ARTICLE IN PRESS
The Veterinary Journal xxx (2004) xxx–xxx
Foot-and-mouth disease: the vital need for collaboration as an aid to disease elimination
Foot-and-mouth disease (FMD) is endemic in many parts of the world and "globalisation" has ensured that those countries with FMD-free status are increasingly at risk of occasional viral introductions – even if they take active measures to prevent such incursions. In countries such as the United Kingdom, which maintain FMD-free status without prophylactic vaccination, a "stamping out" policy (i.e. killing the host) is the current means of dealing with occasional introductions of the FMD virus (FMDV).
Following the experience of the 2001 FMD epidemic in the UK, emergency vaccination "to live" has become an important aspect of future contingency planning. With a stamping out policy alone it is usual to kill animals with all speed on infected premises (IPs) where disease has been diagnosed on either clinical grounds or from positive laboratory testing. Livestock regarded as dangerous contacts (DCs) on the basis of epidemiological tracing are also killed. Movement restrictions and strict biosecurity are additional key features of control measures.
Throughout an epidemic accurate data collection and analysis is crucial, particularly so if any type of modelling is to form part of the decision making process. For all three aspects (data collection, analysis and com-puter-based simulations) close collaborative effort is required to provide the best scientific approach to elimination of disease. During the 2001 FMD epidemic, census data were inaccurate and the collection and analysis of field data were imperfect. Mathematical models were developed during the course of the epidemic, partly as a substitute for poor information, but also because the early indications were that the FMD contingency plan was inadequate (Taylor, 2003). It is, however, important to emphasise that many of the perceived problems in the early stages resulted from the lag between introduction of FMDV into the UK and recognition of its presence. Estimates vary, but Mansley et al. (2003) concluded that 79 premises had been exposed to FMD infection before confirmation of the disease on the 20th February 2001. Haydon et al. (2003) suggested that many premises (from <30 to >80) may have actually been infected by 20th February and a further 40 might have become infected before the introduction of the national movement ban on February 23rd. Gibbens and Wilesmith (2002) and Alexanderson et al. (2003) postulated that infection could have been present in the UK from late January 2001.
The background to the 2001 FMD epidemic is that on February 19th 2001, FMD was suspected in pigs at an abattoir in Essex and confirmed the following day as the sub-type O PanAsiatic strain of FMDV. The origin of the infected pigs was identified as a pig farm at Heddon-on-the-Wall, Northumberland on February 22nd where detailed examination of the pigs on February 24th revealed FMD lesions that were adjudged to be at least 12 days old (Mansley et al., 2003). Precisely how the FMD virus reached this pig farm has not been identified, although untreated pig swill has been implicated.
On February 23rd the disease was confirmed on a mixed sheep and beef cattle farm at Ponteland, some 5 km from Heddon-on-the-Wall; the oldest FMD lesions (estimated to be nine days old on February 25th) were found on a heifer, but serum samples taken from 226 of 355 sheep on the same day indicated that four were antibody positive and eight were virus positive (Mansley et al., 2003).
Movement of sheep to market from the Ponteland farm and subsequent dealing raised the possibility that multiple livestock movements might have already disseminated infection widely by the time of disease confirmation on February 20th. These fears were realised by the detection of disease in several new areas as far south as Devon within a week of the original confirmation of disease on the 22nd February. Subsequent short range modelling of the wind-borne spread of the prevalent strain of the virus indicated that infected pigs at Heddon-on-the-Wall might have excreted sufficient virus to infect livestock on the farm in Ponteland (Gloster et al., 2003).
In this brief historical introduction the complexities of the field situation are fully realised – unknown origin and date of viral introduction; undefined viral characteristics; weather conditions that favoured virus survival; potential spread prior to identification of the index case by, for instance, animals, people, equipment and vehicles; possible airborne transmission; and involvement of numerous animals of different species and susceptibilities in widely varying regions of the UK. Complex biological situations of this type do not lend themselves to an intra-epidemic mathematical modelling approach, particularly when such models are invariably built on a number of assumptions, some of which may be inaccurate. In a comprehensive review of the use of models in informing disease control policy, Taylor (2003) concluded with respect to FMD control that ‘‘tactical decision making should be based more on real veterinary intelligence than on predictive modelling’’.
In this issue of The Veterinary Journal, Kitching, Hutber and Thrusfield (Kitching et al., 2004), examine some of the clinical and epidemiological factors that are of relevance in predictive modelling. The authors stress the importance of farm level analysis, focusing on intra-herd transmission as the starting point for any FMD model. The within-herd parameter of first day incidence (FDI) 1 and regional analysis using the in-ter-herd parameter of first fortnight incidence (FFI) 2 have the merit of being directly measurable. Using these parameters it is possible to predict prevalence, to model within farm infectivity and regional spread and to highlight the differences between species with some accuracy. Control policies can then be targeted more effectively if, for example, large dairy farms demonstrate high FDI and FFI.
