Volume 5: Animal Health, 1989-96
2. The ruminant feed ban, 1989-96
Discussion
Introduction

Introduction

The scheme of the ruminant feed ban

2.558 The terms of the Bovine Spongiform Encephalopathy Order 1988 (the 1988 Order) which imposed the ruminant feed ban, are set out at paragraph 2.3. They prohibited a person from selling or supply for feeding to ruminants, any feedstuff 'in which he knows or has reason to suspect that any [ruminant] protein has been incorporated.' This wording was appropriate to prevent the deliberate use of ruminant protein as an ingredient of ruminant feed. This was, we believe, what the 1988 Order was intended to achieve.

2.559 The scheme of the 1988 Order contrasts with that adopted by regulations under the Agriculture Act 1970, designed to ensure that feedstuffs do not contain deleterious ingredients. Under that Act a feed compounder or merchant commits an offence if it is proved by sample that feedstuffs on their premises contain the proscribed ingredient - see section 73 of the Agriculture Act 1970. In contrast to the 1988 Order, knowledge is not required to be proved - the offence is absolute. There is, however, under section 82 of the 1970 Act a 'due diligence' defence if a person charged can prove:

That he took all reasonable precautions and exercised all due diligence to avoid the commission of such an offence by himself or any person under his control. ¹

2.560 This may be compared with the RFB. Both the 1988 Order and its successor, the Bovine Spongiform Encephalopathy (No. 2) Order 1990, adopted the wording we have quoted in paragraph 2.558. It was suggested by UKASTA in November 1994 that commission of an offence if there were 'reason to suspect' meant that the industry would be liable should ruminant protein be found in ruminant feed, subject to a 'due diligence defence'. It is possible that UKASTA proceeded upon this basis more generally. The concept that their members could be liable if they were not able to show due diligence may have been useful when negotiating with MAFF, but we do not think it was an accurate description of the RFB. This ban did not involve any need to invoke a defence of 'due diligence' as defined by the 1970 Act.

2.561 The scheme of the 1970 Act makes appropriate provision for regulations designed to prevent accidental cross-contamination of feed with deleterious matter. The wording of the 1988 Order was not appropriate to achieve that object. It was not the intention of those responsible for its drafting that it should. As we have explained in vol. 3: The Early Years, 1986-88 cross-contamination in the feedmills was not considered to be a problem when the 1988 Order was drawn up. This was because of a failure to focus on the possibility that a small amount of infectious material might suffice to transmit BSE. This failure was to influence almost everything that followed.

2.562 MAFF officials expected those who manufactured animal feed, and those who fed it to their animals, to recognise the importance of the RFB and to comply with it of their own volition. Thus, in a written statement, Mr Kevin Taylor explained to us:

Meetings were held to publicise the ban and explain the reasons for it. There was, however, no practical way in which the ban could be enforced since there was no test which could identify rendered ruminant protein in animal feed . . . Effectiveness therefore depended on the industry recognising the importance of the ban and complying with it. ²

2.563 Once existing stocks of ruminant feed were disposed of, it was reasonable to expect that feed compounders would comply with the 1988 Order and cease to incorporate ruminant protein in ruminant feed. Alternative sources of protein were available and there was very little financial motive for flouting the law. We are not aware of any evidence of deliberate breach by feed compounders of the ruminant feed ban after existing stocks had been disposed of.

2.564 Equally it was reasonable to expect that farmers, once warned of the danger of incorporating ruminant protein in ruminant feed, would not deliberately continue to run that risk.

2.565 It was not reasonable to expect all feedmills and farmers to anticipate that cross-contamination of ruminant feed with traces of feed for monogastric animals would render the ruminant feed potentially infective unless specifically warned of this danger. No such warnings were given. It seems likely that UKASTA initially assumed that cross-contamination was not a problem. As we view the evidence, UKASTA's concerns in November 1990 about cross-contamination were not that this would render feed unfit for cattle, but that, if detected, it would render their members liable to successful prosecution unless they could establish a due diligence defence.

2.566 For nearly three years MAFF officials proceeded on the assumption that, unless there was maternal or horizontal transmission of BSE, the ruminant feed ban would result in the eradication of the disease. The BABs that were identified from March 1992 onwards did not, initially, disturb this illusion. They were attributed to carry-over of existing stocks after the 1988 Order took effect. As animals born later and later after the ban were diagnosed with BSE, the perceived extent of this carry-over stretched elastically. In June 1992 a three-month time lag in the effect of the ban was considered likely (see paragraph 2.139). By November this period had stretched to six months (see paragraph 2.190). If any thought was given to the possibility of cross-contamination, it was assumed that this would not have been a problem once the animal SBO ban had come into force in September 1990.

