Trends in Google internet searches for norovirus symptoms strongly correlate with rates of norovirus infection, suggesting internet searches could serve as reliable surveillance tools for diseases prone to seasonal variations, according to a study conducted by researchers at the U.S. Centers for Disease Control and Prevention and Google Labs in Tel-Aviv, Israel. The study, published in the July 2012 edition of Clinical Infectious Diseases, tracked rises and falls in Google internet searches for certain keywords and phrases related to gastroenteritis, such as “diarrhea,” “vomiting,” and “stomach virus,” that could indicate a norovirus infection. Norovirus is the most common cause of gastroenteritis in the U.S., responsible for an estimated 21 million illnesses a year. The researchers then compared the Google data to trends in lab-confirmed norovirus infections around the nation. If the searches matched up with the rates of confirmed cases, the researchers would uncover a new method of tracking the activity of norovirus, a pathogen with very scarce amounts of surveillance up until now. The result? The searches and the known cases matched up almost perfectly. “I think we were surprised at just how strong the correlation was,” said Benjamin Lopman, epidemiologist at the Division of Viral Diseases at the CDC and one of the study’s co-authors. “Even having to rely on general search terms, they still match up very well with the actual outcome.” Each year, norovirus infections see a rise in October, November and December before peaking in January or February. Looking at national search data for four years from 2007 to 2010, the researchers found that searches and infections peaked in the same month twice and followed very similar patterns from year to year. They also tested their theory at a smaller scale, analyzing gastroenteritis searches and norovirus illnesses in Boston, Massachusetts from 2006 to 2011. Again, the data showed significant correlation. “We have, potentially, a timely surveillance tool that can be used on various levels of public health surveillance,” Lopman told Food Safety News. “Not only can we look at activity on a national level, we can also look at it on a local level in a way we couldn’t with lab data.” Because the timing and magnitude of norovirus infections varies from year to year, public health officials have few methods at their disposal to monitor norovirus activity fast enough to react. Knowing that internet searches could provide insight into the pathogen’s infection rates means clinics and health officials can react accordingly. Similar techniques have already been put to use when looking at influenza infection rates. But the method isn’t perfect. As Lopman pointed out, media coverage of major illness events — such as the case of H1N1 a few years ago — can disrupt the technique’s reliability, as a far larger number of people begin searching about the event than just those actually affected by it. Lopman added that health professionals could likely use the internet search technique for most pathogens prone to seasonal variation, such as many of the most notorious foodborne illness-related pathogens such as E. coli and Salmonella, which typically see increased infection rates in the summer. Considering an estimated 90 percent of those who suffer from a norovirus infection never seek medical attention, Lopman said he hopes that in coming years the method might provide public health with an economical, instantaneous approximation of infection rates among any given location when more traditional methods are not a possibility.
From the paper: (Panel A) Norovirus Google search data (blue) for the entire US population were compared with estimated norovirus‐associated hospital discharges (red) from January 1, 2004 through June 30, 2007. (Panel B) Norovirus Google search data (blue) for 30 US states were compared with norovirus outbreak surveillance data (red) from January 1, 2007 through April 30, 2010.
From the paper: Norovirus Google search data (blue) for Boston compared with the proportion of llnesses in the Boston Public Health Commission syndromic surveillance system for gastrointestinal illness (red) from January 1, 2006 through July 2, 2011. [Paper link]