Show simple item record

dc.contributor.advisorKao, Rowland
dc.contributor.advisorEnright, Jessica
dc.contributor.advisorBanks, Chris
dc.contributor.authorRuget, Anne-Sophie
dc.date.accessioned2022-06-29T10:08:26Z
dc.date.available2022-06-29T10:08:26Z
dc.date.issued2022-06-29
dc.identifier.urihttps://hdl.handle.net/1842/39237
dc.identifier.urihttp://dx.doi.org/10.7488/era/2488
dc.description.abstractInfectious diseases have played a considerable role in shaping human history. Although their global burden has significantly decreased through the past centuries, they are still among the main causes of human death worldwide. In livestock, infectious diseases can cause substantial production losses but also have detrimental impacts upon human health, and animal health and welfare. Changes in practices and development of treatments and vaccines have helped to dramatically mitigate the impact of infectious diseases, but infectious diseases remain an ongoing challenge, either because they are difficult to control (Tuberculosis, Malaria, HIV, FMD) or because they are emerging or re-emerging pathogens. Human mobility and livestock movements play a crucial role in epidemic spread as they allow for long-range transmission and can act as bridge between otherwise disconnected populations. Repeated importations of cases in disease-free areas make the eradication or control of a disease exceedingly difficult. The patterns of potentially infectious contacts, as recorded in mobility and movement data, can be described as a network. Understanding infection transmission on networks can provide useful insights in disease risk. Mathematical models have played an increasingly important role in helping to control epidemics in animal (FMD, Avian Influenza, Swine Fever) and in human (Measle, Malaria, SARS, Ebola) populations. Modelling tools are now a central feature in the decision-making process for policy makers, as illustrated by the ongoing COVID-19 pandemic and its management. The aim of this work is to show how disease models in combination with movement or mobility data can be useful in different epidemic contexts, namely in peacetime, at the start of an outbreak and once the pathogen is circulating. Part One investigates how livestock movement data and network analysis can be used in peacetime to improve our understanding of disease risk and to propose tools for control. In this part, I consider a fast-spreading disease affecting cattle and sheep. First, I use multi-species movement networks to understand how the combination of cattle and sheep movement affects the potential for disease spread on the combined network. I compare results of single-species vs multi-species and static vs dynamic network analyses to show the importance of interspecies links and temporal network dynamics. My results show that depending on the season, up to 70% of the premises which are likely to drive the epidemic in the multi-species network differ from the ones in both the cattle and the sheep networks. This indicates that their risk is derived from interaction between the two farming systems. Secondly, I propose the use of a dynamic network measure based on contact chains calculated in a network weighted with transmission probabilities to assess the importance of premises in an outbreak. Comparing results with disease simulation model outputs, I demonstrate that the measure proposed allows us to identify around 30% of the key farms in a simulated epidemic, ignoring markets. Whereas static network measures identify less than 10% of these farms. Part Two explores how mobility data within disease models can be used during an epidemic: before the pathogen is introduced (importation phase) and once the pathogen is present (circulation phase). In this part, I use the COVID-19 pandemic and its spread in the Scottish Hebrides, an archipelago off the west coast of Scotland. First, human mobility data and a metapopulation model are used to estimate the risk of introduction in each of the Islands, according to season and potential for control. I show that in some islands the introduction risk is high even in the low season, when activity and movements from the mainland are expected to be reduced. This will be of particular concern if COVID-19 becomes a seasonal respiratory infection affecting temperate areas in winter concomitantly with other seasonal infections such as flu. In the high season, although in most cases movement control will not significantly delay a potential introduction, for some islands a 70% reduction of movements in peak summer tourist season has the potential for delaying the introduction risk for over 6 weeks, i.e. beyond the high risk summer holiday period. Secondly, data from an outbreak localised in Barra Island (Western Hebrides) are used to illustrate how adjusting model parameters to disease data can provide insight in transmission dynamic and control measure efficacy. Using Approximate Bayesian Inference, I estimate the most likely date of introduction, the basic reproduction number at the start of the outbreak and I quantify the impact of voluntary vs policy-induced measures. I find that transmission started to slow down two days after the first cases were reported and a week before restrictions were imposed by the authorities. Thus my analysis is most consistent with the outbreak being mostly contained by a combination of contact tracing and self-imposed measures, whilst the lockdown, which was later imposed, had only a negligible effect on the transmission dynamic.en
dc.language.isoenen
dc.publisherThe University of Edinburghen
dc.subjectn/aen
dc.titleMovement, mobility and disease modelling in three epidemic contextsen
dc.typeThesis or Dissertationen
dc.type.qualificationlevelDoctoralen
dc.type.qualificationnamePhD Doctor of Philosophyen


Files in this item

This item appears in the following Collection(s)

Show simple item record