2.567 Mr John Wilesmith began to become concerned about the possibility of cross-contamination towards the end of 1993 (see paragraph 2.224). It was not, however, until early 1994, when a handful of BABs were confirmed that had been born after the animal SBO ban came into force, that it dawned on MAFF officials that there was a dual problem which was resulting in SBOs first contaminating MBM produced for incorporation in animal feed with the secondary consequence that this feed cross-contaminated feed being prepared for cattle.

2.568 This situation was of concern both to MAFF and to UKASTA. As we have seen, they worked in close consultation in considering how to address the problem. The following options received consideration:

1. A prohibition, voluntary or statutory, on the use of MBM in any animal feed;
2. A requirement that feedmills use separate facilities for the production of ruminant and non-ruminant feed;
3. Sampling the output of feedmills in order to detect any contamination of ruminant feed with ruminant protein.

2.569 We propose to consider each in turn.

Prohibition on the use of MBM in any animal feed

MAFF's attitude to this solution

2.570 We have seen that, from the moment that the Southwood Working Party appeared to be condemning the practice of feeding animal protein to animals, MAFF had been concerned to ensure that this practice should continue. This attitude changed not one whit in the face of evidence that the continuance of this practice was resulting in the contamination of cattle feed with BSE infective material. Mr Richard Packer was asked:

Would it be right to say that the Ministry encouraged the use of ruminant protein other than that derived from SBOs in non-ruminant feed?

He replied:

No I don't think it would. ³

2.571 Despite this answer, we are in no doubt that MAFF officials were not prepared to contemplate the cessation of the use of animal protein in animal feed as the solution to the problem of cross-contamination that was occurring.

2.572 When, on 7 October 1994, Mr James Reed put forward as one option that there should be a stop to using ruminant meal in all feed, he understood that Mr Packer's inclination was to give the feed industry the reassurances it wanted in order to prevent an MBM ban if this was 'remotely possible' (see paragraph 2.353). Mr Packer's report of the meeting to Mr Waldegrave suggests that Mr Reed correctly interpreted Mr Packer's attitude. He commented:

All of this is very unwelcome. Feed manufacturers could switch relatively easily to other raw materials but a decision to stop using ruminant meal would have serious implications for the livestock industry on top of all the BSE and ferry-related blows. Moreover, such a move would undoubtedly generate further public concern about the safety of beef.

It seems to me, at first sight, that there are two possible courses of action. First, we could look to see whether we can legitimately reconstruct our investigations so as to give less worry to the feed industry. Secondly, we could see what other assurances we can give, perhaps about our new rendering rules. ⁴

2.573 Over a year later, in a paper prepared for Mr Hogg, Mr Keith Meldrum set out the options for dealing with the problem of cross-contamination of cattle feed. He commented:

A second alternative would be to prohibit the use of mammalian MBM in animal food. This would prevent it being fed to pigs or poultry and remove any risk of cross contamination. The economic consequences would be devastating: the rendering industry would become a waste disposal operation, producing only tallow as a useable product. ⁵

2.574 If in the autumn Mr Meldrum had privately thought this course might be desirable (see paragraphs 2.508-2.509), he had changed his mind by December.

2.575 The recognition in March 1996 that BSE was transmissible to humans led to an immediate ban on feeding animal protein to animals. We do not, however, consider that this was a course that MAFF should have adopted as a reaction to the discovery that ruminant protein was contaminating ruminant feed. The economic consequences of a total ban on the use of animal protein in animal feed were indeed severe. To have adopted this measure simply to prevent further cross-contamination of cattle feed would have been an admission of defeat. Other, less drastic, viable options were open and MAFF officials and their Ministers acted reasonably in looking to these rather than to a total ban on MBM in feed.

