Data analytics for bioterrorism surveillance

Abstract

Research on techniques for effective bioterrorism surveillance is limited by the availability of data from the few actual bioterrorism incidents and the difficulty of designing and executing simulated bioterrorism attacks for study. In this research, we describe a preliminary study of a naturally occurring incident, the Florida wildfires from January to June 2001, as a reasonable facsimile of a bioterrorism attack. Hospital admissions data on respiratory illnesses during that period are analyzed to uncover patterns that might be expected from an airborne terrorism attack. The principal contribu-tions of this research are the online analytic processing (OLAP) techniques employed to study the adverse effects of this natural phenomenon. These techniques could provide important capabilities for epidemiologist-in-the-loop surveillance systems, ena-bling the rapid exploration of unusual situations and guidance for follow-up investigations. Research implications are discussed for our on-going development of effective bioterrorism surveillance systems. © Springer-Verlag Berlin Heidelberg 2004.