Monitoring social media websites like Twitter could help health officials and providers identify in real time severe medical outbreaks, allowing them to more efficiently direct resources and curb the spread of disease, according to a San Diego State University study published last month in the Journal of Medical Internet Research, Medical News Today reports.
For the study, lead researcher and San Diego State University geography professor Ming-Hsiang Tsou and his team used a program to monitor tweets that originated within a 17-mile radius of 11 cities. The program recorded details of tweets containing the words "flu" or "influenza," including:
- Whether the tweet was an original or a retweet; and
- Any links to websites in the tweet.
Researchers then compared their findings with regional data based on CDC's definition of influenza-like illness.
The program recorded data on 161,821 tweets that included the word "flu" and 6,174 tweets that included the word "influenza" between June 2012 and the beginning of December 2012.
According to the study, nine of the 11 cities exhibited a statistically significant correlation between an uptick in the number of tweets mentioning the keywords and regional outbreak reports. In five of the cities -- Denver, Fort Worth, Jacksonville, San Diego and Seattle -- the algorithm noted the outbreaks sooner than regional reports.
Tsou in a release said that the social media monitoring program detected outbreaks daily, while "[t]raditional procedures" typically "take at least two weeks."
Tsou and his research team next will focus on detecting outbreaks through correlations between data on influenza-like illnesses and specific symptom-related keywords, such as "cough," "sneeze," "congestion" and "sore throat" (Medical News Today, 11/21).