| This thesis is an investigation of the nature of foot and mouth disease (FMD) detection in areas where the disease is not endemic. Twenty-four FMD epidemics that occurred in non-endemic areas between 1992--2003 were studied. Reasons for delayed detection of the viral incursion included: (1) misdiagnosis on the basis of clinical signs; (2) deliberate concealment of disease by a farmer; (3) very subtle/inapparent clinical signs; and (4) false-negative laboratory tests. There was a disparity in the size of these epidemics; 21 of 24 consisted of less than 150 outbreaks, and 3 involved more than 2000 outbreaks. The dispersion of infected animals through hubs (for example, large animal markets) in the animal movement network seemed to be the critical causal factor in the magnitude of the large epidemics.; The remainder of the thesis focussed on the large 2001 UK epidemic. During this epidemic, the identification of cases was based almost solely on clinical signs. The sensitivity, specificity and negative predictive value of clinical diagnosis were 94.7%, 99.6% and 99.9% respectively. The positive predictive value was lower (76.3%), reflecting the low prevalence of disease.; Logistic regression was used to investigate how factors such as species, location, method of detection and time were associated with: (1) the outcome of the diagnostic process, and (2) time to detection of cases. Cases were identified by active surveillance (tracings, patrols, pre-emptive culls), passive surveillance (investigations initiated by farmer or private veterinarian), or serological surveillance (with no prior clinical signs). Most cases were identified by passive surveillance; these were detected earlier than cases found by any other form of surveillance, and more accurately than those detected by active surveillance. Active surveillance detected subtle cases of disease in which clinical diagnosis was more difficult.; These findings suggest that a key goal of a surveillance system for FMD should be to ensure detection prior to the dissemination of infected animals through a hub. Reporting of suspect cases by passive surveillance and initiatives that increase farmer education and awareness should be encouraged. |
