Hopkins analyzes ‘viral’ tweets to track flu

Have you ever tweeted about having the flu? While such a message will certainly gain sympathy from friends and family, it may be even more useful than you think. New research is attempting to use Twitter to track the flu throughout the country.

This tracking, carried out by a program developed by a Hopkins team a few years ago, relies on the categorization of flu tweets: After collecting all of the flu-related tweets, the program differentiates between those from individuals with the flu and those simply from Twitter users interested in the illness or how it affects other people. The algorithm used in this flu-tracking technology can analyze an astounding 5,000 tweets every minute.

The recent research largely focused on the behaviors of individuals with the flu. Specifically, the researchers paid attention to behavior regarding medications and whether sufferers of the flu were improperly taking antibiotics.

While this research regarding antibiotic use is highly important, especially in the face of increased antibiotic resistance, a team of researchers at Hopkins and George Washington University are taking the investigation along a new path.

This change was prompted by the realization that the Hopkins-developed technology could focus the national scope of Twitter to a local area. This ability to track local tweets is an important tool for public information because national tracking may not always convey the health of certain regions of the country. Knowledge about the prevalence of the flu in certain regions can allow health officials to make more precise preparations. With localized information, they can properly distribute staff to better accommodate an increase in sick patients.

This information may also help focus vaccination efforts: healthy individuals living in areas with an increasing number of flu cases can be alerted about their increased risk of contracting the illness and then prompted to get the season’s flu vaccine.

The researchers, focusing on data from New York City, realized that they could predict trends of the virus in the city as well as they could on a national scale. The localization of the tweets was possible because many individuals allow their cell phones to record their particular location using GPS technology.

It turns out that this tweet-tracking technology is extremely accurate. The team of researchers compared their data to the statistics about flu incidence and prevalence collected by the U.S. Centers for Disease Control and Prevention and the New York City Department of Health and Mental Hygiene. They found that an increase in flu-related tweets correlated to an increase in the governmental reports of flu cases. Twitter has blossomed since its creation in 2006. This recent research proves that its influence has reached the field of public health.