Tweeting in the time of cholera

Improving or maintaining access to clean drinking water is the best way to prevent a cholera epidemic. Twitter could prove an excellent way to help stop one. (Julien Harneis/Flickr)

It was after the devastating 2010 Haiti earthquake that mobile-friendly social media services like Twitter and Ushahidi came into their own as disaster management and relief tools. With the nation’s already unsteady infrastructure destroyed, these tools helped speed the deployment of people and supplies to where they were needed by giving relief workers on-the-ground intelligence about what was happening, what was needed and where in nearly real time.

With hindsight, Twitter and other informal data sources could also have sped up efforts to halt the spread of one of the disaster’s most feared aftermaths – cholera.

Looking at Twitter data coming out of Haiti after the earthquake, as well as cholera-related news reports flagged in the online surveillance tool HealthMap, John Brownstein, Rumi Chunara and their team at Children’s Hospital Boston’s Informatics Program (CHIP) found that the numbers of cases reported through these “informal” data sources closely matched official hospital and clinic-reported data in the growing epidemic’s first 100 days. Twitter and HealthMap activity even closely matched two spikes in the official case numbers from Haiti’s public health ministry: at the start of the outbreak, and after Hurricane Tomas brushed in early November 2010.

But more to the point, trends in the Twitter and HealthMap data were apparent two weeks ahead of official statistics from the public health ministry – potentially allowing an earlier response.

The study suggests that self-reported grassroots information can draw a picture of an emerging epidemic as it unfolds, ascertaining in hours what can take weeks for traditional data collection and reporting methods.

“The early findings demonstrate that social and news media could be a powerful way of gaining early insights into an epidemic’s behavior,” says Brownstein, who lead’s CHIP’s Computational Epidemiology Group.

The social data shed particularly strong light on what’s called the epidemic’s reproductive number: a measure of how the disease is spreading. “The spread of an epidemic is closely tied to the number of people who get infected from each case,” Brownstein explains. “It’s a key metric for deciding what control measures you’re going to put in place and where, for cholera and for other infectious diseases.

“The more quickly you can figure out this and other parameters,” he continues, “the more quickly you can start to respond with antibiotics, vaccines or other appropriate measures, depending on the disease in question.”

A map of cholera-related news reports in Haiti in the first year following the 2010 earthquake. (HealthMap)

Brownstein and Chunara’s work is part of a growing body of research showing the value of social media and other unconventional data sources for public health and clinical care. He has shown similar advantages in speed to using Google search data for tracking dengue. He is also hard at work on a new version of HealthMap’s Outbreaks Near Me app, which lets users submit their own disease reports.

With an estimated 6 billion cell phone subscriptions active worldwide at the end of last year (at the time of the earthquake, some 35 percent of Haitians had a cell phone), more than 200 million registered Twitter users globally, and a growing number of smartphone-enabled apps that let people learn about and report diseases on their own, the phone on your pocket – your personal connection to the Internet – could make you part of the solution – not just part of the epidemic.

[Ed. note: Interested in how digital tools can help detect and track infectious diseases? Check out HealthMap’s International Conference on Digital Disease Detection, February 16-17, 2012, at Harvard Medical School’s Joseph B. Martin Conference Center.

Brownstein will also talk about HealthMap at Innovation Day, an event run by Children’s Innovation Acceleration Program (IAP) on February 14, 2012. For more information, including registration information, visit the IAP website.]