While it is nice to tout a system that "picks up" disease outbreaks "some of them before WHO or CDC", it is important to remember that:
- this system does not DETECT outbreaks (nor does it claim to): it presents a map of already reported outbreaks
- no disease outbreak EVER ever been initially detected by an automated system before an alert doctor or other healthcare provider: a human has ALWAYS been the "sensor" that detects disease outbreaks on the ground.
- the WHO and the CDC are not SUPPOSED to identify every disease outbreak in the world, just like a building maintenance staff isn't supposed to identify every coffee spill in the building: most coffee spills can be handled locally, and the maintenance staff never needs to be notified at all. Likewise with disease outbreaks.
It is very interesting, and sometimes useful in generating hypotheses, to view a map of news coverage of outbreaks, but a system like this can never, ever take the place of alert clinicians who, with brain computing power that dwarfs the fastest processor Google can field, are able to spot trends and unusual cases better than any computerized system.
Better to also spend money improving the education and capabilities of those clinicians, and improving the systems by which they can notify a larger world about their unusual cases -- maybe those notifications could feed into a system like the one described and finally give it the ability to truly detect outbreaks.
I find it amazing that people could laud the OLPC "initiative" for wasting millions of dollars in order to . . . get a cheaper laptop. Funny, I thought that's what the market does, year after year after year. I live in a world in which consumer electronics get cheaper every year without any work from Nick Negroponte, but maybe that's different from the world most of the posters at this blog live in.
Funnily enough, while OLPC has worked hard to produce a laptop that finally costs about $175 plus shipping -- and that so few developing countries actually want to buy that now they want US to pay for it -- you can buy a nice Fedora Linux laptop, the Madison Celebrity, from Sweden for $150 including free shipping in the US.
The point is that OLPC could have spent the last few years developing kick-ass SOFTWARE to run on the $100 linux boxes that we'll have by next year ANYWAY instead of wasting money and time promoting themselves and performing the magic act of creating a laptop that costs $25 more than the market created without them.
I'm the developer of EpiSurveyor. While we've had excellent results in the field with PDAs for almost ten years now, working with the American Red Cross, UNICEF, WHO, and many other organizations, we're now shifting from PDAs (ie, unconnected pocket computers) to cell phones (ie, connected pocket computers). Cell phones, as you may know, are rapidly spreading across the developing world: just about every health worker we come across already has one. I believe that before the end of the year we'll have a version of EpiSurveyor that runs on J2ME platforms. Keep an eye out for updates at www.datadyne.org
Slashdot Mirror
- this system does not DETECT outbreaks (nor does it claim to): it presents a map of already reported outbreaks
- no disease outbreak EVER ever been initially detected by an automated system before an alert doctor or other healthcare provider: a human has ALWAYS been the "sensor" that detects disease outbreaks on the ground.
- the WHO and the CDC are not SUPPOSED to identify every disease outbreak in the world, just like a building maintenance staff isn't supposed to identify every coffee spill in the building: most coffee spills can be handled locally, and the maintenance staff never needs to be notified at all. Likewise with disease outbreaks.
It is very interesting, and sometimes useful in generating hypotheses, to view a map of news coverage of outbreaks, but a system like this can never, ever take the place of alert clinicians who, with brain computing power that dwarfs the fastest processor Google can field, are able to spot trends and unusual cases better than any computerized system.
Better to also spend money improving the education and capabilities of those clinicians, and improving the systems by which they can notify a larger world about their unusual cases -- maybe those notifications could feed into a system like the one described and finally give it the ability to truly detect outbreaks.
Joel Selanikio, MD
DataDyne.org
Washington, DC
I find it amazing that people could laud the OLPC "initiative" for wasting millions of dollars in order to . . . get a cheaper laptop. Funny, I thought that's what the market does, year after year after year. I live in a world in which consumer electronics get cheaper every year without any work from Nick Negroponte, but maybe that's different from the world most of the posters at this blog live in.
Funnily enough, while OLPC has worked hard to produce a laptop that finally costs about $175 plus shipping -- and that so few developing countries actually want to buy that now they want US to pay for it -- you can buy a nice Fedora Linux laptop, the Madison Celebrity, from Sweden for $150 including free shipping in the US.
The point is that OLPC could have spent the last few years developing kick-ass SOFTWARE to run on the $100 linux boxes that we'll have by next year ANYWAY instead of wasting money and time promoting themselves and performing the magic act of creating a laptop that costs $25 more than the market created without them.
Hi Josef,
I'm the developer of EpiSurveyor. While we've had excellent results in the field with PDAs for almost ten years now, working with the American Red Cross, UNICEF, WHO, and many other organizations, we're now shifting from PDAs (ie, unconnected pocket computers) to cell phones (ie, connected pocket computers). Cell phones, as you may know, are rapidly spreading across the developing world: just about every health worker we come across already has one. I believe that before the end of the year we'll have a version of EpiSurveyor that runs on J2ME platforms. Keep an eye out for updates at www.datadyne.org
Best,
Joel Selanikio