Leptospirosis is a zoonotic disease caused by infection with Leptospira bacterial species. Human infections occur after exposure to infectious urine, predominantly from rats. This urine is found in contaminated waters and muds in areas where rat populations are abundant. Therefore, leptospirosis is a disease with strong environmental drivers. Global estimates have attributed 1 million cases each year to leptospirosis, with 58 900 yearly deaths. It is therefore considered one of the leading causes of zoonotic disease in the world. There are some important factors of leptospirosis epidemiology that remain unclear. In this thesis, I have focused on two broad aspects: human mobility and environmental perception. The thesis is made up of four papers that show the work I carried out along with collaborating co-authors to examine how these aspects affect leptospirosis epidemiology in urban slums in Salvador, Brazil. The first two papers use telemetry data collected using GPS loggers to analyse human mobility. We begin by examining how step selection functions, a spatio-temporal point process model used in animal movement ecology, can be used to analyse how individuals interact with their environments. We also consider how these interactions may vary across socio-demographic features like age and gender. This research shows that step selection functions were a useful tool to quantify human movement behaviours. The analysis also suggested that there were movement differences based on gender. Movement analysis was expanded in the second paper, where we take a more in depth look at various movement characteristics. We examine how these movement behaviours vary across individual characteristics and daily characteristics, such as rainfall. Our research shows that gender differences in movement behaviours may not follow expected patterns. It also provides a strong foundation to research the effects of rainfall on human movement behaviours.. In the following two papers we used collaborative mapping to capture people’s perceptions of their environment. Collaborative mapping is a type of participatory research method that involves individuals in the map creation process. Through this, we are able to obtain their perceptions on various aspects of their environment. In the third paper, we ask residents to identify areas that they consider risky for their community’s and their own health. We also ask them to classify these areas into various categories. We then used these perceptions to compare them to objective measures of comparable risks. We found there was a low perception of leptospirosis as a health risk and that residents more easily identified visible risks, such as rubbish piles. For the fourth paper, we collected individuals’ perceptions on where interventions should take place in their neighbourhoods. We compared these areas with a prediction for leptospirosis seropositivity, created using household leptospirosis data. Residents’ intervention requests clustered around the communities’ stream rather than the predicted leptospirosis hotspots. Both these pieces of research highlight areas of knowledge that could be improved and disagreements on what interventions should occur and where. Overall, this thesis presents innovative ways to use existing methods to analyse data in the broad context of infectious disease and, more specifically, leptospirosis. The results presented can be used to guide public health interventions and how to assess them. They also present a foundation for further research to build upon to understand the mechanisms of leptospirosis infections in high prevalence, low resource settings.
Pablo Ruiz Cuenca (Thu,) studied this question.