A critical analysis of the DEFRA report "Origin of the UK FMD Epidemic 2001"
by Alan Beat
Jim Scudamore's report published in June 2002 sets out his view of the start of the FMD epidemic, based on the considerable efforts made to establish the possible source and early spread of the disease. However, the full data on which the reports conclusions are based have not been released into the public domain. This paper therefore examines the limited information given in the report, and applies questioning analysis to the conclusions reached - a review process to which all science is normally subject prior to publication.
There is no evidence of clinical disease at Burnside Farm earlier than 12th February
There is no evidence that sheep sent from Prestwick Farm to Hexham market on 13th February were infected as a result of airborne spread from Burnside Farm
Other important issues raised
Pigs can show clinical signs within 24 hours of exposure to virus
Pigs recover quickly from the disease to show no residual signs of infection
There is clear evidence of rising infection at Prestwick Farm to invalidate computer modelling assumptions
Speed and effect of FMD in pigs
The effect of FMD on pigs is very much faster, and milder, than previously published information had suggested. Investigations at Cheales abattoir in Essex, where FMD was first reported, showed that batches of healthy pigs arriving from uninfected farms on the 16th, 17th and 19th February had developed foot lesions within 24 hours of arrival
"suggesting a very short incubation period."
The report continues:
"Pigs usually recover quickly especially once the blisters on the feet have ruptured and provided the horn of the hoof does not separate and there is no secondary infection. Affected but recovered pigs would not be expected to show clinical evidence of disease at post mortem examination."
The report describes the initial investigation at Burnside Farm as follows:
"At the clinical inspection on 22 February, there was little obvious sign of vesiculation in pigs in Shed 1 but the presence of "freckles" lesions on the snouts, indicative of long standing FMD infection, was recorded. None of these pigs appeared notably unwell, distressed or lame. Several, however, showed a growth check line on the hoof, often well down the hoof . . . . . 88% of the 241 pigs blood sampled were seropositive for FMD Type O antibodies. This included pigs that showed no discernible FMD lesions."
Important side issue - this information, that pigs recover quickly to show no physical effects afterwards, stands in complete contradiction to all previous information from DEFRA, of a "killer plague" causing trotters and tongues to disintegrate and causing significant loss of economic value to the carcase.
Infection at Cheales abattoir
Tracing of all farms that had previously sent pigs to this abattoir led investigators to Burnside Farm which:
". . . had sent sows to Cheales abattoir that arrived on the nights of 8/9 and 15/16 February and which were slaughtered on the mornings of 9 and 16 February. Burnside Farm had not sent pigs to the abattoir for a period of 4 weeks from 12 January until 8 February."
The report speculates that disease may have been introduced into the abattoir from Burnside on 8th February but passed unnoticed. However, there is no evidence that the small number of outbreaks linked with this abattoir were exposed to disease before 15th February.
Infection at Burnside Farm
At this farm, the clinical inspection on 22nd February found that in shed 3:
"All the pigs in pens 14 & 16 were clearly unwell. They appeared miserable, were huddled together and reluctant to rise. FMD lesions of varying ages were found in the pigs, ranging from 9 to 10 days old to one-day-old"
A second investigation was carried out on 24th February by FMD experts from IAH Pirbright and found:
". . . approximately 90% of the 527 pigs on the farm had lesions suggestive of FMD. Many pigs exhibited lameness, with the feet showing separation of old horn from the underlying tissue. Vesicles were found on the snouts of some pigs and it was from these that FMD virus of an identical strain to that found at the Essex abattoir was recovered"
The report continues:
"pigs with the oldest lesions were found in Shed 2, where a high proportion exhibited lesions estimated at 10 and 12 days old. In contrast, some of the pigs exhibited lesions only one to two days old, suggesting that disease was still active and spreading within the herd"
This evidence identifies the first appearance of clinical signs at Burnside Farm as around 12th February. Applying the very short incubation times demonstrated at the Essex abattoir predicts a date of first exposure to disease around 10th February, and a date of first infectivity around 11th February.
The report speculates that an incubation period of 2 to 14 days may indicate an earlier exposure to FMD on these premises; and that the first pigs affected could already have been slaughtered at the Essex abattoir to obscure an exposure earlier still. However, no evidence is offered to support the longer incubation period or the speculative events.
Infection at Prestwick Hall Farm
Disease was first reported here on 23rd February. FMD experts from IAH Pirbright investigated these premises on 25th February to find:
"Four groups of cattle housed in a single shed showed evidence of FMD but young stock housed separately did not. Of the four affected groups, 38 heifers in two groups in the southwest corner of the shed all had lesions but only 13 of 37 bulls housed in two groups on the other side of the shed were showing lesions. The oldest lesion (9 days) was observed in a heifer when the livestock were slaughtered on 25 February. The other animals in the group had lesions aged 3 days whilst the majority of affected bulls had lesions aged 1-3 days. If the cattle were the first to be infected, this would suggest that the index case on the farm developed clinical signs of disease on about 16 February."
Important side issue - note the very clear "within-farm epidemic" described here, with spread over time from an "index animal" firstly to others in adjacent pens, then onward to other pens within the building, while well-separated groups remained uninfected. This pattern defines that farm infectivity must increase over time, a crucial fact that changes the computer modelling predictions to NOT support the culling of Contiguous Premises (see appendix below). Detailed studies of "within-farm epidemics" have been published in scientific journals over recent years (see appendix) but despite this knowledge, the computer modelling teams all falsely assumed constant farm infectivity to arrive at the wrong predictions.
