Researchers collected 550 million tweets over several months in 2012. Using an algorithm, they were able to capture and geographically map 9,800 tweets related to potentially risky sexual behaviors and drug use. They compared the data with that of reported HIV-transmission cases and found the two closely matched each other.
This is the first study to suggest that Twitter can be used to predict people's health-related behaviors and as a method for monitoring HIV risk behaviors and drug use
– Sean Young, Assistant Professor at David Geffen School of Medicine at UCLA
Researchers noted the data on HIV was from cases reported in 2009. They said in order to use "big data" culled from social media to predict outbreaks, data would ideally come from a source that updates more frequently.
Researchers Find That Twitter Can Locate HIV Outbreaks | TechCrunch
Researchers found that California, Florida, New York and Texas had the largest proportion of both geo-located and HIV-linked tweets. The areas with the highest per capita rate of HIV-related tweets were Washington D.C., Delaware, Louisiana and South Carolina.