EUROPEAN COMMISSION DIRECTORATE-GENERAL FOR HEALTH AND FOOD SAFETY

EUROPEAN COMMISSION DIRECTORATE-GENERAL FOR HEALTH AND FOOD SAFETY Health and food audits and analysis DG(SANTE) 2016-8737 - MR FINAL REPORT OF AN AUDIT CARRIED OUT IN THE UNITED KINGDOM FROM 11 TO 21 OCTOBER 2016 IN ORDER TO EVALUATE THE CONTROLS ON CONTAMINANTS IN FOOD In response to information provided by the competent authority, any factual error noted in the draft report has been corrected; any clarification appears in the form of a footnote.

Executive Summary This report describes the outcome of an audit in the United Kingdom carried out from 11 to 21 October 2016, as part of the published Directorate General for Health and Food Safety audit programme. The objective of the audit was to verify that official controls in the area of contaminants in food are carried out in accordance with Regulation (EC) No 882/2004 of the European Parliament and of the Council and in compliance with that Regulation and other relevant EU legislation, in particular the implementation of Commission Regulation (EC) No 1881/2006 and related legislation with regard to nitrates, mycotoxins, heavy metals, industrial and environmental contaminants. There is an adequate risk-based system of official controls as regards chemical contaminants in foodstuffs throughout the whole production chain. It allows for satisfactory planning and implementation of the controls, identifies non-compliances and enforces necessary and proportionate corrective actions. There is a sufficient network of official control laboratories. Nonetheless, this system shows some weaknesses that can (a) result in incomplete assessment of food business operators' HACCP-based procedures, (b) call into question the reliability of official sampling and analysis, in particular in terms of representativeness of the samples taken, methods used for sample preparation, availability of replicate samples in the official control laboratories and interpretation of analytical results, and (c) delay undertaking of effective actions due to lengthy turnaround times by the official control laboratories. The report contains recommendations to the competent authorities of the United Kingdom to address the shortcomings identified. I

Table of Contents 1 Introduction...1 2 Objectives and scope...1 3 Legal Basis...2 4 Background...3 5 Findings and Conclusions...3 5.1 Relevant National Legislation and National Guidelines...3 5.2 Organisation and Implementation of Official Controls...4 5.2.1 Designation of Competent Authorities and Resources for Official Controls...4 5.2.2 Registration of Food Establishments...5 5.2.3 Organisation and Scope of Official Controls...6 5.2.4 Procedures for Performance and Reporting of Control Activities...8 5.2.5 Implementation of Official Controls on Food Business Operators' Obligations...9 5.2.6 Official Sampling Procedure...10 5.2.7 Laboratory Performance...12 5.2.8 National Enforcement Measures...15 5.3 Rapid Alert System for Food and Feed...17 6 Overall Conclusions...17 7 Closing Meeting...17 8 Recommendations...18 II

ABBREVIATIONS AND DEFINITIONS USED IN THIS REPORT Abbreviation CA CCA DG Health & Food Safety EC EFSA EU EURL FBO FSA FSS FTE GAP GMP HACCP LA MANCP MS MU NRL OCL PAHs PT RASFF SOP TSO UK UKAS UKFSS Explanation Competent Authority Central Competent Authority Directorate General for Health and Food Safety European Commission European Food Safety Authority European Union European Union Reference Laboratory Food Business Operator Food Standards Agency Food Standards Scotland Full Time Equivalent Good Agricultural Practice Good Manufacturing Practice Hazard Analysis and Critical Control Points Local Authority Multiannual National Control Plan Member State Measurement Uncertainty National Reference Laboratory Official Control Laboratory Polycyclic Aromatic Hydrocarbons Proficiency Test Rapid Alert System for Food and Feed Standard Operation Procedure Trading Standards Officer The United Kingdom United Kingdom Accreditation Service UK Food Surveillance System III