There were also sheep on the farm:
". . .until 12 February a group of 26 ewes were grazed during the day and housed at night in an outside bull pen close to the affected heifers which had the oldest lesions. . . . 15 of the ewes, together with a ram from a separate field and 3 sheep from the owner’s other premises, were sent to Hexham market in Northumberland on 13 February. Examination of the 11 remaining ewes from this group when they were slaughtered on 25 February, revealed healing foot lesions consistent with FMD in five animals; all 11 were seropositive for FMD."
The report continues:
"The sale of 19 sheep, some of which were incubating disease, from Prestwick Hall Farm to Hexham market on 13 February represents the initial stage of the widespread dissemination of FMD virus. Ten of these sheep were subsequently sent for re-sale to Longtown Market on 15 February and the movement of sheep through this market and those held on later dates, resulted in the introduction of infection to the sheep dense areas of England and Wales and the southern borders of Scotland."
Prestwick Hall Farm lies 5km north-east of Burnside Farm and the report concludes that:
"the likeliest source of infection . . . was airborne spread of virus from infected pigs belonging to Burnside Farm"
The sheep concerned had exhibited two periods of lameness that the owner had treated with footbaths on 6th, 10th, 12th and 20th February. The report speculates that this lameness could have been caused by unrecognised FMD. However, no age estimate is given for the "healing foot lesions" seen in 5 out of 11 remaining ewes at slaughter on 25th February.
"In summary, it is considered likely that the affected sheep were infected concurrently with, and possibly earlier than, the affected cattle. None of the other 339 sheep on the premises showed clinical or post mortem evidence of disease."
Timeline of possible airborne transmission between Burnside and Prestwick Farms
From the evidence above, the estimated earliest date for airborne spread to commence from Burnside Farm was around 11th February. The report states:
"During early February there were suitable weather conditions not only for spread of virus to Ponteland but also to other farms in the area. The significance of airborne spread in causing FMD at PrestwicK Hall Farm would be dependent on the timing of virus excretion from the infected pigs at (Burnside Farm)"
At Prestwick Hall, the evidence indicates an incubation period for cattle of six days between exposure and first clinical signs. The one heifer with lesions of 9 days old is presumed to have been the "index case" that infected others with lesions 3 days old or less, the time lapse therefore being six days. The estimated date of first clinical signs on 16th February gives an estimated date of exposure to disease on 10th February. This corresponds closely to the evidence of first infectivity at Burnside Farm of 11th February. The concentration of disease in the south-west corner of the building also supports possible airborne transmission, these cattle making first contact with an airstream from the direction of Burnside Farm.
However, the report concludes that the group of 26 sheep were infected at the same time as the cattle, or possibly earlier, by the same route of airborne spread. The evidence presented argues that this did not happen. Crucially, none of the other 339 sheep on the farm were infected. Furthermore, cattle are far more susceptible to airborne infection, by a factor of 15.2 to 1 when compared with sheep (Keeling et al). The evidence that only one heifer became infected initially implies a low level of exposure that was most unlikely to infect sheep at all.
It is far more likely, and fits the available evidence, that this group of sheep became infected through close contact with the index heifer when penned for the night. In this scenario, the earliest possible exposure of the group to disease occurred as the index heifer became infective, shortly before the appearance of clinical signs on 16th February, perhaps on the 14th/15th February - but early on 13th February, 15 of the group were sent to Hexham market. The report describes these 15 as
but this would not have been possible, as the index Heifer would not have become infective by that date.
The only evidence that supports a possible earlier infection date for this group is the lameness in early February, postulated to have been unrecognised FMD. However, lameness in heavily-pregnant ewes that are housed overnight is a very common condition and is far more likely to have been unconnected with FMD.
For this group of sheep to have been "incubating disease" when sent to Hexham market, they would need to have become infected by airborne means at an earlier date than the adjacent cattle, without any other sheep on the farm becoming infected at all, and before the date at which Burnside Farm (the presumed source) itself became infective.
In conclusion, if movement of these sheep through Hexham market on 13th February was indeed "the initial stage of the widespread dissemination of FMD virus", the source of their infection is not demonstrated within the report.
Alan Beat 11th July 2002
Contact e-mail: email@example.com
From the Imperial College modelling team's paper: "The FMD Epidemic in Great Britain: Pattern of Spread and Impact of Interventions" by NM Ferguson, CA Donnelly and RM Anderson, Science Vol 292, 1155-1160, 11th May 2001:
"We assumed constant infectiousness from 3 days after infection until slaughter, for an average of eight infectious days."
"Our analysis shows that . . . slaughtering on all (infected) farms within 24 hours of case reporting can significantly slow the epidemic . . . and results in rapid control if we assume that infectivity increases throughout the time from infection to slaughter
"If r is greater than one (i.e. farm infectivity is rising), ring culling still accelerates the decline of the epidemic, but at the cost of a larger cull than rapid index case slaughter alone"
Some papers published in recent years on within-farm epidemics:
"The use of vector transmission in the modelling of intraherd FMD" - AM Hutber and RP Kitching, Environmental and Ecological Statistics 3, 245-255 (1966)
"Control of FMD through vaccination and the isolation of infected animals" - AM Hutber, RP Kitching and DA Conway, Tropical Animal Health and Production, 30 (1998) 217-227
"Predicting the level of herd infection for outbreaks of FMD in vaccinated herds" - AM Hutber, RP Kitching and DA Conway, Epidemiology and Infection (1999) 122, 539-544
"The role of management segregations in controlling intra-herd FMD" - AM Hutber and RP Kitching, Tropical Animal Health and Production, 32 (2000) 285-294