Flu, Twitter, and the Future of Public Health Surveillance

Saw a really cool article about public health and social media. Actually, it was about flu and Twitter, titled, "Scientists Use Twitter to Track Flu Epidemics, How to Stay Healthy This Flu Season."

The take-away from this article isn't that scientists used Twitter to track flu. Although, o.k, that's important. The really important development is that scientists developed an algorithm to track public health activity via social media.

First, this thought: Realtime surveillance has been the dream of many public health authorities and researchers. Various models have developed to track diseases, primarily by tracking data submitted by hospital ERs. The University of Pittsburgh has an example of this with the RODDS Laboratory. (If you don't want to click the link, RODDS stands for Real-time Outbreak and Disease Surveillance System).

Second, this thought: I've been told that a true realtime surveillance system would be very difficult to establish in the U.S. in the absence of a single payer health insurance system. In other words, in countries where there is a single payer system, e.g. the government, all patient data are collected and aggregated by a central database system. Theoretically, such a single point of data entry allows realtime surveillance.

But, this article points to a wonderfully efficient, even elegant, way to do an end run around enormously expensive database systems that would have to be woven through the enormously tangled HIPAA regulations and the rules, regulations, and private market interests of the healthcare providers and the healthcare insurers.

To track the flu, scientists--lead biologist was Marcel Salanthe from Penn State U.--analyzed 70,000 Tweets about flu. The Tweet data were analyzed with CDC data on vaccinations. The authors observed patterns that showed crossover between flu outbreaks and vaccinations.

And how did they do this? They developed the algorithm. That's the beauty of their work. An algorithm like this can be used to surveil any number of public health threats. Quoting from "What Twitter Knows About Flu," quoting Salanthe, "You can observe how people feel about certain things in a real-world context....On Twitter, people talk about things that matter to them in a completely normal context."

No worries about confidentiality, no delving into private insurers' proprietary data, no laws and regulations to comply with--these data are voluntarily provided by people on Twitter, the world's most realtime forum.