1 INTRODUCTION The audit took place in the United Kingdom (UK) from 11 to 21 October 2016. The audit team comprised two auditors from Directorate General for Health and Food Safety (DG Health & Food Safety) and one expert from a European Union (EU) Member State (MS). The audit was undertaken as part of the DG Health & Food Safety planned audit programme. Representatives from the central competent authority (CCA), the Food Standards Agency (FSA), accompanied the audit team for the duration of the audit. An opening meeting was held on 11 October 2016 with the FSA and a representative of national reference laboratories (NRLs). At this meeting, the objectives of, and itinerary for, the audit were discussed and confirmed and the control systems were described by the FSA. 2 OBJECTIVES AND SCOPE The main objective of the audit was to verify that official controls in the area of contaminants in food are carried out in accordance with Regulation (EC) No 882/2004 of the European Parliament and of the Council and in compliance with that Regulation and other relevant EU legislation, in particular the implementation of Commission Regulation (EC) No 1881/2006 and related legislation with regard to nitrates, mycotoxins, heavy metals and industrial and environmental contaminants in food. In terms of scope, the audit reviewed planning and implementation of official controls, control procedures, official sampling and laboratory performance. The scope of the audit does not cover environmental contaminants in food of animal origin and environmental contaminants in fishery products. In particular, the audit evaluated the effectiveness of the official controls and the implementation of: Regulation (EC) No 178/2002 of the European Parliament and of the Council of 28 January 2002 laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety; Regulation (EC) No 852/2004 of the European Parliament and of the Council of 29 April 2004 on the hygiene of foodstuffs; Regulation (EC) No 882/2004 of the European Parliament and of the Council of 29 April 2004 on official controls performed to ensure the verification of compliance with feed and food law, animal health and animal welfare rules; Commission Regulation (EC) No 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs; Commission Regulation (EC) No 401/2006 of 23 February 2006 laying down the methods of sampling and analysis for the official control of the levels of mycotoxins in foodstuffs; 1

Commission Regulation (EC) No 1882/2006 of 19 December 2006 laying down methods of sampling and analysis for the official control of the levels of nitrates in certain foodstuffs; Commission Regulation (EC) No 333/2007 of 28 March 2007 laying down the methods of sampling and analysis for the control of the levels of trace elements and processing contaminants in foodstuffs. In pursuit of these objectives, the following sites were visited: Table 1: Audit visits and meetings Visits/meetings Competent Authorities Comments Central 2 Opening and closing meeting with the representatives of FSA and NRLs Local 9 Representatives of the local authorities were met during on-site visits Official Control Laboratories (OCLs) Public 1 Private 2 One OCL was visited on site, representatives from the other OCL were met without visiting the OCL Establishments Importer and processor of nuts and other foodstuffs Producer of foods for infants and young children Producer of canned foods of non-animal origin Producer of breakfast cereals and related foods Producer of green leafy vegetables 1 Large establishment 1 Large producer 1 Large producer 2 Large producers 1 Medium sized primary producer Producer of cereals 1 Medium sized primary producer Producer of smoked fish 2 A large and a medium sized producer 3 LEGAL BASIS The audit was carried out under the general provisions of EU legislation, in particular Article 45 of Regulation (EC) No 882/2004. 2

Legal acts quoted in this report refer, where applicable, to the last amended version. Full legal references are given in Annex 1. 4 BACKGROUND Regulation (EC) No 1881/2006 sets maximum levels for certain contaminants in foodstuffs. The maximum levels are set based on toxicological evaluations carried out by the European Food Safety Authority (EFSA). EFSA's scientific opinions on the risks posed by the regulated contaminants in foodstuffs can be found at: http://www.efsa.europa.eu/en/publications. Information on the existence of direct or indirect risk to human health deriving from food or feed is disseminated via notifications through the Rapid Alert System for Food and Feed (RASFF) to all MSs and to exporting countries. In 2015, a total of 3049 original notifications were transmitted through the RASFF. As regards chemical contaminants in food, 495 notifications (16%) concerned mycotoxins, 219 notifications (7%) concerned heavy metals, 53 notifications (2%) concerned industrial contaminants and 8 notifications (0.3%) were for other chemical contamination. Aflatoxins in nuts, nut products and seeds comprise the largest group of RASFF notifications in recent years. This was the first DG Health & Food Safety audit to the UK on this topic. 5 FINDINGS AND CONCLUSIONS 5.1 RELEVANT NATIONAL LEGISLATION AND NATIONAL GUIDELINES Legal requirements Article 291 of the Treaty on the Functioning of the EU. Findings 1. The directly applicable EU legislation in the scope of the audit has been further implemented through the Food Safety Act 1990 and the Food Law Code of Practice (parallel codes exist for all four nations of the UK). 2. The FSA works closely with various stakeholders to identify and develop good agricultural practices (GAP) and good manufacturing practices (GMP) to prevent and reduce the presence of contaminants in food. These include primary producers, food processors and manufacturers, trade associations, academics and experts in analytical methods. In some cases, the FSA commissions the work to develop GAP and GMP, in others, the FSA works closely with industry to develop these GAP and GMP which are then disseminated to the industry and publicised by the FSA. 3

