Editor's note: The following articles are compiled from Medical Device Daily's Medical Technology and Pandemic Threats: Swine, Avian and SARS. For more information or to order a copy, please visit www.medicaldevicedaily.com or call 800-688-2421.
Medical Device Daily National Editor
The Internet for some time now has been developing as a tool for spotting the outbreaks of infectious diseases such as the swine flu — just keep track of the "hits" for flu information and where they are spiking geographically you have a problem. And this type of application has now spread to what is often considered a less substantial, even frivolous, application of the Internet, social networking.
Launched November 2008, OneRiot is a search engine that crawls the links that people share on Twitter, Facebook, Digg and other social sharing services. It then indexes the content on those pages in seconds, providing a real-time look at what has been described as the most "socially relevant" content from the web.
While some would question the relevance of what most social network participants are talking about, they have shown understandable concern about the swine flu, which OneRiot is tracking.
Seeing value in this real-time application, researchers at the University of Iowa (UI; Iowa City) are using OneRiot to track public perceptions of the swine flu outbreak and other infectious diseases.
The researchers say that the project, named Social Web Information Monitoring for Health (or SWIM for Health) has the potential to enhance disease tracking and forecasting.
The SWIM for Health project is being conducted as part of the interdisciplinary Computational Epidemiology research group (CompEpi) at UI. The UI CompEpi group applies computational methods to help solve public health problems, including new disease surveillance methods and harnessing technology for disease prevention.
SWIMS for Health is headed up by researcher Alessio Signorini, a PhD candidate and Director of Search Technology at OneRiot. Signorini is using real-time data provided by OneRiot, in conjunction with official public health surveillance data, to discern patterns in disease diffusion and public perception of health scare outbreaks.
Signorini told Medical Device Daily that the use of OneRiot really has two purposes, closely linked.
He said that talk on the social networks clearly has spiked following the initial media accounts of the swine flu and its threat to spread globally. Thus, it tracks public concerns, acknowledging that this is often linked to media reports and that as the media moves on to other issues — who will be the newest American Idol winner or whatnot — talk of the flu drops off, then reignites a bit when there is another report of, say, a local outbreak.
But like the prediction markets, also being developed at UI, the talk and information sharing across social networks can also be used to predict flu trends — where it may be appearing next.
However, unlike the predictive markets being offered by the Iowa Electronic health Market, SWIM for Health and OneRiot do not draw on "expert" opinion but rather on anyone participating in these social network conversations.
Signorini told MDD that the system is currently tracking seven social networks, the various offerings gathered from these sources ranging from blog posts to shared news stories and videos.
As evidence of the system's forward-looking abilities, Signorini points to the system's prediction of an American Idol winner in late May 2009, some prominent Internet "authorities" going with Adam Lambert, and OneRiot and Signorini's research accurately picking the eventual "upset" winner Kris Allen.
Philip Polgreen, assistant professor of internal medicine and member of the CompEpi faculty, believes it has this usefulness, and that "analyzing and aggregating real-time data from the social web is a unique enhancement to traditional disease surveillance systems." Agreeing with that assessment, Richard Rothenberg, a veteran of 25 years at the Centers for Disease Control and Prevention (Atlanta) and now a professor of pubic health at Georgia State University (also Atlanta), said that social networking provides "the opportunity to be in touch with one another in a way that has never been the case during a prior epidemic, and that will make a huge difference in keeping track of the disease."