Report on the circumstances leading to the 1996 outbreak of infection with E.coli 0157 in Central Scotland, the implications for food safety and the lessons to be learned.

ANNEX VI (Para 7.37 refers)

REPORT BY THE ACMSF AD HOC WORKING GROUP, MARCH 1997

Background

1. The Pennington Group was convened by the Secretary of State for Scotland on 28 November 1996 to advise him on the implications for food safety and the general lessons to be learned from the E.coli O157 food poisoning outbreak in Central Scotland. It delivered its Interim Report to the Secretary of State on 31 December. The report included a recommendation (4(d)) that the Advisory Committee on the Microbiological Safety of Food (ACMSF) should review the guidance on cross-contamination contained in its 1995 Report on Verocytotoxin-Producing Escherichia coli (the 1995 Report). The Chief Medical Officer for Scotland wrote on 14 January to the Chairman of ACMSF asking the Committee to review its guidance on cross-contamination in the context of the Pennington recommendations for the licensing of premises handling, producing or selling cooked meat or cooked meat products (4(a)) and for the physical separation of raw and cooked meat products (4(c)). The Secretary of State for Scotland also referred to ACMSF the Group's recommendations on research (paras 2.2, 2.4 and 2.5).

2. The ACMSF considered these requests at its meeting on 21 January which was attended by Professor Pennington and Mr Stephen Rooke (a Pennington Group member). The Committee decided to set up an Ad Hoc Working Group to examine the issues in more detail. The Chairman of the ACMSF wrote to CMO Scotland on 30 January to inform him of the Committee's decision and to request information about the circumstances surrounding the Lanarkshire outbreak on which it could base its review of its guidance on cross-contamination.

Terms of Reference and Membership

3. The Ad Hoc Working Group was given the following terms of reference by the ACMSF:

To advise the Advisory Committee on the Microbiological Safety of Food on a response to the requests from the Secretary of State for Scotland and The Scottish Office on matters arising from the Pennington Group.

Meetings and Action Taken

4. We met four times on 3, 17 and 25 February and 6 March. We were assisted by Dr Rosalind Skinner (SODoH), Dr Richard Harding (MAFF) and Mr Tom Murray (DH) all of whom attended meetings of the Pennington Group.

5. After our first meeting Professor Georgala, as the Chairman of the Working Group, sent a letter to CMO Scotland setting out our immediate reactions to the matters which had been referred to us, repeating the request for further information and suggesting that the best way forward would be for the Pennington Group and the ACMSF to work together. We also asked for comments to be conveyed to the Pennington Group.

6. Professor Pennington attended our second meeting accompanied by two members of his Group, Mrs Ann Foster and Dr John Cowden and by Mr K McIntosh and Mr R Weatherston from The Scottish Office. As well as sharing with us information about the circumstances surrounding the outbreak the Pennington Group also agreed to give further consideration to the recommendations set out in paras 10 to 13 below.

Evidence Considered

7. We considered the issues before us in the light of two types of evidence. First, we had discussions with Professor Pennington and members of his Group about the factors which led them to the conclusions in their Interim Report. Second, the limited information available to the Pennington Group on the circumstances surrounding the outbreak was made available to us and discussed with Professor Pennington and the other members of his Group at our second meeting. For legal reasons concerned with pending criminal proceedings and the fatal accident inquiry this information is not set out in this report. We anticipate that the full facts of the outbreak will emerge in due course but a full report on the circumstances of the outbreak was not available. We have also taken into account in our findings the results of a survey carried out by the Meat and Livestock Commission at the request of the Secretary of State for Scotland.

Prevention of Cross Contamination

8. As recommended by the Pennington Group, the ACMSF was asked to reconsider the recommendations in its 1995 Report on the prevention of cross contamination. Important factors which distinguish E.coli O157 from most other foodborne pathogens are the low infective dose and the potential serious consequences of infection. Only very small numbers of the organism may be required to cause illness and as a consequence growth of the organism in food is not necessarily required for contaminated food to cause illness. If growth does take place the consequences are likely to be even more serious. In 1995 the ACMSF published a comprehensive report (the 1995 Report) on Verocytotoxin-producing Escherichia coli. This report set out these distinctive features of the organism which were well known at the time and took them fully into account in drawing up its guidelines (see Chapters 2, 3, 4 and 6 of the 1995 Report). These guidelines remain current but we have reconsidered them in the light of the views of the Pennington Group and all other relevant factors referred to in para 8 above.

9. We conclude, as the Committee did before, that the avoidance of cross contamination is extremely important in the case of E.coli O157. Having considered all the evidence, we remain of the view that cross contamination can be effectively avoided by separation in time or by space, if suitable procedures are properly applied. We have not found any reason to depart from the recommendation of the 1995 Report that the prevention of contamination is best assured by good hygiene and manufacturing practices as determined by hazard analysis and control at critical points of the operation. We are also aware of the risk that compulsory physical separation of cooked and raw meat could lull management and staff into a false sense of security. Further, there may be some instances (for example where staff share other facilities) where time separation of operations may be equally or more effective than physical separation. A proper application of the principles of hazard analysis will indicate what procedures need to be appropriately applied in each individual case. This may well mean physical separation of raw and cooked meats will be necessary in some (but not all) businesses. The essence of food safety is that everyone involved in the production of food has sufficient understanding of the operations to appreciate the importance of necessary safety procedures.

