Thu, 12 Mar 2009
Today at ETech: Baobab Health Partnership
(I mostly live-blogged this; excuse the messiness.)
Mike McKay gave a fantastic presentation about his work at Baobab Health Partnership. To quote his blurb:
Malawi, Africa, has a population of 14 million. One million are HIV positive and there are just 280 doctors in the country.
Baobab Health Partnership took i-openers, added a PIC to create a touch-screen, and hacked on Power over Ethernet, and configured them to be used as data entry and analysis workstations for HIV clinics in Malawi.
The i-opener was an "Internet appliance" sold as a simple web terminal: given a monthly subscription for the i-opener service, you could have an inexpensive, trouble-free web browsing experience. As it happened, the community swiftly repurposed the devices. According to one i-opener hacker, "What follows is discussion and photographs of a most righteous hack, turning the nearly-free Netpliance i-opener web appliance into a full featured pseudolaptop / electronic photo album." Slashdot saw a lot of excitement around these in 2000 and 2001, but they lacked Ethernet or wifi, so they fell out of favor.
I wanted to give you that background so I can explain that the Baobab Health terminal is the coolest thing I have ever heard of anyone doing with an i-opener. This puts my kitchen recipe terminal plan to shame. Check out patient registration video that shows the terminals in use.
The software stack is all open source: They have an Ubuntu server running MySQL and Ruby on Rails, and the user interface is just an AJAX web application running in full-screen mode. It was developed by Malawaian developers who, says McKay, "were trained in VB. Now they hang out on IRC and flame people on mailing lists. They're part of the Internet."
As they bought a few from eBay for about thirty dollars, they noticed that one person in Nebraska seemed to keep listing them. It turned out that he had been stockpiling the until he figured out something great to do with them. In the end, the Nebraskan gave them two thousand i-openers. (The twentieth century rises again.)
As far as impact, McKay pointed out decision support makes a huge difference in healthcare, as Provonost found at Hopkins years ago with checklists. He also talked about how they data validation dramatically changed the clinics' ability to record useful data; most of the previous paper records were found to be useless as they began to import them into the computerized system. Through comparisons to other people of similar age and gender, they can encourage correct weight measurement data entry. Danny O'Brien pointed out that this does not degrade well in the case of equipment or network failure. I believe Danny was hinting at the question of what happens when the entire system fails (backups? data loss?); McKay understood it to refer to what happens during temporary failures. He explained that with their generators, they've had very good reliability. When the computers fail, many clinics take paper records instead, though some simply lock the doors and stop seeing patients. That clinics choose the latter approach has provided strong motivation for their programmers to fix reliability issues!
McKay went on an aside to talk about having his appendix removed when he was in the States. "All of a sudden, I was exposed to the American health care system, and I was shocked at how broken it is. There's no computers anywhere! The only computers around are at billing."
To get to electronic medical records in the US, he says, "Maybe we just need a disruptive piece of technology."
As for how to spread the program to other countries or circumstances, McKay explained, "One of our problems is the next generation of hardware. I'm not going to send this hacked i-opener to a country where there aren't technicians who know what to do with them."
The program was started by Gerry Douglas in 2000. Douglas is now based in Pittsburgh, though he spends four to five months of the year in Malawi. Also, the data collected by the project inform policy discussions at a quarterly HIV Forum in Malawi. A man from the NIH suggested that they work on sharing that data and doing more research.
The presentation was fantastic; the work is brilliant; and the man is friendly and thoughtful.
