[REPLY TO PROFESSOR ANDERSON - VET TIMES, Vol 32 (N0 5) 11th Feb 2002]
Professor Anderson's most recent attack on vets who called for vaccination during the recent FMD epidemic (Symposium on Simulation and Modelling applied to Medicine, Institute of Physics, London, January 2002; reported by BioMedNet News Jan 24th 2002) called into doubt both the integrity of the Veterinary profession and the science on which it is founded. His derisory comments are arrogant, insulting and defamatory and call for a refutation, writes Dr Susan Haywood, honorary senior fellow in the department of veterinary pathology at the University of Liverpool.
Firstly with regard to his dismissal of vaccination and "lessons learned" from the 1967/68 FMD outbreak. The findings from the Report of the Committee of Inquiry on Foot and Mouth Disease 1968, were a comprehensive and balanced investigation into establishing the factual evidence available at the time, with respect to the disease and its control world-wide. Among its more important recommendations the Committee advocated that "contingency plans for the application of ring vaccination be kept in constant readiness". With prescience, Anthony Cripps QC supplied an addendum to the effect that in his opinion, no matter how stringent the border controls, there was no way that the risk of reintroduction of infection could be reduced to zero and that an FMD outbreak could be expected approximately every 20 years. He recommended "the immediate application of ring vaccination to any outbreak which may occur." Why didn't we then vaccinate in 2001? Admittedly the 1991 declaration of the EU as a disease free zone meant that we could no longer take a unilateral decision in that respect. However, emergency or protective vaccination in event of disease was allowed for, albeit with the penalty of the vaccinates being culled and no re-entry to disease free status for 12 weeks (12 months if the vaccinates were allowed to live). In addition protective vaccination with exemption from cull was allowed for rare or endangered species, or animals valuable for research purposes. I applied to MAFF for protective vaccination for my rare breed of sheep in early April since they fulfilled the dual criteria. I was told in May that no decision had been reached. Later in July my solicitor applied formally on my behalf for protective vaccination. Despite having been awarded a valuable research grant by the Wellcome Trust to study the Ronaldsay sheep, DEFRA refused my request. This intransigence seems perverse and quite outside the spirit of EU policy.
Although Anderson now promotes the vaccination of livestock to prevent a future epidemic he still insists that in the last outbreak this would have prolonged the epidemic. What possible evidence has the Professor for his opinion particularly in light of the Dutch experience who did vaccinate and regained disease free status months ago! Could it be that once the cull policy had been decided there must be no change or deviation in case of loss of face!
Now to consider Professor Anderson's vaunted superiority of the modelling method used to determine policy. The criterion by which a model can be said to be a sound model of the situation is how well it can predict the future outcome. In March, the Chief Scientist predicted that the FMD outbreak would be over by early June, just in time for the election. In fact, the number of confirmed cases remained steady between three to five per day from early May until mid September. These facts suggest strongly that Anderson's model was unable to provide a valid prediction.
The type of parametric modelling used relies on identifying the relevant factors that contribute to the causes of the spread of the disease. The art of modelling is to ensure that just sufficient factors are identified to make the model valid - too many and the model becomes merely a description of the situation and becomes unusable (rather like having a road map of the British Isles on a 1 to 1 scale); too few and the model is no longer sound, so that it is unable to make the predictions it should. The other important point about modelling is that the data for each factor is itself accurate and representative. If it isn't, the model becomes unsound - it makes predictions that don't hold. From the Professor's admissions ("the quality of the quantitative data was appalling"), it seems that his difficulty lay in being unable to identify the relevant factors (or parameters) for his model, and secondly, for those that he used, the data wasn't sufficiently accurate to be of use. So, he would have had to rely on possible models of disease that had been used more or less successfully in the past and tried to match their predictive powers retrospectively on data already produced during the current FMD outbreak. The closest fit appears to be, on his own admission, that of human sexually transmitted disease, saying, rather naively, "it wasn't all that difficult to transfer this to a veterinary context". Thus, he had a model which was syntactically correct - it could make predictions - but was semantically incorrect - the predictions it drew were meaningless - in fact little more useful than examining chickens' entrails! It would be interesting for Professor Anderson to tell us just how a category of diseases caused by a disparate variety of organisms, including bacteria, fungi, protozoa and viruses, transmitted almost exclusively by genital contact within human hosts; compares with a viral disease affecting a wide variety of sometimes unrelated animal species both domesticated and wild, with differing modes of transmission and varied expression within a particular species? The differences in clinical expression of the disease between cattle and sheep is that in cattle the clinical signs are obvious blisters whereas asymptomatic infections in sheep (and possibly other ungulates such as deer) are more the norm. This simple fact which has hindered diagnosis of the disease has, in my opinion, been more responsible for the onrolling spread of the disease in this outbreak than farmers' alleged failure to take adequate biosecurity precautions. In the end, the disease was contained more by the application of a 'scorched earth' policy than any use of predictive models. In the Penrith spur this was effected by killing over 80% of the animals in the region, in conjunction with an intensive disinfection policy applied to all vehicles entering and leaving the valley, together with continuously applied disinfectant to vulnerable exit points. Thus, any viral particles that might have been transmitted by road were effectively destroyed. The moorland , being mainly peat bog had an acidity comparable to that of the disinfectant being applied to the road; it is possible that this prevented the spread of the disease on the high moors. This is only speculation since no research has been carried out into the behaviour of this strain of virus in the present outbreak. DEFRA's policy of cull means that no valid data was obtained. So, when the next outbreak occurs, Professor Anderson will be back in the same position of having no quantitative data. Apart from the lack of prediction displayed by the modelling, the unquantifiable destruction of our genetic resource arising from the cull policy has been entirely overlooked by the modellers. Operating no doubt from the contention that science is 'value free', the data incorporated into the models did not distinguish between the real value of the actual animals on the farm, for example between the pedigree breeding stock and the cross bred animals destined for the abbatoir. The former are the result of decades of breeding and as such gene banks in their own right of incalculable value, the latter have only their current market value. The result of the cull has meant the irrecoverable destruction of pedigree blood lines established over a hundred years, perhaps more - in Swaledale alone there is a 50% loss of pedigree breeding stock and this could be repeated many times over around the country.
I return finally to the 1968 Inquiry that stated that slaughter is "a crude and primitive way of dealing with the disease. We recognise the mental anguish of those who suffer..the shattering disaster, not computable in terms of money, that it may bring to a farmer who has to see the work of a lifetime destroyed in a day". What sort of alternative universe does Professor Anderson and his kind inhabit that the welfare of man and beast, the future sustainability of UK agriculture, even biomedical research takes second place to the short term commercial dictates of a niche export market?
In conclusion I suggest that it would be more apposite if Professor Anderson instead of maintaining his statement that "a lot of experimentation in veterinary science needs to be done again" takes his own advice and applies it to his own work.
Susan Haywood PhD,BVSc,MRCVS
Jan: 29th 2002