How Twitter predicts ER visits
Tweeting about your health may no longer be considered over-sharing. New research from the University of Arizona found that health-related tweets, particularly for asthma, may help hospitals predict emergency room visits.
A team of researchers developed a model to predict approximately how many asthma sufferers would visit the emergency room at a hospital in Dallas based on data from medical records, air quality sensors, and tweets.
For three months, researchers collected air quality data from environmental sensors in the Dallas hospital area and tweets containing the keywords “asthma,” wheezing,” and “inhaler” that were traced to ZIP codes where most of the hospital’s patients lived.
Their findings, which will be published in IEEE Journal of Biomedical and Health Informatics’, found that as certain air quality factors worsened around the Dallas hospital, asthma visits to the ER increased. Additionally, asthma-related tweets and Google searches went up. The researchers’ model was able to predict ER visits with 75 percent accuracy as to whether the ER would expect low, medium, or high volume of asthma-related visits on a given day.
These findings underline the important role of social media data as well as the importance of localized environmental data in successfully preparing hospitals to meet the needs of patients. For instance, hospitals usually look at a patient’s medical history to predict future visits, but environmental data could provide a deeper level of accuracy. Previous studies have indicated that Google searches and tweets could predict the spread of the flu and other contagious diseases.
_NEXT: _Weight loss linked to bone loss