Warmwell note:The authorities say that so-called pre-emptive culling was necessary. We still hear the same arguments: these animals cannot be termed "healthy" animals, they were very likely to have been incubating the disease. However, Nicola Morris has collected evidence (below) to show that the massive overkill of animals was indeed unnecessary.
Had they been incubating disease, as the government voices defiantly assert, it would have been obvious at the time of slaughter since delays in slaughter times meant that the disease had time to develop if it was going to. It did not.
We now know that many of these pre-emptive culls (40%) took place around so-called "IP"s where no laboratory evidence was found (Pirbright tests are at least 90% reliable and probably almost foolproof in fact)
The modellers' predictions about likely infection were subsequently found to be very inaccurate. As reported in October in Nature 4/10/2001 vol 413, when the epidemic data was reanalysed in July, the Imperial team found that the spatial relationship between index cases and secondary infections was not as they had previously believed. They had overestimated the risk of exposure on contiguous farms.
The Reality of the Pre-emptive cull in UK 2001 FMD epidemic
Summary of the slaughter statistics for UK 2001 FMD epidemic
(nos) (%) Number of premises with positive
laboratory test results
Premises culled 10509 100 1324 (12.6%)
Premises with suspected
evidence of FMD 2283 22 1324 (58 %)
as a precaution 8226 78 0
Of the premises identified as having suspected evidence of foot and mouth disease
1981(87%) were tested.
Of the premises culled as a precaution (pre-emptive cull) less than 10 % were tested. DEFRA are of the opinion that a significant number of these farms were incubating the disease at slaughter, this is unlikely. For 3 reasons:-
1.delays in slaughtering these farms were such that by the time the animals were
slaughtered if they had been exposed to FMD virus, clinical evidence of disease would
have been apparent.
2.many of these farms were contiguous, or dangerous contacts, to the 42% of 'infected
premises' where foot and mouth lesions were identified but where no laboratory evidence
of disease has been found.
3.farms within 1.5 km of an outbreak were predicted to, according to the mathematical
models, have a 17% chance of contracting FMD, or an 83% chance of not contracting
1.Pre-emptive cull - delays in slaughter
Contiguous cull example
Example given by the Chief Scientific Adviser which has been extracted from NAO Inquiry report (18/06/2002) (3.93 p 70)
Date premises infected 25/03/01
Date culled 28/03/01
Date clinical signs would have appeared
if premises had not been culled 30/03/01
The priniciple is correct but the reality is somewhat different.
The date most contiguous farm would first be exposed to infection will be the day before
clinical signs appear on the infected premises (25 March).
In this epidemic most infected premises were found the day after clinical signs appeared. So the infected farm and the contiguous farm would have been identified on 27 March.
Slaughter times for contiguous farms for the 2001 epidemic are as follows:-
(reference: PQ 5478 24/04/02)
Slaughter within - 48 hrs 137 ( 5%) 29 March
stated hours of - 72 hrs 710 (24%) 30 March
identification - 96 hrs 620 (21%) 31 March
- > 96 hrs 572 (19%) After 1 April
->120 hrs 927 (31%) After 2 April
Thus, the 1-2 day delay in identifying the infected premises and its associated contiguous farms coupled with the 4 to 6 day delay in slaughter on contiguous farms means that: 24% of contiguous farms were slaughtered on the day, and 71% one or more days after, clinical signs would have been apparent if the animals had been exposed to FMD virus.
Notes:- 171 infected farms were not identified until after FMD lesions had healed, and the
only laboratory evidence of disease was antibodies
-there were similar delays in identification and slaughter times for farms that were
deemed to be dangerous contacts.
2.Pre-emptive cull - distribution of Laboratory positive infected premises
On only 58% of premises, where foot and mouth lesions were identified, was evidence of the disease found at laboratory test.
The distribution of laboratory positive premises is not uniform throughout the areas of the country affected by FMD, ranging from 11% (Wigtown) to 100% (Warwicks Northampton)
That is, in Wigtown 11 % of the infected premises returned a positive laboratory test result.
The number of pre-emptive culls per infected premises also varies significantly from county to county, ranging from 17 pre-emptive culls per infected premises (Anglesey) to 1 or less per infected premises (Essex, Oxford & Wiltshire).
Assuming the number of pre-emptive culls per infected premises (within each individual county) is constant, regardless of the laboratory status of each infected premises; for the epidemic as a whole, of the 7296 premises culled as a precaution:-
1736 were linked to a laboratory negative infected premises
1030 were linked to an untested infected premises
4529 were linked to a laboratory positive infected premises
(Note: since this data was compiled DEFRA have added a further 873 premises to the pre-emptive cull total, but as yet have not stated which counties these premises are located)
3. Mathematical models
a) Probability of exposure to disease
In March 2001, the team from Imperial College analysed the relationship between index infections and secondary infections:
'Contact tracing for all FMD-affected farms has produced unique data on the spatial scale of disease transmission (provided by MAFF). Clearly demonstrating that farms closest to index cases of FMD are at greatest risk of infection.'
'We estimate that farms 0.5, 1 and 1.5km from a single farm affected by FMD have a 26, 6, and 2% chance, respectively, of becoming infected.'
When the epidemic data was reanalysed in July ( published -Nature 4/10/2001 vol 413)
The Imperial team found that the spatial relationship between index cases and secondary infections was not as they had previously believed:-
'The newly estimated spatial kernal differed significantly from that previously derived from the infectious contacts identified by DEFRA with considerably more long distance transmission events being predicted. This implies significant biases in the DEFRA contact-tracing process, with closer contacts being more easily identified. '
'The median distance of the newly estimated kernal is about 4km.'
The median distance of the original kernal was about 2 km.
b) Other flaws
Subsequent analysis of epidemic data reveals that:
-up to 1/04/01 of the 909 farms identified as infected only 549 (60%) were positive at
laboratory test (reference: PQ 6650 22/06/02).. It is not clear from the data how many if any
of the 909 farms were not tested. Farms with negative laboratory tests were not removed
from the data used to develop the models: this omission caused the models to over predict
-for many of the farms first infected detailed tracings have revealed not only a source of
infection but also a mode of spread. (VLA Study completed July 2002 - Data witheld)
Thus, it can be concluded that farms closest to an index premises were less at risk of contracting FMD than previously predicted and those most at risk could have been identified and slaughtered or closely monitored as necessary.
Given the delays in identification and slaughter of the pre-emptive cull farms, the fact that around 40% of this slaughter occurred around negative and untested infected premises and the over estimation of risk of exposure to disease by the models, it is highly likely that most of the pre-emptive culling carried out during the UK 2001 epidemic was unnecessary. Significantly the statistical facts are supported by the extensive, detailed anecdotal evidence reported across the country.
However, until detailed independent analysis of the 'truly' infected premises is permitted by DEFRA the true scale of the epidemic can not be established; disease transmission remains a mystery and more importantly the most appropriate disease control strategy for the future can not be determined.
We can say for certain that the scale of the culling in UK 2001 epidemic was unprecedented; the scale and nature of the epidemic is, as yet, unknown.