UKASTA's attitude

2.576 UKASTA were repeatedly threatening that their members would have to cease using MBM in animal feed. UKASTA have disclosed to us a confidential paper that had been circulated to their National Executive Council on 24 August 1995. This set out UKASTA's policy aims as follows:

These have been consistent, although unstated except in FEC discussions, since at least 1989:

• To minimise the risk of farmers' claims for compensation from feed compounders.
• To minimise the potential damage to compound feed markets through adverse publicity.
• To maximise freedom of action for feed compounders, notably by maintaining the availability of meat and bone meal as a raw material in animal feeds, and ensuring time is available to make any changes which may be required. ⁶

2.577 The paper went on to cite as one of the successful examples of the strategy:

UKASTA pressure dissuaded MAFF from publicly linking voluntary ELISA tests on feed on farms with BABs to (possibly compulsory) tests at compounders' premises in June/July 1994; ⁷

2.578 Looking to the future, the paper concluded:

Tests may show that ruminant feeds have been sold which contain illegal traces of ruminant protein. More likely, a few positive test results will turn up but proof that a particular feed mill knowingly supplied it to a particular farm will be difficult if not impossible.

The threat remains real and it will be some years before feed compounders are free of it. The longer we can avoid any direct linkage between feed milling practices and actual BSE cases, the more likely it is that serious damage can be avoided. In issue management terms, the aims and the strategy remain valid, but must be kept under review in the light of further events. ⁸

2.579 It is apparent that UKASTA were as anxious as MAFF that MBM should continue to be used as an ingredient of animal feed. They were naturally anxious, however, to do their best to protect their members from the risk of prosecution or litigation.

2.580 MAFF officials were not prepared to countenance giving

UKASTA any assurance that the law would not be enforced, nor to provide by way of comfort any statement suggesting that cross-contamination with ruminant feed would carry no risk of infection. They were, however, able to provide UKASTA with the statement referred to at paragraph 2.383 which led to UKASTA dropping the threat to introduce a voluntary ban on the use of MBM.

Dedicated facilities

2.581 The suggestion that there might be a prohibition on mills using the same facilities to manufacture ruminant and monogastric feeds did not commend itself to Mr Packer when he considered it in October 1994 (see paragraph 2.352). The following month, however, Mrs Browning expressed the view that the correct course would be to move to a situation in which separate production lines would have to be used for ruminant and non-ruminant feed (see paragraph 2.381). She accepted however, Mr Packer's advice that this was not necessary. Furthermore, this was not, it seems, a practical possibility. Some mills were voluntarily introducing separate facilities, but the smaller feedmills were not able to do this.

2.582 The position had not changed a year later when Mr Meldrum explained the practical difficulties to Mr Hogg. The minute describing the meeting recorded:

The Minister noted that it would not be feasible to require all establishments to have in each factory different lines for different feed, although some large companies had already moved in that direction. ⁹

2.583 In his paper for Mr Hogg of 8 December 1995 Mr Meldrum advised:

A third, almost equally unpalatable, approach would be action in feedmills: at its most extreme insisting that no ruminant feed is produced in premises which use MBM for any other purpose. Again there would be immense financial and practical difficulties, and the action would be difficult to defend in the light of the results of ELISA testing which, in the past year, have failed to find any evidence of ruminant protein in animal feed . . .

At the present time I would not recommend introducing legislation to require that cattle feed is produced in a dedicated mixing line to reduce the risk of cross-contamination because of the various steps that are now in place to prevent infectivity entering meat and bone meal and the measures that we have in hand to further reduce the risk of cross contamination in feedmills. ¹⁰

2.584 We have considered whether, on the facts known to MAFF officials in 1994 and 1995, they should have sought to impose, either by voluntary agreement or by Order, a ban on using common facilities to produce both ruminant and non-ruminant feed. The primary measures that MAFF took in order to prevent contamination of cattle feed consisted of tightening up the implementation and enforcement of the animal SBO ban (see Chapter 4).

2.585 On 5 April 1995, in addition to approving recommendations on tightening the animal SBO ban, Mr Waldegrave asked whether the results of the attack rate study meant that more needed to be done to ensure the cleaning of feed lines which once carried contaminated feed. Mr Meldrum's reply, quoted at paragraph 2.424, ended:

There is no action I can recommend that we should take at this time except to continue to intensify the controls on the disposal of the SBOs. ¹¹

2.586 In fact, at this time, MAFF officials were taking action, which we shall shortly consider, aimed at reducing cross-contamination in feedmills which used common facilities to process ruminant and non-ruminant feed.

