An Agent-Based Modeling Approach to Reducing Pathogenic Transmission in Medical Facilities and Community Populations
MetadataShow full item record
The spread of infectious diseases is a significant and ongoing problem in human populations. In hospitals, the cost of patients acquiring infections causes many downstream effects, including longer lengths of stay for patients, higher costs, and unexpected fatalities. Outbreaks in community populations cause more significant problems because they stress the medical facilities that need to accommodate large numbers of infected patients, and they can lead to the closing of schools and businesses. In addition, epidemics often require logistical considerations such as where to locate clinics or how to optimize the distribution of vaccinations and food supplies. Traditionally, mathematical modeling is used to explore transmission dynamics and evaluate potential infection control measures. This methodology, although simple to implement and computationally efficient, has several shortcomings that prevent it from adequately representing some of the most critical aspects of disease transmission. Specifically, mathematical modeling can only represent groups of individuals in a homogenous manner and cannot model how transmission is affected by the behavior of individuals and the structure of their interactions. Agent-based modeling and social network analysis are two increasingly popular methods that are well-suited to modeling the spread of infectious diseases. Together, they can be used to model individuals with unique characteristics, behavior, and levels of interaction with other individuals. These advantages enable a more realistic representation of transmission dynamics and a much greater ability to provide insight to questions of interest for infection control practitioners. This dissertation presents several agent-based models and network models of the transmission of infectious diseases at scales ranging from hospitals to networks of medical facilities and community populations. By employing these methods, we can explore how the behavior of individual healthcare workers and the structure of a network of patients or healthcare facilities can affect the rate and extent of hospital-acquired infections. After the transmission dynamics are properly characterized, we can then attempt to differentiate between different types of transmission and assess the effectiveness of infection control measures.