Volume 3: The Early Years, 1986-88
4. The ruminant feed ban
Discussion
Cross-contamination

Cross-contamination

4.116 It is now recognised that there was, from the outset, another factor which led to cattle becoming infected by ruminant protein after the ruminant feed ban had been introduced. This was the accidental contamination of ruminant feed, in which no animal protein had been incorporated, by other feed which contained ruminant protein. The circumstances in which this occurred and in which it then came to be appreciated are dealt with in detail in vol. 5: Animal Health, 1989-96, Chapter 2 ('The ruminant feed ban'). In short, cross-contamination occurred at feedmills which were using the same plant to produce ruminant feed, which no longer contained animal protein, and non-ruminant feed, which did contain animal protein. Cross-contamination occurred when the same vehicles were used to deliver in bulk: on one occasion non-ruminant feed and on the next ruminant feed. It occurred on farms where ruminant feed supplied before 18 July remained clinging to hoppers and other containers of bulk feed. And it occurred on mixed farms that handled, or mixed, feed for both ruminants and non-ruminants.

Was cross-contamination foreseen?

4.117 We are satisfied by the evidence, which is set out in detail in vol. 5: Animal Health, 1989-96, that once the pipeline stocks of ruminant feed containing MBM had been consumed, the principal cause of cattle continuing to become infected with BSE was cross-contamination of their feed. Ruminant feed that contained no MBM became contaminated with feed for non-ruminants that did contain MBM. Was the danger of this cross-contamination foreseen?

4.118 It was well known that, where feedmills compounded different batches of feed for different animals, cross-contamination could occur between one batch and the next. This was cause for concern insofar as the ingredients of a feed designed for one animal could have an adverse effect on another. The problem was particularly acute in respect of medicinal additives, an area governed by statutory regulation.

4.119 In 1982 UKASTA issued a Code of Practice, which was updated in 1984, recommending measures to minimise cross-contamination in feedmills with specific reference to medicinal products. ¹ We received evidence from four of the major feed compounders, in addition to evidence from UKASTA, as to the measures that they adopted to reduce cross-contamination. They all applied the UKASTA Code. More generally they told us that as and when they updated their plant, they attempted to do so in a manner which reduced the possibility of cross-contamination. All were clear, however, that a degree of cross-contamination was inevitable. ²

4.120 When the ruminant feed ban was introduced, neither officials at MAFF concerned with its introduction nor the feedmills which had to comply with it foresaw that cross-contamination of ruminant feed with non-ruminant feed might render the former infective. One reason for this was a belief that the quantity of any contaminating feed would be too small to contain an infective dose of BSE.

4.121 In a supplementary witness statement to the Inquiry Mr Meldrum said:

. . . if I, or indeed any other of the MAFF or industry representatives, had known at that time that the infective dose was really so low as to lead to cross-contamination problems, the issue would have been pursued, as indeed it was at a later stage as a result of investigations into BABs and the use of the ELISA test in the field. The lack of knowledge at the time as to cross-contamination in the context of BSE is reflected by my response to two queries raised at a cattle industry meeting on 28 June 1988. ³

4.122 The queries to which Mr Meldrum referred were raised at a meeting, which he chaired, attended by representatives of the cattle industry and by MAFF officials who included Mr Cruickshank, Mr Kevin Taylor, Mr Wilesmith and Mr Lawrence. Mr Meldrum is recorded as saying, in the course of discussion about standards in renderers:

. . . feed compounding mills presented, at worst, only a low contamination risk and would not be investigated. ⁴

4.123 To a question whether the use of MBM in fertiliser was acceptable, Mr Meldrum replied:

At present there was no evidence of a need to ban the use of meat and bone meal in fertiliser as the dose received by animals grazing the fields thus fertilised would almost certainly be too low to cause disease. ⁵

4.124 We raised with Mr Meldrum the question whether he ought to have warned the feedmills of the danger of cross-contamination. His response was:

It logically follows . . . that if, as at June 1988, the evidence was not available as to either the size of the infective dose or the extent of any cross-contamination that might occur through continued production of ruminant-derived meat and bone meal, it is unreasonable to expect me to ensure that a clear warning of the danger of cross-contamination was publicised. The existence of such a danger was not considered to exist at that time. ⁶

4.125 We asked Mr Meldrum about his understanding in 1988 of the amount of infective material a calf would have to eat in order to become infected with BSE. His answer was:

. . . we simply did not know what volume of material, what weight of material in meat and bone meal would cause disease in calves. ⁷

