Bayesian Spatio-Temporal Modeling of Multi-Pathogen Infectious Diseases Using Data from Multiple Sources

• Activity Number: 130
• Type: Contributed
• Date/Time: Monday, August 1, 2016 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Hunter Merrill* and Xueying Tang and Yang Yang and Nikolay Bliznyuk
• Companies: University of Florida and University of Florida and University of Florida and University of Florida
• Keywords: data fusion ; hierarchical modeling ; spatiotemporal statistics ; surveillance data

Abstract

Mathematical modeling of infectious diseases plays an important role in the deployment and evaluation of targeted interventions. These interventions, such as vaccine development, are usually pathogen-specific, but laboratory testing of all pathogen-specific infections is rarely available. For epidemics of several co-circulating pathogens, it is desirable to jointly model the pathogens to fully understand their transmissibility. This work creates a modeling framework for important multi-pathogen infectious diseases and develops statistical methods for understanding their transmission mechanisms. Our framework is flexible and capable of identifying drivers of transmission from large biosocial and environmental data. We build a hierarchical Bayesian model with a latent process to link disease counts and lab test data. Inference and prediction are carried out by a computationally tractable MCMC algorithm. We study operating characteristics of the algorithm on simulated data, then apply it to hand, foot and mouth disease surveillance data in China. The data set consists of weekly counts of reported cases in 334 prefectures and laboratory test data for a small subset of reported cases.