This article is more than 5 years old. Information may no longer be current.
Asthma-related Twitter posts forecast ED visits
Click Here to Manage Email Alerts
BALTIMORE — An analysis of real-time Twitter data was able to prospectively project population-level asthma ED visits and hospital admissions, according to study findings presented at the Pediatric Academic Societies Meeting.
“By using real-time Twitter activity, health departments could actually anticipate asthma ED visits or hospitalizations in the following days and possibly intervene before some of them occur,” Yolande Mfondoum Pengetnze, MD, medical director at Parkland Center for Clinical Innovation in Dallas, said in a press release. “For instance, a notification might be sent by the health department when there is an increase in asthma-related tweets in the community, giving people with asthma a heads-up to take necessary precautions, like avoiding exposure to asthma triggers or being more assiduous in taking their asthma medications.”
To evaluate the correlation between real-time Twitter data and asthma ED visits or hospital admissions in the Dallas-Fort Worth area, the researchers sampled 3,810 tweets posted between October 2013 and June 2014 that used common asthma terms. Pengetnze and colleagues then created a dataset of tweets to distinguish asthma-relevant tweets, those where the user mentioned their ongoing asthma symptoms, from nonrelevant tweets, in which the user only mentioned asthma as a comment.
The researchers used Twitter location data to isolate tweets from the Dallas-Fort Worth area, and only included asthma-relevant tweets in their analysis. Using linear regression analyses, they evaluated the chronological association between the number of asthma-relevant tweets and asthma ED visits/admissions data from the Dallas Fort Worth Hospital Council Foundation, a network of more than 70 hospitals.
A positive association between the average number of asthma-relevant tweets in a 7-day period and the number of asthma-related ED visits in the subsequent 7 days was observed (beta = 0.58; 95% CI, 0.47-0.69).
In addition, the researchers found similar correlations with other time windows of asthma-relevant tweets with asthma-related ED visits/admissions. After days with a significantly increased number of asthma-relevant tweets, the researchers calculated the ensuing 7 days as “high” asthma days or days when the number of asthma visits were greater than average.
“Our research is innovative and unique because it harnesses the power of Big Data from social media and other sources to address the problem of anticipating ED visits for a chronic condition, in this case asthma, in close to real-time conditions,” study investigator Sudha Ram, PhD, director of the INSITE Center for Business Intelligence and Analytics at the University of Arizona, said in the release. “We believe this work paves the way to address signal extraction and prediction for other chronic conditions and goes beyond current work that mostly looks at infectious conditions.” – by Bob Stott
Reference:
Pengetnze YM, et al. Abstract 4600.8. Presented at: Pediatric Academic Societies Meeting; April 30-May 3, 2016; Baltimore.
Disclosure: The researchers report no relevant financial disclosures.
Perspective
Back to Top
Daniel M. Fein, MD
[This study] comes out of the Dallas metro area where researchers used ‘big data’ and Twitter tweets to see if they could predict a rise in the visit for asthma to the emergency department or hospitalizations for asthma depending on the number of asthma-related tweets available in the previous short time period.
I think this is a novel idea. This is the first study like this that I have seen and I think this is just the tip of the iceberg: ‘Big data’ is a buzz word across all fields – finance, politics, social networking – and now, it’s really going to get into medicine.
The researchers were able to show that for a rise in tweets about asthma, there was an increase in ED visits for asthma-related symptoms. This is fascinating, not only because it can provide emergency medicine physicians ways to think about how we can use this to our advantage in terms of staffing, predicting what will happen not only for exacerbations of chronic conditions like asthma, but also the flu or outbreaks of norovirus and gastrointestinal symptoms.
We can also have the forethought to go proactively to the social media users. If people are tweeting about their asthma symptoms, and there is a built-in base of children who are following the institution on whatever social media platform, then we can proactively approach those children, and say “Be careful – please use your asthma medication. We know that there are a lot of asthma allergens around now and people are getting worse.”
I think this is just the start of research that will focus on ‘big data’ and how it affects all different fields, not only pediatrics or emergency medicine, but everything from birth to geriatrics.
Click Here to Manage Email Alerts