4.126 He and his colleagues envisaged pockets of infectivity in the MBM but did not know the size of the pockets.

4.127 Mr Meldrum was referred to a note of an earlier meeting with UKRA, UKASTA and GAFTA that he chaired in March 1988. Mr Wilesmith was recorded as saying:

The effective dose of the disease was very small, and the smaller the dose the longer it took to show itself. ⁸

4.128 Mr Meldrum said he had no recollection of that statement, and that:

That was now the view, the corporate view that we held at the time based on the advice we received from experts. We did not know. And I do not think John [Wilesmith] knew either, frankly. ⁹

4.129 Mr Kevin Taylor commented in a supplementary statement:

The risks of cross-contamination were not and could not have been foreseen in the absence of epidemiological evidence or knowledge of the effectiveness of rendering procedures and of the size of the oral infective dose of cattle tissues to cattle . . .

The possibility of cross-contamination of feed which did not include ruminant MBM by other feeds which did was not considered at that time, so far as I am aware, and on the evidence then available I can see no reason why it should have been. ¹⁰

4.130 In oral evidence Mr Taylor made the point that he did not share Mr Meldrum's knowledge of the feed industry. He had no knowledge of UKASTA's Code of Practice. He had no recollection of discussing dose at the time of the introduction of the ruminant feed ban. ¹¹

4.131 When giving oral evidence to the Inquiry, Mr Cruickshank said:

I think it is important to bear in mind that at that time and for some considerable time afterwards everybody believed that for a cow to go down with BSE, it had to ingest a massive dose of infected material. ¹²

4.132 Mr Cruickshank elaborated on this when he returned to give evidence in Phase 2 of the Inquiry. He said that he recalled many conversations in which veterinary colleagues said: 'Of course you can only pass on the disease if the animal takes in a large dose.' ¹³ He understood dose to mean 'the amount of contaminated sheep-origin material in the rations that were fed to the cattle'. ¹⁴

4.133 He was referred to evidence that Mr Wilesmith had not believed that a large dose was necessary to infect. He said:

I find it quite amazing, because I had many conversations with John Wilesmith, and it just never came across. ¹⁵

4.134 Mr Lawrence stated that cross-contamination was not perceived as being a problem at the time. The consensus was that BSE was 'dose-related' and therefore required a significant quantity of infective material in order to transmit the disease. ¹⁶

4.135 Witnesses from the feed industry told us that they did not understand that cross-contamination was a matter for concern in relation to BSE. Thus Dr Brian Cooke of Dalgety Agriculture Ltd said:

Both the industry in the form of UKASTA and ourselves in Dalgety Agriculture were reassured in the period up to 1994 by indications from MAFF that a significant level of cross-contamination of meat and bone meal from pig and poultry feeds into ruminant feeds would be necessary to infect a cow with BSE. ¹⁷

4.136 Mr Reed, the Director-General of UKASTA, told us that statements by MAFF officials over a very long period suggested to UKASTA that, for infection to occur, it was necessary to have a very high quantity of contaminated feed or feeding over a substantial length of time, or both. ¹⁸

4.137 Some support for this statement can be derived from a UKASTA file note of 2 October 1992 which recorded:

It was noted that MAFF were not concerned about cross-contamination of feeding stuffs in mills because the dose rate of meat and bone meal would be too low. ¹⁹

4.138 There is, however, no documentary evidence of any discussion with UKASTA about the size of the infective dose or the risk of cross-contamination at the time that the ruminant feed ban was introduced.

4.139 In this respect the position in Great Britain differs from that in Northern Ireland. On 2 June 1988 a meeting took place between officials in DANI and representatives of the animal feed and rendering industries at which a proposal to follow MAFF's lead in introducing a ruminant feed ban was discussed. The minutes record that the question was raised:

. . . of how much cross-contamination would be acceptable in compounding mills and how mills could control this. Because of the possible risk of cross-contamination it was suggested that some firms may refuse to handle meat and bone meal. ²⁰

4.140 Nevertheless, Mr Edward Sullivan, the Chief Veterinary Officer for Northern Ireland at the time, told us that their thinking then was that a considerable dose of the infective material would need to be incorporated into the MBM to cause the disease and that the various processes involved would produce tremendous dilution:

So in the back of our minds we thought that contamination trouble from that source was practically near to negligible. ²¹

Should the possibility of cross-contamination have been foreseen?

