Twitter may help hospitals anticipate a spike in asthma attacks

Researchers are leveraging the microblogging platform to predict the volume of asthma-related emergency room visits one Dallas hospital can expect on a given day.

For those who can see beyond the chatter, Twitter contains a trove of potentially useful data, including information that may be of interest to the medical community. In a recent example, researchers at the University of Arizona (UA) and the Parkland Center for Clinical Innovation in Dallas, Texas mined the social network for information that may help hospitals anticipate peaks in asthma-related emergency room visits.

A chronic condition, asthma currently affects an estimated 25 million people in the US, and its prevalence is rising steadily. Each year, asthma is responsible for around 2 million emergency room visits, half a million hospital admissions and 2,500 deaths, according to the researchers, citing CDC data from between 2001 and 2009.

While there is no cure for asthma, pharmaceutical treatments that are generally effective in preventing attacks. But prevention also involves awareness of certain triggers, including environmental factors such as pollution.

In a special “Big Data” issue of the IEEE Journal of Biomedical and Health Informatics, the researchers explain their innovative system for predicting the frequency of asthma-related emergency room visits at the largest hospital in Dallas.

“You can get a lot of interesting insights from social media that you can’t from electronic health records,” said UA professor Sudha Ram, one of the leaders of the study. “We think that prediction models like this can be very useful, if we can combine various types of data, to address chronic diseases.”

Their method involves monitoring air quality data using environmental sensors and scanning Twitter for posts with relevant key words such as “asthma,” “inhaler” and “wheezing.” Text-mining techniques are then used to hone in on the tweets from the zip codes where most of the hospital’s patients live, according to electronic medical records.

Unsurprisingly, the researchers observed that as air quality worsened, the number of asthma-related ER visits tended to increase. They also found a positive correlation between the number of asthma-related tweets and the volume of asthma-related ER visits.

By analyzing both of these data sets, the researchers developed an algorithm that predicted with 75% accuracy whether the emergency room could expect a low, medium or high number of asthma-related visits on a given day. These predictions allowed the hospital to be better prepared, namely by ensuring that qualified staff and appropriate equipment were in place on peak days.

After testing their method for three months, the researchers are expanding the study to include 75 hospitals in the Dallas-Fort Worth area.

• Health Care Industry
• Twitter
• asthma