Spatial Epidemiology of Yaws and Tuberculosis and Implications for Public Health Interventions

Decision making about what role, if any, spatially targeted interventions should play in infectious disease elimination programs is complex and depends on a range of considerations, including the underlying spatial epidemiology of the disease. This dissertation investigates the spatial epidemiology of yaws and tuberculosis. It also explores the implications of spatial epidemiology on the relative merits of a spatially targeted approach to yaws eradication versus a population-wide approach. The first chapter analyzes patterns of spatial-temporal clustering of yaws on Lihir Island, Papua New Guinea. We found that clusters of yaws typically transcend village boundaries and we argue that villages may be epidemiologically linked to a degree such that mass drug administration (MDA) may be more effectively implemented at a spatial scale larger than the individual village. The second chapter uses a stochastic compartmental model of yaws transmission to investigate how total targeted treatment (TTT) performs compared to MDA in different settings. We found that the performance of TTT is improved when there is little mixing between subpopulations and when there is spatial heterogeneity in transmissibility, but even in these settings, our model suggests that MDA can achieve the same result as TTT more quickly and probably at lower cost. The third chapter uses tuberculosis whole-genome sequence data from Lima, Peru to explore the spatial scale at which transmission occurs. We found that transmission risk was highly elevated at short spatial distances but that long-distance transmission events occur, too. Taken together, these 3 studies illustrate a range of tools that can be used to better understand the dynamics of infectious diseases in space and to understand the implications of spatial epidemiology for the design of public health programs.