4.141 Mr Kevin Taylor expressed the view that to suggest that steps should have been taken to address the risk of cross-contamination in 1988 was to use hindsight. The problem was viewed as an animal health problem, not a public health problem:

. . . in animal control terms it is perfectly normal to put in place what seems to be a logical and simple control. There is no virtue in going for complexity, it is better to go for simplicity, and to observe what happens. When you get evidence that what you have done is either unsuccessful or could be improved, if you find evidence that it could be improved or that it is failing in some way, then you look again at the situation and you make the changes which are necessary to block the leakage, if you like. It is like blocking a dam with stones. It is easy to block a big part of the transmission and it is only later that you become aware that it has been leaking a bit round the edge, so then you block the bits round the edge.

That is exactly what was done with BSE and it is exactly what historically has been done with other animal diseases. ²²

4.142 We accept that the approach described by Mr Taylor may be appropriate when dealing with animal diseases with a short incubation period. It was not, however, appropriate for BSE. The lengthy incubation period of the disease meant that several years would elapse before it would become apparent whether the dam had leaks, or was even holding. That factor alone meant that a rigorous appraisal of any steps necessary to ensure the practical efficacy of the ban was desirable - at least once it was appreciated that MAFF was dealing with a sizeable epidemic.

4.143 In addition, it was not correct to proceed on the footing that BSE posed only an animal health problem. The fact that BSE posed a potential risk to human health, even if the risk was seen as remote, was appreciated at the time that the ruminant feed ban was introduced.

4.144 In 1995, in a paper on risk communication strategy, Mr Danny Matthews cited, as an example of crisis management which had not occurred, anticipation, as early as 1988 or 1989, of the partial failure or delayed implementation of the ruminant feed ban:

By asking ourselves 'What do we do if cases occur after the ban?'we might have anticipated the means by which we could ensure early compliance . . . ²³

4.145 We believe that this comment is particularly apt in relation to the risk of cross-contamination.

4.146 At the heart of the complacency that existed in relation to this risk was a failure on the part of those most directly concerned with the introduction of the ruminant feed ban to appreciate the possibility that a very small quantity of infective material might suffice to infect an animal orally.

4.147 Evidence from many quarters has demonstrated that there was a widespread belief that a substantial quantity of infective material had to be eaten before a cow would become infected with BSE. That belief persisted until early 1995, when results from the attack rate experiment were released (see vol. 2: Science). These demonstrated that 1 gram of infective brain tissue had sufficed for oral infection of a cow. The volume of 1 gram is equivalent to two peppercorns. ²⁴

4.148 The many to whom this result came as a surprise included Dr David Shannon and Dr Kenneth MacOwan of the Chief Scientists Group of MAFF ²⁵ and Professor Eric Lamming, who said of his Committee's understanding in 1991:

We were under the impression from the evidence available that it would require much larger quantities than that. ²⁶

4.149 Mr Bradley told us that he 'would not have had a clue, in a nutshell' about the size of an infective dose, though it would 'not necessarily be expected to be extraordinarily large'. He continued:

I mean I am talking about small quantities, less than let us say 100 grammes of brain material. But what that amount would be to cause disease in the same species in regard to BSE was speculation, which I could not really comment upon. ²⁷

4.150 Dr Watson endorsed Mr Bradley's evidence, observing that it was 'purely conjecture' and that both he and Mr Bradley were surprised at the small amount of material from the central nervous system that did transmit orally to cattle. ²⁸ Mr Wells also expressed surprise that a single gram of brain homogenate proved to be an effective dose. ²⁹

4.151 The widespread misconceptions about the amount of infective material necessary to transmit BSE to a cow reflect the fact that this question does not appear to have been thought to be of sufficient significance to merit detailed consideration. The attack rate experiment was designed not to ascertain the minimum amount of material necessary to infect a calf, but as a precursor to the pathogenesis study. It was not begun until 1992. The Tyrrell Committee had not recommended an attack rate study - which, with hindsight, Mr Meldrum found surprising. ³⁰

4.152 Attack rate experiments take time, but discussions with Mr Wilesmith, Dr Kimberlin and the scientists at the NPU would have made it possible, at the time of the ruminant feed ban, to reach an informed view that it was at least possible that a very small quantity of infective material would suffice to infect a cow. We shall summarise the views on dose that these scientists held.

