Bayesian Inference for Early Detection of Influenza Epidemics

• Activity Number: 550
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Xian Yu*+ and Didun Peng
• Companies: University of Arkansas at Little Rock and University of Arkansas at Little Rock
• Keywords: Bayesian inference ; Triangular prior ; Binomial Markov chain ; Influenza epidemics ; Hierarchical Bayes Model

Abstract

Early and accurate detection of outbreaks of epidemics is one of the most important objectives of influenza surveillance systems. We propose a hierarchical Bayesian change-point model for detection of influenza epidemics under binomial Markov processes framework. Prior probability of a change-point depends on factors that affect the spread of influenza. The proposed hierarchical Bayesian models with triangular prior distribution couple with CUSUM detection procedures are applied to the 2001-2009, 2010-2013 seasonal influenza and 2009 H1N1 influenza pandemic data published by the Centers for Disease Control and Prevention (CDC).