Google Flu Trends for the win

Update 3/29/14: 
Okay, not so much. In reality, Google Flu has been a bit of a disaster. It turns out that when people self diagnose themselves with flu, they're wrong at least 80% of the time. Which of course means that people who do Google searchers on influenza hardly ever actually have the flu. Oops.



Google Flu Trends (GFT) is a tool for predicting where and when the flu will strike, based on the number of people doing Google searches for flu-like terms and symptoms.  The number of flu searches correlates surprisingly well with the number of actual flu cases.  In fact, a new study led by Andrea Dugas of Johns Hopkins showed that GFT could predict emergency room visits.

Traditionally, doctors’ offices and hospitals rely on reports from the Centers for Disease Control (CDC) to prepare for influxes of flu patients.  Because these reports are based on actual hospital admissions and laboratory tests, by the time a community hears from the CDC about flu epidemics in their area, it may be too late.  In contrast, clinicians can access the data from GFT in real time.  A regional spike in interest in the flu tells local health care providers to prepare for incoming flu patients.

Like citizen science projects, GFT relies on the power of the community of internet users.  However, unlike other projects, participants don’t need to sign up for anything or get any kind of instruction.  Simply by using Google however they like, they end up advancing science and benefiting their neighbors.  It makes me feel all warm inside.