2.587 If defects in the implementation of the SBO ban could be remedied, infectivity of MBM entering the feedmills for incorporation in monogastric feed should have been minimal. Effective steps were being taken to tighten up the implementation of the SBO ban, including a requirement that SBOs should be processed in dedicated facilities - see Chapter 4. We consider that it was reasonable for MAFF officials and Ministers to conclude that it was not necessary to require feedmills to undertake, in parallel with the renderers, the expense of installing duplicate lines.

2.588 In reaching this conclusion we have borne in mind that MAFF officials could reasonably assume that the steps already taken would have reduced the scale of infection to a fraction of that at the height of the epidemic. Figure 2.1 gives an indication of the extent to which they had been successful in controlling the disease before the additional measures taken in 1994 and 1995.

Figure 2.1: Confirmed cases of BSE with known dates of birth, plotted by month of birth

2.589 We are now in a position to judge the efficacy of those steps. As at 1 June 2000, 1,010 BABs have been confirmed which were born in 1994. The number for 1995 is only 179 and for 1996 only 1 BAB has been confirmed, born in January of that year. While further cases can be expected to be identified in the years to come, it seems clear that these are likely to be no more than a very small trickle.

Sampling

2.590 An effective ruminant feed ban would have coupled a legal power to enter feedmills and to take samples of feedstuff with an effective test of whether the samples taken contained ruminant protein.

Right of entry

2.591 In his witness statement to us, Mr Meldrum commented:

On 18th July 1995 Commission Decision 95/287/EC came into effect requiring the monitoring of feedmills on a routine basis to identify any incorporation of ruminant protein into ruminant feed. This overcame the problem of right of access to feedmills which had not been overcome in the meantime. ¹²

2.592 This chapter includes some of the confused and conflicting advice provided to officials by MAFF lawyers in relation to right of access to feedmills, the right to take samples of feedstuffs, the right to inspect records, and the manner in which samples had to be analysed. This is not a matter for criticism: the lawyers were being asked to solve a problem which the legislation had not been designed to meet. We suspect that there was probably more than one pathway through the regulatory maze that would have enabled compulsory sampling of animal feed, if only a reliable sampling technique had been established. One approach would have been an appropriate addition to the Feedingstuffs Regulations under the Agriculture Act 1970. The problem was the unreliability of the ELISA test.

The ELISA test

2.593 A separate report could be devoted to the history of the attempts to produce a successful ELISA test. We shall simply pick out a few incidents from the account given in this chapter.

1. Mr Ansfield's project proposal of April 1989 anticipated, if we understand it correctly, that the major work in preparing the test would be completed by the end of the year (paragraph 2.54).
2. On 24 May 1990 Mr Ansfield predicted that validation would take at least to the end of the year (paragraph 2.71).
3. On 1 November 1990 Mr Ansfield said that he was confident that validation would be carried out by the end of the year and the test would then be available for use (paragraph 2.88).
4. Up to this point the test had only been applied to processed MBM. When it was applied to compound feed it was found that in the majority of cases it produced positive results even without the inclusion of MBM (paragraph 2.105).
5. By September 1991 progress had been made towards resolving this problem (paragraph 2.118).
6. In August 1992 Mr Ansfield reported that detection of bovine and ovine protein in MBM was now possible down to one part in 6,400 and in compound feed to one part in 200. High standards of laboratory expertise and a composite range of equipment were, however, required to carry out the test, which would take three days (paragraphs 2.155-2.156).
7. In March 1993 it was reported that development of the test was halted by relocation of Worcester VIC to Luddington (paragraph 2.208).
8. By March 1994, when consideration was being given to a voluntary sampling scheme, Luddington had capacity to test only 20 samples a week (paragraph 2.263).
9. In July 1994 Luddington were refusing to release details of their sampling techniques as the test was to be patented (paragraph 2.297).
10. By October 1994 the sensitivity of the test was such that a 5 per cent carry-over of a feed with a 5 per cent ruminant MBM content would be necessary for the threshold of detection to be reached (paragraph 2.348).
11. On 30 November 1994 UKASTA were told that the test had been validated (paragraph 2.385).
12. By February 1995 the validity of the test had been put in doubt because of experience of both false positives and false negatives (paragraphs 2.413-2.414).
13. On 5 June 1995 it was reported that a number of vegetable products gave positive reactions (paragraph 2.453).
14. In December 1995 Mr Ansfield advised that further work on the test was necessary because appropriate controls had not been used for compound feeds (paragraph 2.536).
15. In February 1996 the first round of mandatory sampling of feedstuffs at mills took place.

