Modeling the dynamics of disease elimination

Elimination of an infectious disease is often a goal of the public health community. Although that goal is rarely achieved, the tremendous expansion of epidemiological databases provides new opportunities to test hypotheses concerning elimination with mathematical modeling. Besides improving our scientific understanding of disease transmission, hypotheses validated through mathematical modeling provide public health practitioners with a more structured, quantitative assessment of how elimination of specific pathogens can be achieved. This proposal aims to develop an interconnected set of modeling tools to support elimination of communicable diseases. A variety of processes used to achieve disease elimination will be considered including use of mass drug administration to eliminate neglected tropical diseases such as trachoma, vaccination for preventable diseases such as SARS-CoV-2, and antibiotic stewardship efforts to curtail drug resistant infections such as methicillin-resistant Staphylococcus aureus (MRSA). A key theme is the requirement of subcritical transmission for disease elimination, meaning that the average number of new infections each case causes is less than one. A major goal is to elucidate the transmission dynamics of subcritical diseases on the brink of elimination. Transmission heterogeneity may arise from many mechanisms including super-shedding of certain individuals, pockets of susceptibility such as in a community with low vaccine uptake, and contact structure in which some individuals have the potential to infect many others. Simulations of various patterns of disease transmission will be used to develop distinct measurements of transmission heterogeneity. In addition, new techniques to infer and compensate for observation error will be developed that integrate data on the observation process, such as the proportion of cases identified retrospectively via contact tracing programs. Models of transmission dynamics will be used to identify transmission-hotspots and superspreaders that can jeopardize elimination. People, areas, or events that have increased transmission potential can maintain endemic disease transmission even though the population- level average value of R may be less than one. In the first stage of this objective, we will use existing models to construct a suite of in silico simulations to compare the performance of various scan statistics designed to detect disease burden beyond what is expected by chance. In the second stage, we will apply these scan statistics to observational data. Identification of transmission-hotspots and supersreaders permits optimization of disease elimination strategies. To eliminate disease, it is insufficient to merely identify transmission- hotspots or superspreading activity. A strategy is needed for suppressing the sites, events, or people that cause higher levels of transmission. We will use mathematical and computational models for disease elimination to address 1) the impact of control interventions, 2) the optimal distribution of a limited treatment supply, and 3) monitoring of treatment efficacy.