10. We therefore strongly support the Pennington Group's recommendation (4(f)) in their Interim Report that the implementation of HACCP should be accelerated in high risk premises. Indeed we believe this recommendation should go further and should be extended to cover not only premises in categories A and B as defined in Code of Practice No 9, but category C as well. Implementation of existing legislative requirements for hazard analysis and their enforcement together with appropriate training should be taken forward as a matter of urgency. Hazard analysis would lead to consideration of all the elements contributing to safe food production and take account of different levels of training skills and equipment. The measures and procedures that would flow from hazard analysis would address and be proportionate to the hazards identified in individual circumstances. It will identify those processes and procedures for which physical separation is essential for food safety. As such it provides greater assurance than any practicable prescriptive legislative requirements.

11. In addition, we believe that it is essential that hazard analysis plans applied to individual premises should be documented where appropriate. Written procedures to be followed in individual premises would greatly assist management and staff in their day to day operations, and also enforcement. We recommend that the Codes of Practice under Section 40 of the Food Safety Act, which are currently being reviewed, should contain guidance to enforcement officers to ensure the efficient and effective enforcement of the hazard analysis requirement in the categories of premises mentioned in paragraph 10.

12. We would agree with the conclusion in the Interim Report of the Pennington Group that implementation of hazard analysis has been too slow. While we appreciate that it cannot be introduced immediately we believe that it is essential that there is urgent action to take this forward expeditiously for high risk premises. We recommend an Industry Guide to good hygiene practice be developed in accordance with article 5, 6 and 7 of Council Directive 93/43 EEC (and deployed by enforcement authorities as described in Regulation 8(2)(c)(ii) of the Food Safety (General Food Hygiene) Regulations 1995) by the relevant trade bodies in consultation with enforcement authorities, representatives of consumers and Government. The Industry Guide should be specifically for butchers and producers and retailers of cooked meats and cooked meat products. Its purpose would be to show in practice what was necessary to comply with legislation and to confirm good manufacturing practice and would provide both a basis for operations within industry and help for those responsible for inspection and enforcement. In particular the Guide should focus on the basic steps of good manufacturing practice, key issues in hazard analysis and training. It should set out a methodology for the application of hazard analysis in these kinds of businesses and the training required to make this effective. This will require action by Government, the enforcement authorities and the industry itself but the necessary legislation is already in place to enable it to happen. We understand that such a Guide could be produced within a short period.

13. Training is vital to this approach. Training/instruction is already a legal requirement, and we recommend that, where this has not been given, industry ensures that it is undertaken without delay. In our view, this approach should go beyond food handlers themselves, and embrace management and supervisory staff. Staff at these levels will also require training in safe food manufacture and hazard analysis. We expect that the Industry Guide would fully describe the training required.

14. There is also the corresponding need to ensure that enforcement officers also posses an appropriate level of expertise to adequately enforce the requirements of the hazard analysis sections of the regulations.

Research Recommendations

15. The Pennington Group's Interim Report recommended research in three areas: the prevalence in Scottish cattle of the sub-clone of E.coli O157 which caused the recent outbreaks, forecasting the likely future incidence/prevalence of E.coli O157 and its sub-clones in Scottish cattle, and in the resolution of the DNA-based typing methods. We believe that any information to be collected in Scotland needs to be matched by comparable information on the prevalence of E.coli O157 in the rest of the UK to provide a broader perspective. Because of the factors which can affect the apparent prevalence of this organism among livestock, we recommend careful attention is paid to sampling protocols to ensure that whatever information is collected is reliable and comparable. As for typing methods, we welcome and support the Pennington Group's proposals to commission further research to increase the resolution of DNA-based typing methods. We believe such research should cover not only DNA analysis but other approaches as well.

16. We strongly recommend that any further work in these areas should be considered in the context of work already under way. Wherever appropriate this work should be commissioned through open competition inviting proposals which would then be subject to peer review.

The Wider Context

17. The Pennington Group's Interim Report and the work we have been asked to undertake have both focused, for obvious reasons, on the manufacture and sale of meat and meat products, both raw and cooked. However it should always be remembered that E.coli O157 poses a danger in all sections of the food industry and vehicles other than meat have been responsible for outbreaks. We would therefore urge that the principles advocated in the ACMSF's 1995 Report should be applied throughout the food chain. The ACMSF's 1995 Report recommended this and current events have only served to underline the importance of this task and the urgency with which it should be undertaken.

Summary of Recommendations

18. We recommend that:

18.1 Procedures to ensure effective separation of raw and cooked meats in all premises should be determined by the application of hazard analysis (para 9).

18.2 Given the urgency of the situation every effort should be made to implement and stringently enforce existing legislative requirements for hazard analysis and training (para 10).

18.3 Hazard analysis plans should be documented where appropriate (para 11).

18.4 Codes of Practice under Section 40 of the Food Safety Act which give guidance on efficient and effective enforcement of the hazard analysis requirement for the categories of premises mentioned in paragraph 10 should be developed (para 11).

18.5 An Industry Guide to good hygiene practice should be urgently developed under the Food Safety (General Food Hygiene) Regulations 1995 specifically for butchers and producers and retailers handling or preparing cooked meat and cooked meat products (para 12).

18.6 Appropriate training for food handlers, including supervisors and managers, is vital and should be accelerated by this sector of the industry (para 13).

18.7 The collection of data on the prevalence of E.coli O157 in livestock should be planned on a UK basis (para 15).

18.8 Further research into strain typing should cover not only DNA analysis but other approaches as well (para 15).

18.9 Such research should take account of work already in hand, be commissioned by open competition and be peer reviewed (para 16).

18.10 The butchery sector is not the only part of the food industry at risk from E.coli O157 and the principles advocated here should be applied throughout the food chain as recommended in the ACMSF's 1995 Report (para 17).