Wikipedia Reliable in Tracking Influenza Attacks in the US: Study

The number of hits on Influenza symptoms and related topics in Wikipedia helps accurately estimate the severity of the flu outbreak in the U.S., according to a study.

This way of tracking countrywide disease-outbreak is precise than conventional methods and data records of surveillance systems like the Centers for Disease Control and Prevention (CDC) which lag behind by two weeks. These agencies take a long time to gather facts and update the information from researches every week that are published at least a day or two later.

Experts from the Boston's Children's Hospital and Harvard Medical School observed the internet traffic seen on certain influenza related articles in Wikipedia during six flu seasons between December 2007 and August 2013. They created a model that supervised Wiki searches on Influenza-like illness in the country and compared its exactness with CDC's statistical data reports and Google Flu Trends (GFT).

It was found that Wikipedia browsing was extremely high in the 2009 H1N1 pandemic case and as well as the 2012 and 2013 influenza attacks. These had 17 percent accuracy in estimating peak weeks with surging flu outbreak than the data by the GFT that seemed faulty in a number of ways. The Google Flu trends misreported the numbers in seasons that received exaggerated media attention. In addition this method was found far more reliable than any traditional surveillance system.

David McIver, study author and postdoctoral fellow at Boston Children's Hospital said, "We were able to get really nice estimates of what the [flu] level is in the population," reports Live Science.

The authors believe each time there is a widespread attack or breaking news about a dangerous virus, people begin searching for news articles in the internet just to be informed. But their system of tracking only noted the number of Wikipedia articles that were browsed which, "to act as markers for general background-level activity of normal usage of Wikipedia,"

But, the study suggests these internet-based monitoring can be served as information that is parallel to the data estimates by sources like CDC.

The researchers add, "We are not trying to create something that will replace the CDC or anything like that. Our goal is to get both things to work well together, to give us a more holistic view of what is going on," reports the Live Science.

More information is available online in the journal PlOS's Computational Biology.