News Feature | April 24, 2015

Google Trends Tracks Diseases

Katie Wike

By Katie Wike, contributing writer

An analysis of internet searches could help researchers determine disease prevalence, according to a recent study.

Data collected from web searches could fuel Google’s disease tracking machine. According to a study published in Journal of Epidemiology & Community Health, web search activity for non-communicable diseases, or NCDs, “has potential to provide real-time information on population risk during policy implementation and other population-level NCD prevention efforts.”

iHealth Beat explains researchers led by Svetha Venkatesh, director of the Center for Pattern Recognition and Data Analytics at Deakin University in Geelong, Australia, used Google Trends to study certain search terms over a one year period. This data was then compared to CDC data on risk factors associated with the diseases. These included cardiovascular disease, diabetes, exercise frequency, high blood pressure diagnosis, and tobacco use.

Reuters reports web search activity can often predict behavior. For example, a person who visits the gym may search for the hours. Or, a person who searches for the number of a restaurant may order takeout. “A diagnosis or (being) suspicious of heart problems is associated with searching for symptoms, side effects and so on,” Venkatesh said in an email to Reuters Health.

Researchers discovered their results were strongly tied to disease risk estimates from the CDC. iHealth Beat summarized the results when it came to predicting diabetes prevalence:

  • 11.2 percent in Alabama, compared with CDC's 11.8 percent estimate
  • 8.1 percent in Nevada, compared with CDC's 10.3 percent estimate
  • 9.4 percent in New Jersey, compared with CDC's 8.8 percent estimate

“These may provide a means to understand the response to changes in policy or other interventions in close to real time,” Venkatesh said.

“I’m not surprised that they found an association between search behavior and chronic disease prevalence and it might have some value in predicting prevalence,” said Diane T. Finegood, who leads the Chronic Disease Systems Modeling Lab at Simon Fraser University in Burnaby, Canada. But, she believes, it may not yet have direct policy making applications.