Computational modeling of spontaneous behavior changes and infectious disease spread

Abstract

In the Spring of 2009, a new strain of pandemic influenza virus emerged in the human population and spread to major countries worldwide. This caused panic that the world was witnessing another influenza outbreak potentially of the size of the 1918 Spanish Influenza outbreak where a fifth of the world’s population was affected. Although, this fear did not come to pass, the threat of a potentially deadly outbreak remains. The ability to mitigate and contain a disease is a vital aspect of any country’s response strategies. Through modeling and simulation of the spread of an outbreak, decision-makers can better plan mitigation and containment strategies. This dissertation investigates how changes in human behavior affect the spread of pandemic influenza in the U.S. population using an agent-based computational model. The dissertation argues that more aspects of human behavior such as people’s attitudes and trust in government-issued health advisory infor- mation about the disease need to be integrated into population-level models of pandemic influenza to improve model realism. I present a framework for incorporating such factors into computational models of disease spread to simulate possible scenarios that the spread may take to improve policy insights. I created models to represent different configurations of the attitudinal disposition of the population and then examined how agents representing individuals responded to the interventions implemented. The study revealed that a popu- lation that responds positively to government interventions reduced overall disease impact in comparison to the other scenarios modeled. Although the model is built on the U.S. population, it may be generalized for other synthetic populations in the future.