4.153 We have already referred to Mr Wilesmith's contemporary statement that 'the effective dose of the disease was very small' (see paragraph 4.127). In oral evidence, he told us that he considered that a very small amount of the tissues with high infective doses would provide an effective exposure. He considered that this view was, or should have been, widely shared by administrators at MAFF, though he declined to name individuals:

. . . they were perfectly aware of what the likely amount, that being a small amount, was necessary to infect animals, because everybody was aware of what those inclusion rates were. They were somewhere between 2 and 4 per cent of calf rations if any meat and bone meal was included. So there was a knowledge of it. But I think my concern is that it seems to have been ignored. ³¹

4.154 In 1991 Dr Kimberlin was to express the view to the Lamming Committee that the dosage level required to infect an animal was likely to be very low. ³² He advised on the design of the attack rate experiment. His selection of 1 gram for the lowest dose in the experiment indicates that, as he confirmed to us, he considered it possible that this dose would achieve oral transmission. ³³

4.155 Between June and August 1988 the NPU set up an experiment titled 'Transmission of BSE and natural scrapie into sheep and goats by intracerebral and oral routes'. It used, for this experiment, brain material from BSE-affected cattle supplied by MAFF. The quantities used were determined simply by the amount of this material left over after other experiments. An oral dose of 0.5 grams was given to six 'positive-line' sheep, six 'negative-line' sheep and three goats. The doses used indicate that those designing the experiment at least envisaged the possibility that quantities as small as 0.5 grams would prove orally infective.

4.156 There was no valid basis upon which Mr Meldrum could be confident in 1988 that cross-contamination in the feedmill would not involve sufficient quantities of infective material to give rise to a risk of transmission. It seems to us that he must have simply assumed this to be the case. It is right to note that Mr Wilesmith was present at the meeting where Mr Meldrum stated that feed compounding mills would not be investigated because they presented, at worst, only a low contamination risk. Mr Wilesmith did not comment on that statement.

4.157 We have already observed that, at the time the ruminant feed ban was introduced, the true scale of the BSE epidemic was latent. The apparent rate of infection was only about 60 cases a month. In those circumstances we can understand a failure to attach significance to the possibility of infection as a result of cross-contamination of feed. After the introduction of compulsory notification, the picture changed. In the course of the month of September 1988, 435 cases were reported in Great Britain. Once these figures were apparent, if not before, Mr Meldrum should have addressed the problem of cross-contamination.

4.158 We have considered whether there were other officials at MAFF who should have given consideration to the potential risk of cross-contamination in feedmills. While we have indicated our dissent from Mr Kevin Taylor's approach to controlling BSE (see paragraphs 4.141-4.143), we accept that he lacked the knowledge of the operation of feedmills that was needed if the risk of cross-contamination there was to be appreciated.

4.159 Dr Peter Dawson was the ACVO who headed the VI Service. He has made the point that he had no involvement in the introduction of the ruminant feed ban. That is correct. What is more to the point is that he had no personal experience of what happened in feedmills and, thus, could not have been expected to foresee potential problems of cross-contamination.

4.160 Mr Cruickshank has told us that he had been advised by his veterinary colleagues that cross-contamination was not a problem. ³⁴ His understanding about this would have been confirmed as a result of being present when Mr Meldrum stated to representatives of the cattle industry that feed compounding mills presented, at worst, only a low contamination risk. ³⁵ Furthermore, he had been informed by his colleagues that a very large dose was required to infect.

4.161 We are not aware who provided Mr Cruickshank with the information to which he has referred, but his understanding about the need for a large dose, by which we understood him to mean a large amount of infective material, was one that was shared by a number of witnesses, including Mr Lawrence. ³⁶ We do not consider that Mr Cruickshank should have appreciated the risk of cross-contamination.

What should have been done?

4.162 We think that the first thing Mr Meldrum should have done was to seek guidance from those best placed to form a view on the quantity of infective material that might be capable of transmitting BSE. This was no easy question, involving as it did the extent to which the processes of rendering and feed compounding would dilute infective material. He would not have received definite advice, but whether he turned to the CVL, to Dr Kimberlin or to the NPU, he would have been likely to be told that there was a real possibility that a small quantity of brain or spinal cord might suffice to infect.

4.163 He should then have ensured that a warning of the dangers of cross-contamination was given, not merely to feedmills, but also to farmers, many of whom mixed their own feed. An example of the guidance that would have been appropriate and helpful is the 'Notice to Manufacturers and Mixers of Animal Feedingstuffs' sent out by Mr Kevin Taylor on 10 August 1995. ³⁷

4.164 Mr Meldrum told us that he was well aware that those who manufactured feed knew of the risk of cross-contamination with medicinal products and the means by which this could be avoided with the aid of the UKASTA Code of Practice. He assumed that they would employ the same precautions to guard against the risk of cross-contamination of ruminant feed by feed prepared for non-ruminants. He added:

I am as certain as I can be that when I would have discussed this with them in 1988 I would have said : 'Yes, I am of course aware of what routine precautions you take; and on that basis I am not unduly worried about the risks from BSE.' ³⁸

4.165 We do not believe that such a conversation took place. Had Mr Meldrum considered at the time that it was necessary for feedmills to apply the UKASTA Code in order to prevent cross-contamination between ruminant and non-ruminant feed, we believe that he would have ensured that specific advice was given to feedmills of the need for this. It would not have been safe to rely upon them to do this on their own initiative.

