Analyzing tweets could help identify public health trends for influenza, depression

Tweeting could one day be used by public health officials to predict the spread of influenza in populations. Research, published in EPJ Data Science, conducted at the U.S. Department of Energy's Pacific Northwest National Laboratory (PNNL), outlined how future public health workers could utilize social media to identify trends of influenza or other health issues.

Using social media platforms to identify trends in health issues is more difficult than searching for messages such as “I feel sick.” In this study, researchers outline how to pinpoint patterns in social media behaviors to accurately identify rises in influenza, depression or other health issues based on area.

"Opinions and emotions are present in every tweet, regardless of whether the user is talking about their health," said Svitlana Volkova, a data scientist at PNNL and lead author of the study. "Like a digital heartbeat, we're finding how changes in this behavior relate to health trends in a community."

Currently, public health officials take weeks to discover a trend in influenza by calculating the number of people coming into the clinic. By using social media to examine the trends in health patterns of a certain area researchers hope to improve public health tools to be more efficient in catching onto a disease before it spreads. The study analyzed 171 million tweets from U.S. military members to determine opinions and emotions to test if they could relate to physician visits for influenza-like illnesses.

The tweets were found to be good indicators of how people behave according to the world around them. Researchers noted that military populations had more negative opinions with increased emotions of sadness, fear, disgust and anger than other populations. By identifying these locational patterns, researcher next turn their focus on whether these trends can be used to identify public health trends.