Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.633250
Title: The analysis of epidemiological data on bovine tuberculosis in a wild badger population : an investigation of approaches to statistical modelling
Author: Walker, Neil J.
Awarding Body: University of Bristol
Current Institution: University of Bristol
Date of Award: 2012
Availability of Full Text:
Access from EThOS:
Abstract:
Bovine tuberculosis is a chronic disease common to a wide range of mammalian specie~ caused by infection with the bacteria Mycobacterium bovis. Eurasian Badgers Meles meles are implicated in the spread of bovine tuberculosis to domestic cattle in the United Kingdom. A long-term study into the epidemiology of bovine tuberculosis in a wild badger population at Woodchester Park in Gloucestershire, England, has been ongoing since 1975. This has given rise to a dataset of more than 15000 observations to date consisting of various de: mographic and biological data on resident badgers in addition to results from tests used to diagnose infection with Mycobacterium bovis. Work presented in this thesis builds on earlier analyses of these data through the inclusion of more recent records and through the application of novel statistical approaches designed to deal with a number of complexities. The present analyses encompass (i) an updated model for population size estimation (ii) a method for estimating bovine tuberculosis prevalence that accounts for uncertainty in results from the diagnostic tests and (iii) a model for spatiotemporal covariance in group-level disease incidence. Finally, elements of the above are synthesised in a single epidemiological model relating the risk of infection to a range of covariates. In this integrated model, efficient use is made of all clinical data in individual badger's life histories. For this and the afore-mentioned analyses, Bayesian methods using Monte Carlo Markov Chain simulation are adopted. Results are compared from this fully developed model and comparable analyses using frequentist methods.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.633250  DOI: Not available
Share: