Researchers are developing innovative ways to harness health-related posts on Facebook and updates on Twitter to track public health trends. Now new research from Penn State University offers additional insights into how social-media sites affect and reflect disease networks.
In the study, 318,379 tweets
containing vaccination-related keywords or phrases were collected from August 2009 to January 2010 and rated as either positive, negative or neutral using a specially-designed computer algorithm. The tweets were geocoded to categorize Twitters users' expressed sentiments by U.S. region. Data from the Centers for Disease Control and Prevention (CDC) was also used to determine how vaccination attitudes correlated with CDC-estimated vaccination rates. Additionally, information about Twitter users' followers was examined to identify like-minded groups.
The study unearthed a number of interesting findings:
- Negative expressions spiked during the time period when the vaccine was first announced
- More-positive sentiments emerged when the vaccine was first shipped across the U.S.
- The highest positive-sentiment users resided in New England, which also had the highest H1N1 vaccination rate
- Communities are dominated by either positive or negative sentiments towards the novel vaccine
- Simulations of infectious disease transmission show that if clusters of negative vaccine sentiments lead to clusters of unprotected individuals, the likelihood of disease outbreaks is greatly increased
In a university release, study author Marcel Salathe, PhD, discussed how this approach to mining Twitter data could be used in the future to study non-communicable diseases such as obesity hypertension and heart disease:
We think of a disease such as obesity as noninfectious, while a disease such as the flu is clearly infectious. However, it might be more useful to think of behavior-influenced diseases as infectious, as well. Lifestyle choices might be 'picked up' in much the same way that pathogens -- viruses or bacteria -- are acquired. The difference is simply that in the one instance the infectious agent is an idea rather than a biological entity.
Previously: Facebook app models how viruses spread through human interaction, Mining Twitter data to track public health trends, Following Google Flu Trends, researchers use queries to track MRSA, Modeling the spread of H1N1 flu and Department of Energy lab develops new software for evaluating and responding to pandemics
Photo by Daniel Paquet
