Since 2008, Google has been looking annually at influenza infection statistics around the world and creating models to estimate how serious flu outbreaks are in a wide swath of nations, including the United States.
But this year, Google is updating its flu data analysis methods for U.S. cases after last year’s 2012 to 2013 Google Flu Trends model overestimated the severity of flu cases that actually occurred in this country during that period, according to an Oct. 29 post by Christian Stefansen, a Google software engineer, on The Official Google.org Blog. The models look at the number of Web searches that are conducted by people seeking information about the flu, which Google says are good indicators of flu levels.
“When people get sick, they turn to the Web for information,” wrote Stefansen. That connection is what inspired Google, through its Google.org philanthropic arm, to begin compiling its Google Flu Trends (GFT) studies each year, using real-time, aggregated Google search data, in regions around the world, he wrote.
The overestimated number of flu cases described by the Google Flu Trends data in January 2013 were discovered after the estimates were compared to the numbers of actual health care visits for influenza-like illnesses (ILI) reported by the Centers for Disease Control (CDC), which were much lower, wrote Stefansen. The actual CDC numbers for the week of Jan. 13, 2013, showed an estimated flu incidence of 4.52 percent, compared with a Google Flu Trends estimated flu incidence of 10.56 percent, according to a related report by Google.org.
After studying the discrepancies, Google.org experts theorized that the reason for the different CDC and GFT estimates was that “heightened media coverage on the severity of the flu season resulted in an extended period in which users were searching for terms we’ve identified as correlated with flu levels,” wrote Stefansen. “In early 2013, we saw more flu-related searches in the U.S. than ever before.”
To correct those inaccurate, higher estimates, the GFT team is improving the model by using peak estimates from the 2012 to 2013 season, which “provided a close approximation of flu activity for recent seasons,” wrote Stefansen. “We will be applying this update to the U.S. flu level estimates for the 2013-2014 flu season, starting from August 1st. A casual observer will see that the new model forecasts a lower flu level than last year’s model did at a similar time in the season. We believe the new model more closely approximates CDC data.”
The Google Flu Trends reports “can help estimate the start, peak, and duration of each flu season—all important information for public health agencies,” wrote Stefansen. “This is an iterative process. We will keep exploring how we can build resilience to accommodate the effect of news media. In the meantime, stay healthy!”
The data used in the GFT reports is gathered using IP address information from Google server logs to make a best guess about where queries originated, according to Google. The flu search estimates include search data collected from more than 25 countries, including Australia, Belgium, Canada, Germany, Japan, New Zealand, Poland, Spain, Sweden and Switzerland, in addition to the United States.
In the Northern Hemisphere, the flu season typically spans from November to March, according to Google. In the Southern Hemisphere, the flu season typically spans from May to September. In tropical countries, a strong seasonal pattern may not exist.
In September, Google announced it is launching a health company, called Calico, to fight human aging and disease. Calico will work to find ways to improve the health and extend the lives of human beings, according to Google. Much of the details behind the new operation, however, have not yet been announced, including just what that goal means and how Google will take on its mission in these areas.
Calico wasn’t the first health care-related initiative undertaken by Google. Back in 2008, Google launched its Google Health initiative, which aimed to help patients access their personal health records no matter where they were, from any computing device, through a secure portal hosted by Google and its partners, according to earlier eWEEK reports. Google Health eventually shut down in January 2013.