(373f) Improved Estimation and Control for Large-Scale Models of Infectious Disease Spread

Infectious diseases continue to be a significant public health problem. In developing nations limited resources and other factors prevent sustained administration of effective public health programs. This, coupled with the universal threat of a possible pandemic associated with an emerging infectious disease, necessitates the development of reliable models of infectious disease spread for improved public health decision-making. In this work, we focus on state and parameter estimation techniques for the development of reliable models for the spread of infectious disease and an improved understanding of the important factors affecting the dynamics.

In previous work, we developed an efficient approach for estimating seasonal transmission parameters in deterministic, discrete-time infectious disease models. While this further showed that the patterns of seasonal drivers were correlated with school terms, the use of discrete-time models had several drawbacks. Here, we extend this approach to continuous time models including both measurement and dynamic noise.

In large urban centers (e.g. > 500,000 people), without the influence of vaccination, many childhood diseases are endemic (self sustaining in the absence of imported infections). However, in smaller communities, stochastic fadeout can cause extended periods of time where there are no observed cases. To estimate seasonal drivers, we present a nonlinear programming formulation for maximum likelihood estimation of stochastic continuous time infectious disease models. Estimation is performed using incidence of childhood infectious diseases from three independent locations, the UK, New York City, and Thailand. This approach allows characterization of key time-delays associated with birth rate, estimation of imported cases caused by spatial-coupling, and the physical significance of model parameters. We further demonstrate the use of these models in optimal planning for control of infectious disease spread.