potent enabler.
A foundational effort towards this vision is the need for society to evolve standard ways to communicate and represent medical content. Within the OpenMRS community, we’ve intentionally avoided trying to create standard notions of how to represent HIV care information (and healthcare information in general), instead creating a technical “commons” of sorts that gathers the differing notions of everyone and presents them back to the community in an optimized way. This lets the most common conventions rise to the top and become “de facto” standards. This is in stark contrast to the way vocabulary standards development organizations currently operate, and is more consistent with the vision of the “Wisdom of Crowds” and the revolutionary work of Wikipedia. If we can move towards the direction of standard information representation models as alluded to recently by Tim Berners-Lee in his WHIT keynote, “user-generated” healthcare becomes much more than a novel idea.
- Don’t underestimate the motivations of individuals.
When we began evolving our open source plans, we mistakenly believed that the only people that would be interested in getting involved with OpenMRS would be those that had some “skin in the game” and directly worked in patient care settings. What’s actually happened however is that people from all walks of life have approached the project and helped in ways that we wouldn’t have ever anticipated. We’ve had seasoned software developers who want to write code in their spare time but have no background in healthcare, other open source communities involved in an aspect of software development complementary to healthcare, and individuals who merely want to advocate or raise resources on our behalf. They all seem to share a very Health 2.0 vision: improving the health of others through efficient access to information. That sense of philanthropy which permeates through each of us shouldn’t be underestimated as a potent realizing force behind Health 2.0 efforts.
As I continue to become more and more involved in the wonderful efforts in Western Kenya and other developing countries, it’s becoming clear that the leading HIV care initiatives are beginning to focus on prevention instead of treatment. As Don Berwick has famously described, “Information is Care”. Getting in front of the rampant spread of the disease in these environments is only possible once individuals have access to health information. It is becoming an increasing focus of our efforts in OpenMRS, as a result. If you are personally interested in participating in our little microcosm of a Health 2.0 effort, one which will hopefully soon enable individuals with HIV and other life threatening illnesses in the developing world to be empowered in caring for themselves, you are welcome to come join us!
Who knows? Perhaps the product of this work could ultimately serve as a potent nidus towards realizing some of the visions espoused by Health 2.0 in other health care environments.