By monitoring certain “risk behavior” keywords, researchers can track HIV infections in cities.
Here’s some crazy shiz to wrap up your Wednesday: A team of researchers at UCLA is using Twitter and other social media platforms to track HIV outbreaks in major cities.
In a recent study, the team determined that there was a link between certain Twitter keywords related to various risk behaviors, such as drug use and “sexually risky” behavior, and major HIV outbreaks. By monitoring people’s tweets for these keywords, they could track—and, in some cases, predict—the geographical distribution of the HIV virus in various cities.
Published in the journal Preventive Medicine, the study collected more than 550 million tweets from May to December 2012, creating an algorithm to find tweets with keywords that referenced risk behaviors, such as “sex” or “get high.” When they plotted the tweets on a map, they found there was a significant link between the tweets and geographical locations where HIV infections had been reported.
This is not the first time public health researchers have used social media to track disease outbreaks. Other studies have focused on using Twitter to track influenza infections, and Google Flu Trends also uses aggregated worldwide flu data to track recent outbreaks. But this is the first time that researchers have used Twitter to track HIV infection, and they’re optimistic that the method will serve as an effective means of “monitoring HIV risk behaviors and drug use,” Sean Young, an assistant professor of family medicine at UCLA, said in a press release.
Of course, the study has various weaknesses, chief among them being that their data on HIV infections had not been updated since 2009, (the second, more obvious reason: tweeting about having “sex” and “getting high” doesn’t necessarily indicate you’re engaging in those behaviors IRL, nor are either of those things inherently hazardous). But at the very least, the results have fascinating implications for the future of public health and social media, as well as the power of a public platform like Twitter to predict our private behaviors.