A Spatio-Temporal Bayesian Model for Syndromic Surveillance: Properties and Model Performance

• Activity Number: 21
• Type: Topic Contributed
• Date/Time: Sunday, July 29, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Jian Zou*+
• Companies: Indiana University Purdue University Indianapolis
• Keywords: syndromic surveillance ; spatio-temporal ; conditional autoregressive ; SIR model

Abstract

Analysis of syndromic surveillance data is often complicated by spatio-temporal dynamics, uncertainty in disseminate multiple sources of epidemic, and correlation in different syndromes. In this talk, We propose a flexible hierarchical Bayesian model to partition the variability and quantify uncertainty in a unified framework. The methodology incorporates Gaussian Markov random field (GMRF) and spatio-temporal conditional autoregressive (CAR) modeling. The methodology has some nice features including timely detection of outbreaks, robust inference to model misspecification, reasonable prediction performance, as well as attractive analytical and visualization tool to assist public health authorities in risk assessment. Our hierarchical model decompose the variability of the source into different components, which has reasonable epidemiological interpretation. Numerical results suggest that the model performs sensibly and robust to various less than ideal conditions such as various signal-to-noise ratio (SNR), choice of prior, missing and superfluous edges in the network structure, and other possible misspecifications. Multivariate extension of the methodology will also be discussed.