................................................................. Estimating the human health risk from possible BSE infection of the British sheep ¯ock N. M. Ferguson, A. C. Ghani, C. A. Donnelly, T. J. Hagenaars & R. M. Anderson Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College of Science, Technology and Medicine, St Mary's Campus, Norfolk Place, London W2 1PG, UK .............................................................................................................................................. Following the controversial failure of a recent study 1 and the small numbers of animals yet screened for infection 2 , it remains uncer- tain whether bovine spongiform encephalopathy (BSE) was trans- mitted to sheep in the past via feed supplements and whether it is still present. Well grounded mathematical and statistical models are therefore essential to integrate the limited and disparate data, to explore uncertainty, and to de®ne data-collection priorities. We analysed the implications of different scenarios of BSE spread in sheep for relative human exposure levels and variant Creutzfeldt± Jakob disease (vCJD) incidence. Here we show that, if BSE entered the sheep population and a degree of transmission occurred, then ongoing public health risks from ovine BSE are likely to be greater than those from cattle, but that any such risk could be reduced by up to 90% through additional restrictions on sheep products entering the food supply. Extending the analysis to consider absolute risk, we estimate the 95% con®dence interval for future vCJD mortality to be 50 to 50,000 human deaths consider- ing exposure to bovine BSE alone, with the upper bound increas- ing to 150,000 once we include exposure from the worst-case ovine BSE scenario examined. The aim of this study was not to evaluate the probability that BSE has entered the sheep ¯ock, but rather, given the pessimistic assumption that infection has occurred, to explore its potential extent and pattern of spread. In this, we used epidemiological parameter estimates from experimental BSE infections of sheep, and, where data are unavailable, assumed (given the observed similarities in BSE and scrapie pathogenesis in sheep) that other aspects of disease epidemiology resemble those of scrapie. Analyses were constrained to be consistent with the failure to detect the BSE agent in a small sample of 180 brains 2 collected between 1996 and 2000 from sheep diagnosed with scrapie (giving an upper bound for BSE prevalence within apparently scrapie-affected sheep of 2%; Fig. 1a), and are also broadly consistent with an assessment of historical exposure of the ovine population to meat and bonemeal (MBM) 3 . Key to our analysis are estimates of the infectivity in animal tissues during disease incubation. Data are limited for BSE in sheep acquired by oral challenge, but using new experimental results and published data from studies of both scrapie 4±7 and sheep BSE 8,9 pathogenesis, we constructed an infectiousness pro®le. This pro®le was based on temporal changes in the density of the agent in different tissues, weighted by the proportion of such tissue in the host's body (Fig. 1b). This pro®le differs from that of BSE in cattle 10,11 , with a more rapid rise in overall infectivity early in the incubation period in a wide range of tissues (for example, spleen and lymph nodes). The sheep±human infectiousness pro®les (Fig. 1b) adjust for tissue-speci®c usage in food 12 and the effect of the 1997 speci®ed risk materials (SRM) ban in sheep. The distribution of the BSE incubation period in sheep is not well characterized, but on the basis of the limited available data, we used letters to nature 420 NATURE | VOL 415 | 24 JANUARY 2002 | www.nature.com 0 1 2 3 4 5 6 7 8 1980 1983 1986 1989 1992 1995 1998 Year Infectivity entering food supply (per year ×10 3 ) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 BSE feed risk BSE cases Infectivity BSE feed risk d 0 0.2 0.4 0.6 0.8 1 9 8 7 6 5 4 3 2 1 0 10 Years Survival probability 0 0.1 0.2 0.3 0.4 0.5 IPD density Survival probability Incubation period distribution c 10 –4 10 –3 10 –2 10 –1 1 0 6 12 18 24 30 36 Incubation time (months) Infectiousness Human infectivity before SRM Human infectivity after SRM Human infectivity with offal ban Sheep–sheep infectivity b 10 7 10 6 10 5 10 4 10 3 10 2 10 –4 10 –3 10 –2 10 –1 1 Prevalence (%) Sample size Upper 95% bound on prevalence if no cases detected Sample size for 95% chance of observing at least 1 case a 0 1 2 3 4 1997 1999 2001 2003 10% maternal No maternal Figure 1 Epidemiological inputs to transmission model. a, Relationship between sample size and detectable prevalence in screening studies. b, Infectiousness of sheep as a function of time from infection (see Methods). Sheep and human pro®les are separately normalized to give maxima of 1. c, Survival probability of sheep as a function of age, and assumed incubation period distribution (IPD) of BSE in sheep. The survivorship function was estimated from annual data from the June census, slaughter, export and disappearance statistics, and data on the seasonality of lamb slaughter. d, Before mid-1988, both reported and unreported clinical BSE cases could be used for food (red-shaded curve). After that time, BSE was made noti®able and cases were destroyed, so we assume none entered food. The blue-shaded curve represents the rate of slaughter (per year) of pre-clinical infected cattle weighted by infectiousness relative to disease onset, assuming infectiousness grows at the exponential rate of 4 per year. The over-30- month scheme ended most bovine exposure in 1996, with estimates of residual levels (inset) being dependent on the extent of maternal transmission of BSE. The solid black curve represents the estimated infection hazard to cattle and sheep posed by infectivity in contaminated feed, relative to the maximum level reached in 1988. © 2002 Macmillan Magazines Ltd