4.166 Mr Reed of UKASTA provided us with some information as to whether the UKASTA Code of Practice would always have been applied by mills in order to segregate ruminant feed from non-ruminant feed. He informed us:

The measures included in the 1984 UKASTA Code would have been adopted as routine in feedmills whenever it was appropriate. If a non-ruminant feed did not contain materials known adversely to affect a ruminant, then the degree of clean-out between products could be less rigorous. Some non-ruminant feeds contained neither meat and bone meal nor medicinal additives and hence would not adversely affect the following feed even if the conveying equipment carried over a small quantity.

On the other hand if any such non-ruminant feed contained a known potentially deleterious material or an ingredient which would adversely affect the next product, the plant would be purged with 'clean' material such as wheatfeed or ground wheat, or production runs re-scheduled to ensure that no sensitive feed followed. ³⁹

4.167 In the course of 1994 MAFF received information that made it plain that, while some feedmills were taking steps to prevent cross-contamination, others were taking none: see vol. 5: Animal Health, 1989-96.

4.168 This information should have occasioned no surprise - nor did it. It was not reasonable to expect that all feedmills would take steps to prevent cross-contamination of ruminant feed unless given express guidance that this was necessary.

4.169 As to this, Mr Reed stated:

At the time when the ruminant feed ban was introduced in July 1988, we were given no reason to believe that the risk to ruminants arising from cross-contamination with MBM existed at the minimal levels likely to occur where the normal practices of scheduling, cleaning out, etc laid down in industry and company codes of practice were applied in multi-species mills.

The issue of the size of infective dose for BSE was given no special significance in discussions between the Government and industry until much later - after a number of BAB cases had occurred. In the past, the removal from a formulation of a specific ingredient known to cause a problem, combined with the procedures already in place to minimise cross-contamination, had proved successful in preventing the problem continuing even when the same ingredient was still included in other products for other species. In the absence therefore of any knowledge of what might constitute a dangerous level of carry-over or guidance from MAFF as to what if any additional measures to take, the existing methods were deemed adequate. ⁴⁰

4.170 On 14 October 1994 Prosper De Mulder met with UKASTA to discuss cross-contamination. UKASTA's minute of the meeting records:

Reference was made to the discussions, for which no official records were known to be available, in the late 1980's when trace contamination was not considered to be a problem. ⁴¹

We had no other evidence about any discussions involving UKASTA in the 1980s when cross-contamination was discussed. UKASTA did, however, discuss with MAFF concerns about cross-contamination in late 1990. ⁴² It seems to us possible that they may have received some reassurance on that occasion (see vol. 5: Animal Health, 1989-96, Chapter 2).

4.171 There was also a need to inform farmers of the dangers of cross-contamination. We had little evidence as to the extent to which cross-contamination may have taken place on the farm. We have concluded, however, that there must have been some scope for this.

4.172 Evidence given to the Lamming Committee suggests that as much as 40 per cent of feed is mixed on the farm. ⁴³ Methods of mixing varied, widely from a use of specialist machinery to mixing feed in a bucket. It was thought in 1995 that home mixers were unlikely to mix rations for more than one species. ⁴⁴ Whether or not they were home mixers, however, farmers should have been warned of the need to avoid any contamination of their ruminant feed by any trace of MBM or non-ruminant feed that might contain it.

The need for a test

4.173 In vol. 5: Animal Health, 1989-96 we examine the leisurely course of attempts made by MAFF 'in-house' to develop an ELISA test for the presence of ruminant protein in animal feed. This became a matter of critical importance only when it was realised that cross-contamination was probably resulting in transmission of BSE via cattle feed. Had this possibility been identified at the outset, we believe that this task would have been given higher priority and might have been treated as an important R&D project.

The lesson

4.174 Failure to address the risk of cross-contamination of ruminant feed with non-ruminant rations resulted from a failure to give rigorous thought to the question of dose. The best time for rigorous consideration of the practical implications of a set of Regulations is at the time of their introduction. Once Regulations are in place, there is a tendency not to consider potential problems, but to wait to see whether any arise in practice. The question whether cross-contamination would be a problem was an obvious one to anyone with knowledge of how feedmills operated. The question was raised, but dismissed by Mr Meldrum without the rigorous consideration that he should have seen it received. No one could know the minimum quantity of infective matter necessary for oral transmission. In the absence of knowledge, the ruminant feed ban should have been implemented on a 'worst case' assumption.