Statistical modeling of disease emergence

Infectious diseases seem to be appearing at an unprecedented rate: within the last few years alone, a sequence of novel diseases like MERS-CoV, Chikungunya, and Zika have emerged. Concurrently, a number of previously known diseases have re-emerged like the 2009 H1N1 pandemic and the 2014 Ebola epidemic. While these known and unknown emergence events have all begun with a wildlife or livestock spillover transmission event into humans, they each present unique subsequent public health challenges. Quantitative prediction of either the re-emergence of a known disease or potential for global spread of a novel disease can help optimize public health responses and resource allocation, but these events are usually analyzed in retrospect. In this dissertation, I developed quantitative frameworks that can be used in real-time for predicting disease emergence risk. In Chapter 2, I identified a seasonal trend to pandemic influenza emergence events, and proposed a hypothesis to explain the seasonal patterns and predict pandemic emergence risk for seasonal flu data. In Chapter 3, I developed a framework to both predict the number of imported Zika cases into a region, and subsequently assist public health decision-making during an uncertain outbreak. Finally, in Chapter 4, I developed a method that can be used to update regional transmission risk estimates of a novel disease before transmission occurs. Altogether, the results presented in this dissertation suggest that statistical modeling can be an important tool to assist real-time public health predictions and responses.