default
CONTACT
REQUEST A DEMO

Twitter, MD: Using Social Media to Track Health Risk

twitter-on-a-computerConsumers are changing the healthcare game — and hospitals need to be nimble, developing strategies that simultaneously meet the consumer and regulatory demands of healthcare reform.

Psychographic segmentation allows healthcare organizations to classify patient groups by shared motivations and attitudes towards wellness to create more effective approaches to patient engagement. And now, according to a recent University of Pennsylvania study, social media may help hospitals predict, and thereby better manage, chronic disease within the community.

How Much Can 140 Characters Tell Us?

Not everyone who is sick visits the doctor or a hospital. Yet in recent years, researchers have been able to more accurately and quickly predict flu outbreaks because data reported by healthcare providers is being complemented by healthcare consumers who google their symptoms or tweet about their illness.

An article in Emergency Management magazine earlier this year quoted HealthMap co-founder John Brownstein who said, “There are real opportunities for using this data that is scattered across the Web in news, blogs, chat rooms and social media. We’re focused on collecting all that information using data scraping, machine learning and other processes and combining it into one platform that will enable clinicians, public health practitioners and consumers to see what’s happening.”

He’s not alone in his assessment. Ming-Hsiang Tsou, a professor at San Diego State University, published a study titled, “The Complex Relationship of Realspace Events and Messages in Cyberspace: Case Study of Influenza and Pertussis Using Tweets.” Tsou’s research team aggregated user tweets near 11 different cities across the country, then compared the social buzz mentioning flu or whooping cough-like symptoms to national flu data. According to Tsou, “The correlation between the weekly flu tweets versus the national flu data was almost 86 percent.”

Results were even better when Twitter® data was compared with the more precise data available on a local level.

Gaining a Deeper Understanding of Healthcare Consumers

The University of Pennsylvania study, which was just published in Psychological Science, suggests that social platforms like Twitter may act as “a dashboard indicator of a community’s psychological well-being and can predict rates of heart disease.” In addition to risk factors like low income and smoking, anger, stress and fatigue can all exacerbate cardiovascular conditions — and tweets are ripe with these emotions, as well as more positive ones. Semantic analysis of two years’ worth of tweets allowed the researchers to identify commonly used emotional language, as well as word clusters related to behaviors or attitudes, which can be used to filter tweets and predict where negative emotions may play a role in the prevalence of chronic diseases.

Researchers at Johns Hopkins are also using algorithms to monitor Twitter posts to gather mental health data ranging from depression and bi-polar disorder for a joint study on post-traumatic stress disorder with the U.S. Naval Surface Warfare Center. A press release from the University of Pennsylvania similarly noted that social media data could become “an invaluable public health tool if able to be tied to real-world outcomes.”

Of course, tying the data to real-world outcomes is where it gets complicated. Apart from the privacy issues that must be addressed, healthcare providers are well-acquainted with the difficulties of moving from insight to solution. The first problem? Patients do not want to be engaged with their chronic conditions; they want to be engaged in their lives. Hospitals need to understand what motivates individuals rather than taking a generic approach to a population segment based on a diagnosis or demographics. The proprietary psychographic segmentation model developed by PatientBond classifies healthcare consumers based on unique features, such as their propensity to be self-motivated versus needing direction regarding healthcare. The model also shows which psychographic segments are most likely to use — or not use — social networking sites.

Of the five psychographic segments identified by PatientBond, Willful Endurers are statistically less likely (95% confidence) than the other four segments to participate in a social networking app or website.  Willful Endurers are also the most reactive and disengaged segment when it comes to health and wellness, and struggle with many health conditions. The implication is that social media may actually under-estimate community health issues and may not be a sufficient outreach tool for many healthcare consumers who need care.

However, for other psychographic segments that are regular participants in social media, apps and sites like Twitter and Facebook are critical resources. Hospitals may find, as a Boston Globe editorial suggested that, “Twitter is, apparently, the quiet therapist to whom we reveal much more that we realize. As such, it could be a valuable public-health tool.”

But without deep insights into what will motivate individuals to engage with healthcare providers, social media data alone will not improve patient outcomes. Want to explore how psychographic segmentation can complement social media and mHealth as a means to engage and motivate healthcare consumers? Check out our whitepaper on patient activation or contact PatientBond to learn more.

Psychographic Segmentation and its Practical Application in Patient Engagement and Behavior Change

DOCUMENT_ICON_WHITEPAPER_WHITE
DOWNLOAD WHITEPAPER ⇩
psychographic-segmentation-whitepaper



Submit a Comment

Request a Demo