Speaker: Qi “Tracy” Dong, Graduate Student, UW Biostatistics
Abstract: Supplementary immunization activities (SIAs) are implemented worldwide to improve vaccination coverage and to prevent infectious disease outbreaks. The estimation of SIA efficacies, defined as the fraction of susceptibles removed from the susceptible pool after an SIA, remains a critical programmatic challenge. We develop a stochastic discrete time susceptible-infected-recovered (SIR) model to estimate SIA efficacies while accounting for under-reporting. The model is designed specifically for modeling time series of reported measles cases although it can be adapted for other infectious diseases with permanent immunity. We examine two estimation procedures, a full Bayesian method involving the Metropolis-Hastings-within-Gibbs sampler and a two-stage method involving the Hamilton Monte Carlo (HMC) sampler, to make inference about the unobserved infected and susceptible populations and the unknown parameters of interest. We investigate the performance of the model via simulations and propose future steps for improvement.