During the 2001 epidemic in Cumbria, dairy farms faced a five-times greater risk of infection than other farms (Taylor et al., 2001) and yet sheep were the species most intensively slaughtered. It would seem that host species differences were not taken fully into account and the risk factors were imperfectly modelled. The 3km pre-emptive slaughter of sheep in Cumbria, Dumfries and Galloway announced on March 15th 2001 was, perhaps, not the best means of dealing with FMD in a species in which extensive husbandry, subclinical infection and low disease transmission are the
First day incidence (FDI) is a measure of the number of animals showing clinical signs on the first day of disease diagnosis for a given infected premise (IP), alternatively, prevalence at first report can be used.
First fortnight incidence (FFI) is the number of identified infected premises within a given region during the first fortnight of the regional focus, alternatively, prevalence at first report can be used.
norm. Kitching and colleagues (2004) discuss other ways in which the problems posed by sub-clinical infection can be addressed, which would result in the loss of fewer uninfected animals.
Intra-epidemic predictive modelling played a disproportionate role in the management of the 2001 epidemic in the UK. For example, the deterministic, state-transi-tion type model of the Imperial College group published in May 2001 (Ferguson et al., 2001a), based on differential equations, did not differentiate between host species, or take into account different farm types and assumed constant infectiousness from three days after infection until slaughter. The latter situation does not occur in the field, as virus excretion from a group of animals increases over time from first infection to slaughter. In addition, farms closest to the index cases of FMD were declared to be at greatest risk of infection, but this conclusion was incorrect and based on inaccurate input data and the observation ‘‘that, on average, animals on 34% of premises within a radius of 1.5 km of infected premises came down with FMD’’ (Anderson, 2002b). It was not possible to identify the source of infection for the majority of infected premises (Gibbens and Wilesmith, 2002), so it was assumed that the probable source was one of the recent nearby (within 3 km) infected premises
(i.e. due to "local spread"). It would have been more scientifically robust if the source of infection had been reported as "unknown".
In a later paper published in October 2001 (Ferguson et al., 2001b) the Imperial College group acknowledged some of the imperfections of their earlier publication, noting that the newly estimated spatial kernel differed significantly from that derived previously in that considerably more long-distance transmission events were predicted (median distance 4 km). Taylor et al. (2004) have since shown that in over half the cases in the most heavily affected parts of Cumbria, occurrence of the disease was not directly attributable to a recently infected premises being located within 1.5 km.
On these and other grounds the justification for a contiguous culling (CC) policy, ‘‘founded on a statistical concept’’ (Anderson, 2002b), with a 48 h target (report to pre-emptive slaughter) can be challenged. Failure of the informally constituted FMD Official Science Group to agree a definition of "contiguous" more than a month after the contiguous culling policy was introduced did not inspire confidence; further, those conducting the Lessons to be Learned Inquiry stated that ‘‘We have been unable to establish the precise rationale for the target of 48 hours, nor ascertain the source of that time scale
Incidentally, contiguous slaughter was rarely achieved within the 48 h target in the field and many species of animal on contiguous premises remained in good health and went beyond the accepted incubation periods for FMD before they were killed.
Honhold et al. (2003)examined the effect of speed of animal slaughter on IPs and the intensity of culling on other premises on the rate of spread of FMD and concluded that the need for a contiguous cull was not supported by analysis of the field data. Similar conclusions were reached by Taylor (2003) and Honhold et al. (2004) and their analysis of the data for Cumbria, the worst affected region where almost half the IPs in the epidemic occurred, indicated that the epidemic peaked prior to full implementation of the DC, contiguous and 3 km culling policies. Kitching and colleagues (2004) discuss what they term "ineffective pre-emptive culling" in some detail and their analysis clearly demonstrates the importance of field veterinary epidemiology and effective collaboration in disease control strategies; a view echoed by Haydon et al. (2004).
Mathematical models were a novel and contentious means of directing policy in 2001 and there have already been a number of publications assessing their strengths and weaknesses (Haywood and Haywood, 2002; Green and Medley, 2002; Kao, 2002; Lusmore, 2002; Taylor, 2003; Ap Dewi et al., 2004). The paper by Kitching and colleagues (2004) provides further evidence that in-tra-epidemic predictive modelling should not determine control polices.
