Quotes of the Day

Back in February 2003, if you subscribed to the ProMED e-mail list — a clearinghouse for intelligence about infectious-disease outbreaks — you might have seen quizzical messages about a strange spate of respiratory disease cropping up in the southern Chinese city of Guangzhou. Those messages — some of which were translated versions of news items appearing in the mainland-Chinese press before government censors stepped in — were the first public descriptions of what would later be known as severe acute respiratory syndrome or SARS.

In just a few months SARS would escape from southern China and spread — thanks to international jet travel — to Singapore, Canada, Taiwan and Europe, infecting more than 8,000 people and killing more than 900 in the first serious pandemic of the 21st century. Before the Centers for Disease Control and Prevention (CDC) or the World Health Organization knew anything about the new disease, the information was out there on the wire — and if the right people had known earlier, there's a chance the outbreak could have been halted before it became a pandemic.

A century ago — before jet planes — it took new diseases months to travel around the world, and many pathogens probably never made it out of their isolated rural stomping grounds. But now no place is truly isolated, no matter how remote. We live in a world that's more connected than ever before, one where humans — and the viruses hitchhiking inside us — can circle the planet in a day. As a result, we're at greater risk from new infectious diseases than ever before.

But there's an upside to our interconnectedness as well. Thanks to the Internet and cell phones, we can know what's happening in nearly every corner of the globe almost instantaneously — and that's a boon for epidemiology. In the arms race between us and the viruses, communication is our advantage. By analyzing the Internet's everyday wealth of data, we can catch new diseases before they've emerged — and stop them before they become a deadly threat.

That's what John Brownstein, a digital epidemiologist at Children's Hospital Boston, is working to do with his HealthMap project. HealthMap automatically trolls news sites, eyewitness reports, government data, even wildlife-disease cases to identify new patterns in outbreaks, presenting the results on a clickable map. Want to know about an ongoing polio outbreak in Angola? HealthMap will show you where it's occurring and who's dying. "We bring in all the cases we can on one site, which enables early detection and heightened awareness of new diseases," says Brownstein, who developed the site with software expert Clark Freifeld. "This is a way for epidemiology to move forward."

HealthMap first launched about five years ago, but it has just relaunched with a new focus toward what Brownstein calls "participatory epidemiology." HealthMap will tap the wealth of potential information on social media — think tweets about flu outbreaks and Facebook postings about contaminated food. The result is more finely tuned intelligence about emerging outbreaks, presented in a personalized format — Facebook by way of the CDC. HealthMap already has a related mobile app called Outbreaks Near Me, which gives users news about public health around their location — and allows them to report information as well. "It's really taking the local-weather-forecast idea and making it applicable to disease," says Brownstein. "We're trying to make these ideas that much more relevant to the general population."

The ideas behind digital and participatory epidemiology have merit. Scientists have already managed to find spikes in seasonal flu before the CDC by analyzing online search queries and Twitter feeds for flu-related items. (If an unusually large number of people in Pittsburgh are suddenly searching for influenza symptoms and treatment, chances are an outbreak is already under way.) Brownstein himself has worked with Google and the CDC to use the same methods to track outbreaks of mosquito-borne dengue fever, which has recently reappeared in the U.S. The neat project Bio.Diaspora, at the University of Toronto, combines intelligence on outbreaks from HealthMap with real-time information on international travel. And Global Viral Forecasting Initiative — a group I spent time with recently in Cameroon — is working to get on-the-ground intelligence in disease hot spots like the Congo and China.

The challenge with HealthMap and other digital epidemiology projects is the same one that all intelligence experts face: separating the signal from the noise. Brownstein points out that HealthMap could show unusual cases of respiratory illness in Mexico in the early spring of 2009, before what would become the H1N1 flu pandemic burst onto the global stage, but it's still difficult to separate truly dangerous events from run-of-the mill outbreaks. The hope is that sharper data collection will allow future digital epidemiologists to identify the patterns that indicate a potentially deadly new disease in time to actually do something to stop it. "We're trying to improve our algorithms to make sure that we catch the really important events," says Brownstein.

One way to do that might be to get more people participating in participatory epidemiology. To that end, HealthMap is part of a public-education campaign built around the upcoming disease thriller Contagion. The Steven Soderbergh–directed film — which opens in U.S. theaters on Sept. 9 and stars Gwyneth Paltrow, Matt Damon and Jude Law, among others — tells the story of a deadly pandemic that spreads around the world. (Imagine H1N1 if it had killed 20% of those it infected, instead of at most 1%.) "We want to get people talking about this threat, so they can take it seriously without being scared," says Brownstein. Fighting new infectious diseases is no different than any other war — the first step is good intelligence.

• Bryan Walsh
• By mining the Internet's massive volume of data, digital epidemiologists hope to identify disease outbreaks in real time -- and eliminate deadly epidemics before they start