Following dengue fever in Brazil with Twitter

New software that filters tweets containing "dengue" and the user's location can help authorities stay current with the locations of outbreaks for faster responses.
Written by Janet Fang, Contributor

When dengue season begins in Brazil this November, social networking could help the country better control outbreaks by tracking the disease’s spread down to individual cities. New Scientist reports.

According to the World Health Organization, about two-fifths of the world’s population are at risk of dengue. The mosquito-borne disease kills hundreds of people a year.

A collaboration between two Brazilian National Institutes of Science and Technology, led by Wagner Meira of the Federal University of Minus Gerais, created software to identify a tight correlation between the time and place where people tweet they have dengue and the official statistics for where the disease appears each season.

For example: “My mother is suspected of having dengue," tweets a woman in Rio de Janeiro. "I think I have dengue. Hopefully I'm wrong!" tweets a man in São Paulo, 350 kilometers away.

Dengue breaks out every year in Brazil, but exactly where varies every season. That means it can take weeks for health authorities to plan where to concentrate their resources.

With increased access to the internet in Brazil, using Twitter messages could mean a much faster response, Meira says. "It isn't predicting the future but the present," he says. "This means we aren't weeks behind like we used to be."

  1. The approach uses software to filter tweets containing the word "dengue" and information on the user's location.
  2. Tweets expressing personal experience of the disease are identified using sentence structure and wording; those mentioning public campaigns or telling jokes are all filtered out.

Testing the software on 2,447 tweets containing "dengue" and a location sent between January and May 2009 shows that ‘personal experience tweets’ was highly correlated with outbreaks identified by the Brazilian Ministry of Health.

  • To further investigate how tweets mirror trends seen in cities where changes in the outbreak are known to have occurred, the team plans to analyze 181,845 tweets sent between December 2010 and April 2011, after the ministry's 2011 data come in.
  • Terms such as "bone pain" and "eye pain” (typical symptoms of dengue fever) are now being included in software.

This isn’t a first for Twitter. It was used for real-time surveillance of the 2009 swine flu pandemic, for example. This is, however, the first time it’s used to follow dengue, and the first time data on the scale of individual cities has been collected like this. It'll be complemented nicely by Google Dengue Trends, unveiled last month, which records spikes in web searches for dengue fever.

The work [pdf] was presented at the Web Science Conference in Koblenz, Germany, last month.

Via New Scientist.

Image: Aedes aegypti from CDC / James Gathany

This post was originally published on Smartplanet.com

Editorial standards