The epidemiology of BSE in cattle herds in Great Britain. I. Epidemiological processes, demography of cattle and approaches to control by culling

Abstract

This paper explores the key epidemiological processes and demographic factors that determined the pattern of transmission of the aetiological agent of bovine spongiform encephalopathy (BSE) in cattle herds in Great Britain (GB). The analyses presented utilize data from published and unpublished experimental studies and from the GB central database of confirmed BSE cases. We review the experimental and epidemiological evidence that has both confirmed indirect horizontal transmission via the consumption of infectious material as the major transmission route and provided information on the duration and variability of the dose–dependent incubation period of BSE in cattle. The epidemiological and genetic data pertaining to the possible existence of maternal transmission and/or genetically variable susceptibility to infection is discussed. The demography of British cattle is characterized and the impacts of key demographic features on the observed epidemic profile are discussed. In the main BSE case database, analyses reveal that BSE cases cluster significantly at both the holding and county scale. Furthermore, analysis of longitudinal patterns reveal substantial temporal within–holding correlation. Such clustering of cases suggests a highly heterogeneous infection process. The paper ends with a discussion of how analyses of spatio–temporal clustering inform the design of targeted culling programmes aimed at reducing future disease incidence. We show how the retrospective implementation of culling policies on the BSE case database allows the qualitative evaluation of policy performance, but that model predictions of future trends in case incidence are required to estimate the precise impact of any current or future programme.

The epidemiology of BSE in cattle herds in Great Britain. II. Model construction and analysis of transmission dynamics

“…The form of this distribution, coupled with knowledge of the relatively low peak incidences (even at the herd level, less than 30%) seen, allow much tighter bounds to be (simultaneously) put on both the forms and means of the incubation-period and age-dependent-susceptibility distributions. Indeed, we explored an exhaustive range of forms of both distributions (much wider than the subset presented here), and all produced best-fit estimates of the mean of the incubation-period distribution in the range 4.5-5.5 years, agreeing convincingly with the limited experimental data available (Anderson et al 1996;Donnelly et al 1997b). …”

Estimation of the basic reproduction number of BSE: the intensity of transmission in British cattle

“…Under this, or most other scenarios in which a subpopulation is su¡ering BSE transmission with R 0 b 1, one would have expected to see an increase in the degree of case clustering (Donnelly et al 1997b following the MBM ban. However, statistical analysis of clustering at the herd level has so far failed to reveal any signi¢cant di¡erence between the degree of clustering seen before and after the MBM ban (Donnelly et al 1997b). (b) Infectiousness of cattle through the feed-borne transmission route Figure 3a,b shows estimated annual averages of I FF , the expected number of infections per maximally infectious primary infection slaughtered in an entirely susceptible population.…”

“…if the MBM consumed by cattle in one holding was manufactured from cattle in that holding. Under this, or most other scenarios in which a subpopulation is su¡ering BSE transmission with R 0 b 1, one would have expected to see an increase in the degree of case clustering (Donnelly et al 1997b following the MBM ban. However, statistical analysis of clustering at the herd level has so far failed to reveal any signi¢cant di¡erence between the degree of clustering seen before and after the MBM ban (Donnelly et al 1997b).…”

Implications of BSE infection screening data for the scale of the British BSE epidemic and current European infection levels

“…The total number of animals born in the time interval t 0 to t 0 1 D, N t 0 , and the timedependent birth rate, B(t), were obtained from analysis of annual agricultural census data (Department of Agriculture and Fisheries for Scotland 1975-1980, 1981-1990Ministry of Agriculture, Fisheries and Food 1975-1990, 1992-1995Scottish Of ce 1991-1995 and data on the seasonality of births (Donnelly et al 1997b). Thus, ignoring additive constants, the log likelihood of the clinical case incidence data can be written as…”