3. In the current planning, any implementation of the EU s and national Guidelines and Codes of Practice by the food business operator (FBO) does not result in any significant changes concerning frequency and control methodology of the official controls. Conclusions on Relevant National Legislation and National Guidelines 4. Through the existing legal framework at national level, the competent authorities (CAs) have adopted all measures of national law necessary to implement legally binding Union acts. The CAs actively participate in the development of GAP and GMP to ensure the safety of food. 5.2 ORGANISATION AND IMPLEMENTATION OF OFFICIAL CONTROLS 5.2.1 Designation of Competent Authorities and Resources for Official Controls Legal requirements Articles 4(1) and 6 of Regulation (EC) No 882/2004. Findings 5. In accordance with Article 4(1) of Regulation (EC) No 882/2004, the CAs are clearly designated. 6. An overview of how control systems are organised in the UK, based on information supplied by them, is provided in the Country Profile for the UK and is available at the following link: http://ec.europa.eu/food/audits-analysis/country_profiles/details.cfm?co_id=gb 7. The FSA is the CCA in the scope of the audit. The detailed description of the FSA's role, structure and activities is presented in the Country Profile. 8. The majority of food law enforcement is delegated to local authorities (LAs) throughout the UK who carry out checks on all food businesses in their area to ensure compliance with food safety, traceability and labelling requirements. LAs are defined as food authorities under the Food Safety Act 1990. 9. LAs are responsible for enforcing the legislative provisions, and the Official Food and Feed Regulations 2009 (parallel versions exist for the four nations of the UK) set out the responsibilities of the CA and certain local responsibilities in relation to Regulation (EC) No 882/2004. Enforcement is carried out by Environmental Health or Trading Standards Departments and authorised officers are typically Environmental Health Officers or Trading Standards Officers (TSOs). Chapter 3 of the Multi-Annual National Control Plan (MANCP) for the UK 2013-16 provides more information and can be found at: http://www.food.gov.uk/enforcement/regulation/europeleg/feedandfood/ncpuk 4

10. The Food Law Code of Practice is issued under the Official Feed and Food Controls Regulations, the Food Safety Act and the Food Hygiene Regulations, and is directed to LAs enforcing food legislation in the UK. Parallel codes exist for all four nations of the UK. Authorities are required to have regard to this Code when undertaking their duties and the FSA has powers to audit LAs in accordance with the Food Standards Act: http://www.food.gov.uk/enforcement/codes-of-practice/food-law-code-of-practice-2015 11. Local Trading Standards Authorities have an enforcement role with respect to ensuring that food law is complied with and this is normally done at the point of sale (including retail and catering level). Chapter 1 of the Food Law Code of Practice (parallel codes exist for all four nations of the UK) outlines arrangements between CAs, and the division of enforcement responsibilities. 12. The FSA keeps enforcement authorities and FBOs abreast of developments in EU legislation via a number of routes. Interested parties letter (called "ENF letters to enforcement") updates on the FSA website and topic specific meetings with individual stakeholders groups. 13. In 2014-15, LAs reported a total of 2,303 full time equivalents (FTEs) delivering food hygiene and food standards work; this includes 343 FTEs in Scotland. 14. Training enabling the staff performing official controls to undertake their duties competently and to carry out official controls in a consistent manner is regularly arranged. However, based on meetings and interviews with LA inspectors (see paragraphs 40, 41, 42 and 49), the audit team concluded that the training provided did not allow LA inspectors to assimilate all the skills enabling them to undertake their duties competently and to carry out official controls in a reliable manner (this relates predominantly to the assessment of HACCP plans and interpretation of analytical results). This is not in line with the requirements laid down by Article 6 of Regulation (EC) No 882/2004. Conclusions on Designation of CAs and Resources for Official Controls 15. The CAs have been designated and in general avail of sufficient resources to carry out official controls. 16. The training provided is not sufficiently robust to ensure that LA inspectors undertake all their duties competently and carry out official controls in a reliable manner. 5.2.2 Registration of Food Establishments Legal requirements Article 6 of Regulation (EC) No 852/2004, Article 31 of Regulation (EC) No 882/2004. Findings 17. In the UK countries under remit of the FSA, most types of food business, including catering businesses run from home and mobile or temporary premises, such as stalls and vans, are required to register (and apply for approval, when necessary) all premises with their LA/Food Authority at least 28 days before opening. This is in line with the 5

