Public health, the science of epidemics and large-scale prevention efforts, rests on data collection. Knowing what diseases are spreading, what strains, where, how, how quickly– these are the weapons on the arsenal of the protectors of the health of the people. The internet is particularly good for this. Since it connects millions of people, around the world, in real time, it’s a great way to find out what people are doing, how they’re feeling, and what’s happening to them, including illness. In fact, Google is harnessing its Web data analysis capabilities to measure and track flu trends.
If Google can do this with the data it has access to, it stands to reason that Twitter, the microblogging service on which more than half a billion users share their thoughts and activities on an almost minute-by-minute basis. And indeed, epidemiologists and other researchers are using a variety of techniques to sift through these posts and dig out information on people’s health. Looking at 5,000 public messages per minute, computer scientists at Johns Hopkins University have gotten a picture of patterns of flu outbreaks that match the maps generated by the government– but faster, in minutes rather than the two weeks required by the Centers for Disease Control and Prevention
The hardest part, experts say, is to separate discussions from reports, and metaphorical from actual illnesses and symptoms. For example, “fever” can mean high body temperature, but it can also mean extreme enthusiasm. Certain bodily responses have been compared to sneezing, but do not indicate sickness. And people will talk about fear of a disease, or precautions against a disease, or rumors or a disease, or having had a disease in the past, and these need to be distinguished from current instances for the data to be useful.
Already, some interesting patterns have emerged. A private company that has also been working on tracking disease through public internet postings found that the densely populated corridor between Hartford, Connecticut and Washington, D.C., is a hotbed of disease activity, and that disease mentions go up around events such as the Super Bowl that draw large crowds of tourists.
Follow Medex on Twitter: @MedexSupply.