435. The Use of Twitter to Track Public Concerns about Novel H1N1 Influenza
Session: Poster Session: Hospital-acquired and Transplant Infections
Friday, October 30, 2009: 12:00 AM
Room: Poster Hall A
Background: Twitter is a free micro-blogging service that enables its millions of users to send and read each other's messages, or "tweets." Tweets, limited to 140 characters, can be public, or may be restricted to preselected “followers;” often sent from handheld platforms, they convey a sense of immediacy. The goal of this study was to track rapidly-evolving public sentiment with respect to H1N1 or swine flu using Twitter data.
Methods: Starting 4/28/09, we collected all public tweets matching a set of prespecified search terms (e.g., “oseltamivir,” “h1n1,” “swine,” “flu,” “(face|surgical)+mask,” etc.). Each tweet was timestamped and annotated with the author's geolocation. We created a client-side JavaScript application to continuously update a Google map with the 500 most recent matched tweets, yielding a real-time view of flu-related public sentiment. Users can read any tweet by placing the cursor over its corresponding colored dot on the map.
Results: From 4/28/09 to 5/15/09, we recorded a total of 592,543 H1N1 related tweets. Tweets containing references to antivirals peaked on 4/29, the same day WHO raised their pandemic warning level to 5; hand-washing tweets peaked the next day. H1N1 flu related tweets that also contained travel related terms (“airport,” “travel,” “visit,” etc.) peaked on 5/2, as did those H1Na flu related tweets containing reference to face masks.

Conclusion: We used Twitter traffic to gauge and track users’ interest and concerns related to H1N1 influenza. Twitter user’s initial interest in antivirals such as oseltamivir fell as official disease reports indicated most cases were reltively mild, despite the fact that the overall number of cases was still increasing. Tweet analysis promises to provide an inexpensive way to determine not only levels of anxiety and concern but also to gauge response to news and official public health messaging.
Philip M. Polgreen, MD1, Alberto Segre, PhD1, Alessio Signorini, MCS1 and  A. Signorini,
OneRiot (an Internet services company) Role(s): Employee, Received: Salary.
A. M. Segre, None..
P. M. Polgreen, None., (1)University of Iowa, Iowa City, IA