2.594 These incidents take the story up to the end of the period that falls within our terms of reference, and up to the point at which, after events leading up to 20 March 1996, all animal protein was banned from animal feed. We were told in a statement from Dr Matthews, however, that difficulties in validating the test had persisted. He summarised the position as follows:

While expectations of the test have been exaggerated, it is true to say that it provides an opportunity to seriously question quality control measures in feedmills. There are still difficulties to resolve, not least because of the complexity of the legislation. The test detects protein - not meat and bone meal. Some protein can be legitimately included in ruminant and non-ruminant rations. Consequently a positive test result is not in itself sufficient evidence of a contravention. Further investigations are essential in every case, and it is these rather than prosecutions that produce serious review of working practices in affected plants. ¹³

2.595 We have no criticism to make of Mr Ansfield. He worked diligently at attempting to perfect the ELISA test and it is clear that he has met with considerable technical difficulties. Development of the test was not, however, treated as a matter of urgent priority. Mr Meldrum decided that it should be developed 'in house' rather than in collaboration with the Agriculture and Food Research Council (AFRC), explaining that this was due to 'redundancies at AFRC and other difficulties' (see paragraph 2.53). In his statement to us, Mr Ansfield spoke of eagerness of the AFRC Institute of Food Research, Bristol, to become involved in test production, and commented, 'For reasons of which I am unaware this particular direction of test development was not followed up'. ¹⁴

2.596 Mr Ansfield also commented in his statement:

With hindsight there are always ways to consider how a test could have been developed more quickly. On the evidence available at that time it is my opinion that the considerable resources available to the State Veterinary Service (which included a wealth of expertise at the Central Veterinary Laboratory at Weybridge and numerous contacts with commercial and research establishments) was the appropriate option to take at that time.

In respect of the speed of developments, it is possible that involving a team of experts working full-time rather than having one person based at one laboratory liaising with other experts, (usually those based at the Central Veterinary Laboratory at Weybridge and with commercial companies involved with protein separation) may have speeded up the development of the test. Having said this, development can only be achieved if there is some form of breakthrough and this occurred early on in the test development with the removal of gelatine and the concentration of remaining proteins for testing. ¹⁵

2.597 Mr Ansfield was not told that development of the test was a matter of urgency (see paragraph 2.59). We do not believe that in 1989 it was seen as a matter of urgency. This was not unreasonable, if the only purpose of the ban was to prohibit the deliberate inclusion of ruminant protein in ruminant feed. A reliable test only became a matter of urgency when it was appreciated that there was a need to ensure that ruminant feed was not subject to accidental cross-contamination.

2.598 On 20 July 1993 Mr Kevin Taylor, in a minute to Dr Peter Dawson, commented:

As I understand it the development of the test at Worcester and now Luddington VICs has largely been a consequence of their existing expertise in species testing of meats, and has not been identified as an area requiring major R&D funding. ¹⁶

2.599 He added:

I fear that we will be vulnerable to criticism if we are not able to test feed intended for ruminant animals fairly soon. ¹⁷ (See paragraph 2.214.)

2.600 In May 1994 Mr Bradley commented that there seemed to have been inordinate delay in getting testing established (see paragraph 2.280).

2.601 We consider that the delay was primarily due to the failure, when the ruminant feed ban was introduced, to appreciate the possibility that the amount of infective material necessary to transmit the disease might be very small, so that the problem of cross-contamination needed to be addressed (see vol. 3: The Early Years, 1986-88, Chapter 4).

2.602 If the importance of a reliable test had been foreseen in 1988, it might well be that an alternative approach to achieving this would have been adopted, which would have been successful by 1994. As it is, MAFF officials found themselves in a position where the test was not sufficiently reliable to demonstrate the presence of ruminant protein to the standard required for criminal proceedings.

2.603 The latter part of the period with which this Inquiry is concerned saw the ELISA test used on farms, with the consent of farmers, in order to check on recent supplies from feedmills. On at least two occasions, follow-up to mills had beneficial results (see paragraph 2.409). At the same time many feedmills submitted samples for testing as part of their own due diligence to check that their products were free of contamination.