Sheila M Crispin
Department of Clinical Veterinary Science University of Bristol, Langford House Langford, Bristol BS40 5DU, UK E-mail address: firstname.lastname@example.org
Alexanderson, S., Kitching, R.P., Mansley, L.M., Donaldson, A.I., 2003. Clinical and laboratory investigations of five outbreaks during the early stages of the 2001 foot-and-mouth disease epidemic in the United Kingdom. Veterinary Record 152, 489–496.
Anderson, I., 2002a. Foot and mouth disease 2001: lessons to be learned inquiry Report. London, The Stationery Office,
Anderson, I., 2002b. Foot and mouth disease 2001: lessons to be learned inquiry Report. London, The Stationery Office,
Ap Dewi, I., Molina-Flores, B., Edwards-Jones, G., 2004. A generic spreadsheet model of a disease epidemic with application to the first 100 days of the 2001 outbreak of foot-and -mouth disease in the UK. The Veterinary Journal 167, 167–174.
Ferguson, N.M., Donnelly, C.A., Anderson, R.M., 2001a. The foot-and-mouth epidemic in Great Britain: pattern of spread and impact of interventions. Science 292, 1155– 1160.
Ferguson, N.M., Donnelly, C.A., Anderson, R.M., 2001b. Transmission intensity and impact of control policies on the foot and mouth epidemic in Great Britain. Nature 413, 542–547.
Gibbens, J.C., Wilesmith, J.W., 2002. Temporal and geographical distribution of cases of foot-and-mouth disease during the early weeks of the 2001 epidemic in Great Britain. Veterinary Record 151, 407–412.
Gloster, J., Champion, H.J., Sørensen, J.H., Mikkelsen, T., Ryall, D.B., Astrup, P., Alexandeson, S., Donaldson, A.I., 2003. Airborne transmission of foot-and-mouth disease virus from Burnside Farm, Heddon-on-the Wall, Northumberland, during the 2001 epidemic in the United Kingdom. Veterinary Record 152, 525–533.
Green, L.E., Medley, G.F., 2002. Mathematical modelling of the foot and mouth disease epidemic of 2001: strengths and weaknesses. Research in Veterinary Science 73, 205–210.
Haydon, D.T., Chase-Topping, M., Shaw, D.J., Matthews, L., Friar, J.K., Wilesmith, J., Woolhouse, M.E.J., 2003. The construction and analysis of epidemic trees with reference to the 2001 UK foot-and-mouth disease outbreak. In: Proceedings of the Royal Society of London B, 270, pp. 121–127.
Haydon, D.T., Kao, R.R., Kitching, P., 2004. The UK foot-and-mouth disease outbreak – the aftermath. Nature Reiews Microbiology 2, 675–681.
Haywood, S., Haywood, G., 2002. Modelling and FMD. Veterinary Times (British Cattle Veterinary Association Congress Edition), 16th–18th July. Veterinary Business Development Ltd., Peterborough, pp. 6–8.
Honhold, N., Taylor, N.M., Mansley, L.M., Paterson, A.D., 2003. The effect of speed of animal slaughter on infected premises and the intensity of culling on other premises on the rate of spread of foot and mouth disease. In: Proceedings of the Society for Veterinary Epidemiology and Preventive Medicine, Warwick, UK, March 31–April 2, pp. 183-194.
Honhold, N., Taylor, N.M., Mansley, L.M., Paterson, A.D., 2004. Relationship of speed of slaughter on infected premises and intensity of culling on other premises to the rate of spread of the foot-and-mouth disease epidemic in Great Britain, 2001. Veterinary Record 155, 287–294.
Kao, R.R., 2002. The role of mathematical modelling in the control of the 2001 FMD epidemic in the UK. Trends in Microbiology 10, 279–286.
Kitching, R.P., Hutber, A.M., Thrusfield, M.V., 2004. A review of foot-and-mouth disease with special consideration for the clinical and epidemiological factors relevant to predictive modelling of the disease. The Veterinary Journal.
Mansley, L.M., Dunlop, P.J., Whiteside, S.M., Smith, R.G.H., 2003. Early dissemination of foot-and-mouth disease virus through sheep marketing in February 2001. Veterinary Record 153, 43–50.
Taylor, N.M., 2003. Review of the use of models in informing disease control policy development and adjustment. Report for Department of the Environment Food and Rural Affairs pp. 94.
Taylor, N.M., Hornhold, N., Mansley, L.M., 2001. Modelling the FMD outbreaks. Letter to Veterinary Record 149, 600.
Taylor, N.M., Hornhold, N., Pateson, A.D., Mansley, L.M., 2004. Risk of foot-and-mouth disease associated with proximity in space and time to infected premises and the implications for control policy during the 2001 epidemic in Cumbria. Veterinary Record 154, 617–626.