Data Mining Predicted Swine Flu 18 Days before WHO

Dennis Faas's picture

According to reports, Washington state surveillance firm Veratect Corp. raised the first warning about a possible swine flu outbreak in Mexico more than two weeks before the World Health Organization (WHO) made its own announcement. (Source: pnwlocalnews.com)

It also appears as that federal and international health officials had access to the Veratect Corp. warning. Emails relating to the possible outbreak spreading in Mexico were sent to 10 officials of the Centers for Disease Control (CDC) and prevention. It's not known why health officials didn't act sooner than they did.

Data Mining Technique Detected Outbreak

Kirkland, Washington-based Veratect uses a data mining technique to automatically search tens of thousands of web sites daily for early signs of looming medical problems or civil unrest anywhere in the world.

When items of interest are found, the results are turned over to a team of 35 analysts who post legitimate threats on the company's web site. Government agencies, businesses and other groups have access to the Veratect web site.

Veratect has been unsuccessful in trying to sell its services to the CDC, the WHO and the Department of Homeland Security, though the CDC and the WHO were given access to the Veratect web site on a trial basis earlier this year.

Flu Detected Weeks Before WHO's Announcement

18 days before WHO issued the alert about a possible Swine flu pandemic, Veratect reported on a strange outbreak of respiratory disease in La Gloria, Mexico, noting that local residents thought the flu outbreak was related to contamination from pig breeding farms nearby. (Source: mcclatchydc.com)

The information was available to the CDC as well as many state and local health authorities. Records show that an epidemiologist at the Pan American Health Organization -- part of the WHO -- viewed the message about the La Gloria outbreak on April 10 and on April 11.

An automated email was sent to 10 people at the CDC on April 16 to notify them that a disease with an 'unspecified number of atypical pneumonia cases' was detected at a hospital in Oaxaca.

Data Mining Technique for Detecting Outbreaks Untested

When there was evidence of the outbreak spreading, Veratect's chief scientist called people he knew at the CDC's emergency operations center on April 20 to make them aware of what was happening in Mexico.

During that time, the CDC was focusing its attention on possible swine flu events happening in Texas and California. Apparently the CDC was not aware that the outbreak was spreading in Mexico. Four days later, WHO made its announcement about the outbreak.

Veratect hopes to be part of an effective global health monitoring system that is badly needed. The use of data mining to track possible disease outbreaks is, however, currently unfounded. (Source: mcclatchydc.com)

Visit Bill's Links and More for more great tips, just like this one!

Rate this article: 
No votes yet