2.604 It does not seem to us that UKASTA showed any enthusiasm for the programme of unannounced mandatory checks on mills that MAFF proposed to initiate in order to comply with the Commission decision - and this should occasion no surprise - their primary duty was to their members who had nothing to gain from the proposal. However, one round of tests was carried out in February 1996. This identified four mills out of the 25 sampled where feed was contaminated with animal protein.

2.605 An interesting feature of the sampling operation was the fact that MAFF staff carried this out themselves without involving the Trading Standards Officers of the County Councils who had statutory authority for enforcing the regulations. In December 1993 Mr Taylor had received a somewhat bizarre suggestion from Mr Howard that it would be 'an extremely risky business' to allow Local Authorities to carry out sampling because the results would not remain confidential. Mr Taylor had understandably responded that test results could hardly be suppressed and that he did not see that it mattered much whether MAFF or the Local Authorities collected samples (see paragraphs 2.237-2.238).

2.606 In May 1994 Mr Eddy, in a submission which had been approved by Mr Meldrum, Mr Haddon, Mr Taylor and Dr Cawthorne, sought and obtained Mr Soames's approval to MAFF introducing the new test under MAFF's control. The submission suggested that some Local Authorities could not be trusted to 'treat the existence of this test in a prudent way', and that it was better for MAFF to keep the test capacity in-house so as to be able to monitor the situation.

2.607 It seems to us that, whatever may have motivated this course, the limitations of the ELISA test made it desirable for MAFF to retain overall control of the sampling exercise, rather than to hand this over to individual Local Authorities.

Guidance to the feed industry and farmers

2.608 Sampling was not the only means open to MAFF officials by which to seek to reduce cross-contamination of ruminant feed. When MAFF officials realised that cross-contamination was probably responsible for BABs, and when they later learned that one gram of infective material would suffice to infect, the question arose of the steps that should be taken to bring these matters to the attention of all feed compounders. The need to do so was underlined by the report from a VO of the sources of cross-contamination that he had identified in a mill that he had visited (see paragraphs 2.307-2.308).

2.609 That report led Dr Matthews to suggest to Mr Taylor that the SVS should carry out an inspection of feed plants (see paragraph 2.305). His manuscript note of 14 July warned that while some companies were aware of the risks and were taking steps to avoid cross-contamination, others were not. Mr Bradley gave a similar message a week later, when he wrote to Mr Taylor stating that several mills in East Anglia were wise to the problem and taking steps to deal with it, and emphasising the need to see that all the loopholes were plugged (see paragraph 2.306). This led Mr Eddy to draw Mr Taylor's attention to the desirability of sending out guidance to farmers and to the feed industry (see paragraph 2.311).

2.610 These suggestions do not appear to have resulted in immediate action. On 15 May Mr Fleetwood put up to Mr Taylor and Dr Cawthorne a suggestion, agreed by him with colleagues, that the SVS should carry out advisory visits to feedmills to help them to avoid cross-contamination.

2.611 Dr Cawthorne's response, on the assumption that all UKASTA members would have carried out a self-audit, suggested that attention should be focused on the small mills, who 'were unlikely to be aware of cross-contamination problems'. He drew attention to the staffing problems that a more ambitious programme would raise (see paragraph 2.440).

2.612 It was soon discovered that the assumption that all UKASTA members would have carried out a self-audit was unjustified. UKASTA had not kept all their members aware of MAFF's concerns about cross-contamination. Furthermore, non-UKASTA members were not aware that the attack rate experiment had shown that a single gram of infective material sufficed to transmit the disease. It was decided to send a letter of guidance to all mills. Dr Matthews commented that visits to mixed species mills should follow, and were more important (see paragraph 2.462).

2.613 UKASTA's cooperation was sought. They assisted in the preparation of a leaflet designed to be sent to feedmills and to farmers who mixed their own feed (see paragraph 2.489). This was a useful leaflet, but bland in tone. We find it surprising that it did not emphasise the finding that a very small quantity of infective material would suffice to transmit the disease. The leaflet was sent out on 10 August 1995.

2.614 A programme of SVS visits was planned, giving priority to mills producing feed for both ruminants and pigs and poultry (see paragraph 2.491).

2.615 Later in 1995 steps were taken to send out a letter to farmers, giving specific guidance on how to avoid cross-contamination which was much more detailed than that which had been included in the joint leaflet to feedmills and farmers. It was thought necessary that this should be approved both by SEAC and by the Parliamentary Secretary. SEAC approved it, with some suggested improvements, on 23 November 1995. It reached Mrs Browning on 6 February 1996 and received her immediate approval, but had not been sent out by 20 March.

