How Analyzing Social Media Can Improve Healthcare

  2 m, 6 s

health_care_and_sentiment_analysisSocial media monitoring is crucial for brand health, but an increasing number of studies are proving that SSM may provide new tools for improving population health as well. Here are three examples of science using social media to improve healthcare:

Linking Sentiment and Heart Disease:

A study by researchers from the University of Pennsylvania were able to predict rates of coronary heart disease by county using Twitter. They gathered tweets and analyzed them using sentiment analysis. Their research showed that expressions of negative sentiment such as anger, stress, and fatigue within a county were associated with higher risk for heart disease.

Predicting Emergency Room Visits:

A team of researchers led by UA professor Sudha Ram created a model that was able to predict the amount of asthma-related emergency room visits a Dallas hospital could expect that day. The model used data gathered from text mining Twitter as well as from electronic medical records and air quality sensors. They found that asthma-related emergency room visits rose as air quality in the area fell, and that ER visits correlated with an increase in tweets containing asthma keywords such as “asthma”, “inhaler”, or “wheezing”. By combining these two data sources, the team was able to use machine learning to predict the volume of asthma-related ER visits with 75% accuracy. Predictive models are a new tool that may help tackle health-care challenges such as how to best allocate resources and staff within a hospital.

Tracking HIV Outbreaks:

A study published in the scientific journal Preventative Medicine used real-time social media to track behaviors linked with HIV risk. Researchers collected more than 550 million tweets and analyzed them for words and phrases related to risky sex and drug related behaviors. They then ran statistical models to see if the geographical locations matched up with areas where HIV cases had been reported. They found a significant relationship between tweets indicating risky behavior and countries with the highest number of reported HIV cases. The study found that the largest number of HIV risk-related tweets per capita came from the District of Columbia, Delaware, Louisiana, and South Carolina.

Healthcare Research is Going Social

How we choose to express ourselves on social media says a lot about us, and may provide clues to our wellbeing. Researchers are turning to social media as a possible alternative to traditional survey methods, which are more time-consuming, expensive, and limiting. Combining social media sources with big data analytics may provide healthcare experts with vital information on how to improve our healthcare system and quality of life.

Categories: Social Media, Text Mining, Twitter