Slashdot Mirror
Online Tool Flagged Ebola Outbreak Before Formal WHO Announcement
Taco Cowboy (5327) writes Nine days before the announcement from WHO regarding the Ebola outbreak in West Africa, an online tool had the incident flagged. HealthMap, a team of 45 researchers, epidemiologists and software developers at Boston Children's Hospital founded in 2006, hosting an online tool that uses algorithms to scour tens of thousands of social media sites, local news, government websites, infectious-disease physicians' social networks and other sources to detect and track disease outbreaks. Sophisticated software filters irrelevant data, classifies the relevant information, identifies diseases and maps their locations with the help of experts. The tool was introduced in 2006 with a core audience of public health specialists, but that changed as the system evolved and the public became increasingly hungry for information during the swine flu pandemic. To get a feel of how HealthMap works, in the case of the Ebola outbreak, visit the site.
If it can make these kinds of predictions without a tonne of false positives, then we have something we can call a tool - otherwise it's just a more efficient but no more reliable form of gossip and rumour.
Go to the site. Click to the head of the timeline. Look:
So, if you label the "mystery hemorrhagic fever" as ebola, after the fact or without waiting for confirmatory tests, you too can beat the WHO by 9 days.
If you ignore that the WHO's detection regime is the one that has doctors and hospitals sending samples laboratories for confirmatory testing, you too can beat the WHO by 9 days.
If your algorithm identifies dengue fever as ebola based upon "tens of thousands of social media sites, local news, government websites, infectious-disease physicians' social networks and other sources," keep quiet about the fact. Announce your success four months after everyone is sure that it is what you think it is to avoid embarrassing press releases.
This does not appear to be early epidemiological detection by connecting the social-media-dots. This is jumping-the-gun based on early reporting of the processes of an existing early detection program.
TFA doesn't make this clear which WHO announcement this tool is being compared to, which makes it really hard to judge the effectiveness of HealthMap.
The WHO declared the Ebola outbreak a Public Health Emergency of International Concern only 2 days ago on August 8th. However I am not aware - nor can I find - any record of the WHO declaring an epidemic, as TFA states. (Does the WHO even declare epidemics?)
If HealthMap is being compared to the PHEIC announcement, then for all practical purposes its useless as this outbreak has been going on for some number of weeks now. More likely HealthMap is being compared to an earlier WHO announcement, but without knowing exactly when that is, there's no way to tell if the HealthMap analysis would have actually been of any use.
I see you've managed to reverse the polarity of the humor flow...
Ezekiel 23:20
One successful detection of an outbreak is meaningless. This is like how everybody claimed frogs had predicted the earthquake in China a few years ago.
To judge the success of this system, we need three pieces of information:
* How many outbreaks has this system actually flagged?
* How many outbreaks has this system missed (false negatives)?
* How many outbreaks has this system flagged that turned out not to be (false positives)?