2.616 The course of events described above suggest that in some respects there was a lack of urgency in the response of both MAFF officials and UKASTA to the facts that emerged in 1994.

2.617 The discovery that cross-contamination in feedmills was probably resulting in the infection of cattle came as a surprise. So did the result of the attack rate experiment. In the absence of countervailing reasons, we would have expected that the immediate reaction of UKASTA would have been to warn their members of this problem, and of the result of the attack rate experiment, and to give advice on how to minimise the risk of cross-contamination. Equally we would have expected MAFF officials to ensure that this was done, and to warn all feedmills that were not members of UKASTA in the same way, providing them with suitable guidance on avoiding cross-contamination. It appears to have taken 12 months for such action to be put in place.

2.618 We have noted earlier in this chapter (paragraph 2.584) that the primary measures that MAFF took in order to prevent contamination of cattle feed consisted of tightening up the implementation and enforcement of the animal SBO ban (see Chapter 4). It was reasonable for MAFF to give priority to those measures. So far as cross-contamination at feedmills was concerned, two factors should be borne in mind. First, it was reasonable for MAFF to seek to advance the development of the ELISA test. The perfection of this test, along with the tightening up of the animal SBO ban, would have diminished the dangers of cross-contamination considerably. Second, MAFF officials were anxious to avoid precipitating a waste disposal crisis if the industry reacted by refusing to use MBM in animal feed generally. UKASTA were clearly alive to this concern on the part of MAFF. In these circumstances we do not think that the failure by MAFF and UKASTA to move faster on additional measures in relation to cross-contamination of ruminant feed is a matter for criticism.

Tallow and gelatine

2.619 Before leaving this topic, we should draw attention to two further potential sources of BSE infection in cattle feed.

Tallow

2.620 Tallow was an ingredient of both ruminant and non-ruminant feed. The ruminant feed ban applied to ruminant protein. At the time that the ban was introduced it was considered that tallow that was extracted in the course of the rendering process did not contain protein, so that the ban did not preclude tallow's continued incorporation in animal feed. Thus UKASTA advised their members through Feed Circular 412 that:

The Order does not cover fishmeal, poultry meal, poultry offal or tallow. ¹⁸

2.621 In 1991 approximately 5,000 tonnes of tallow was used in cattle and calf rations. ¹⁹

2.622 By 1994 MAFF officials had made the following discoveries:

1. Much of the tallow being used in the production of animal feed was derived from SBOs.
2. Tallow might include between 0.1 and 0.3 per cent protein.
3. Distillation residues which settled as sediment in tank bottoms were sold for animal feed. Dr Taylor of NPU considered the distillation process and advised that it was likely to inactivate the BSE agent.

2.623 SEAC were not concerned at the use of tallow in cattle feed provided that it was not derived from SBOs. In these circumstances MAFF officials recommended that a 'technical amendment' be made to the 1991 Order to allow minimum quantities of ruminant protein to be fed to ruminants. Mr Gummer rejected this suggestion for presentational reasons, but indicated that he was content that tallow, other than that derived from SBOs, should continue to be incorporated in animal feed. ²⁰ There is scope for argument as to whether this constituted turning a Ministerial blind eye to the possibility of technical infringements of the law. ²¹ Having regard to SEAC's risk assessment we do not think it involved significant risk to animal health.

2.624 In accordance with SEAC's advice, the animal SBO ban was subsequently extended to include tallow derived from SBOs. ²² It must be possible that prior to this extension the occasional BAB was infected as a result of the incorporation in cattle feed of tallow derived from SBOs. Having regard to the probable inactivating effect of the distillation process, we do not consider that this is likely to have been a significant source of infection.

Gelatine

2.625 The bones from which gelatine was produced included, on occasion, the skulls of cattle. Although under the SBO ban the brain should have been removed from any skull used for this purpose, traces of brain will often have remained. Equally, spinal column used in the manufacture of gelatine may, on occasions, have included segments of spinal cord. The processes involved in the manufacture of gelatine may not always have resulted in the inactivation of such infectious matter. Waste food, containing gelatine, was commonly incorporated in animal feed. This, then, was a further potential source of infection of cattle feed, although once again we do not consider that the scale of any transmission by this route can have been significant.