Researchers at Boston Children’s Hospital, USA, have developed a method of estimating levels of influenza-like illness in the American population by analyzing Internet traffic on specific flu-related Wikipedia articles.
David McIver and John Brownstein’s model, published recently in PLOS Computational Biology, estimates flu levels in the American population up to two weeks sooner than data from the Centers for Disease Control and Prevention becomes available, and accurately estimates the week of peak influenza activity 17% more often than Google Flu Trends data.
McIver and Brownstein calculated the number of times certain Wikipedia articles were accessed every day from December 2007 to August 2013. The model they developed performed well both through influenza seasons that are more severe than normal and through events such as the H1N1 pandemic in 2009 that received high levels of media attention.
“Each influenza season provides new challenges and uncertainties to both the public as well as the public health community. We’re hoping that with this new method of influenza monitoring, we can harness publicly available data to help people get accurate, near-real time information about the level of disease burden in the population.”
Following further validation, the model could be used as an automatic system to model flu levels in the USA, providing support for traditional influenza surveillance tools.
The research was funded by the National Institutes of Health and National Library of Medicine.
Read the study at PLOS Computational Biology: Wikipedia Usage Estimates Prevalence of Influenza-Like Illness in the United States in Near Real-Time.