requirements laid down by Article 31 of Regulation (EC) No 882/2004. The registration can be made online through a dedicated internet portal. 18. For Scotland, primary food producers are required to be registered (and approved, when necessary) with their LA. Making use of the provision in Article 31(1)(b) of Regulation (EC) No 882/2004, producers are registered by the Scottish Government s Rural Payments Inspections Division and this information is shared with Food Standards Scotland (FSS) and LAs. Template registration forms are found within the Food Law Code of Practice (Scotland). Lists of registered food businesses are collated by the LA. Conclusions on Registration of Food Establishments 19. The mechanisms that the CAs have put in place to register FBOs are sufficient. 5.2.3 Organisation and Scope of Official Controls Legal requirements Article 17 of Regulation (EC) No 178/2002 and Article 3 of Regulation (EC) No 882/2004. Findings 20. There is a risk-based approach in place for the planning of official controls regarding chemical contaminants in foodstuffs, which is in line with the requirements laid down by Article 3(1) of Regulation (EC) No 882/2004. 21. In the process of planning of the official controls, priorities are decided based on risks evaluated from analysis of RASFF notifications, intelligence gained from enforcement authorities and other MSs and European Commission (EC) Recommendations to monitor some contaminants. On this basis, an annual control plan is created. The plan includes types of commodities, types of analytes (contaminants) and numbers of samples that shall be taken by the LAs. 22. Official control samples are taken throughout the whole distribution chain. The sampling plan designed by the FSA is funded by the FSA. In addition, LAs may have their own sampling plans (funded locally) in case of specific priorities (for example, based on previous experience) identified by the LAs. 23. For the purpose of planning of official controls (Trading Standards official controls cover chemical contaminants), all FBOs are rated in the following classes: Category A: high risk FBOs, minimum intervention frequency every 12 months Category B: medium risk FBOs, minimum intervention frequency every 24 months Category C: low risk FBOs, alternative enforcement strategy or intervention every five years 24. The audit team noted that some LAs use National Trading Standard Board risk rating (as of September 2012): Category A: high risk FBOs, interventions required every 12 months Category B1: upper medium risk FBOs, interventions required every 24 months Category B2: lower medium risk FBOs, interventions required every 24 months 6

Category C: low risk FBOs, alternative enforcement strategy or intervention every five years Unrated FBOs 25. The audit team noted that predominantly high risk FBOs are chosen for routine official controls 1. 26. In Scotland, the inspection programme is determined by LAs subject to risk rating. The Resource Working Group determines sampling priorities for LAs. The FSS provides sampling grants for LAs to look at additional sampling priorities. 27. For the year 2015-16 in relation to contaminants, for example, the EC had recommended monitoring for ergot alkaloids, pyrrolizidine alkaloids and tropane alkaloids; these were included. RASFFs on aflatoxins in maize products from Africa and India continue to be issued, therefore these were also monitored. A few of the priorities for sampling were acrylamide from food service as part of the ongoing work to develop controls in this area and to help inform EU risk management discussion. Another priority was arsenic in rice and fruit juice, as the new EU maximum limits for arsenic in rice would come into force at the start of 2016. 28. During 2015-16 as well as in preceding years, the measurement of polycyclic aromatic hydrocarbons (PAHs) in banana chips, dried herbs and spices and herbal supplements were included in national coordinated monitoring as these were recognised as higherrisk commodities and there were ongoing discussions with the EC with regard to the setting of limits. UK data were presented to the EC during the negotiations. PAHs limits in these commodities were eventually introduced and monitoring in the UK continued to assess the extent of compliance with the new limits. 29. For 2016-17, sampling priorities remain acrylamide due to ongoing work to mitigate its occurrence and arsenic to ensure that the adapted limits would be respected. A number of issues were reported for erucic acid in mustard oil: several non-compliances were reported, with some FBOs trying to get around the maximum levels by selling the oil as non-food. Mustard oil (both edible and non-food) will be monitored and advice on health effects will be provided. Another example is the higher levels of fusarium mycotoxins reported in European-grown maize due to adverse weather conditions. Therefore, maize and maize products will be monitored. Following recent concern from cyanide in apricot kernels, these have also been included in the sampling. 30. To avoid monitoring the same issues year on year without compelling reasons, for the current food sampling programme the FSA requested from LAs information for prioritisation. It covers, among other things, what the data will be used for, the risks involved and whether enforcement authority sampling is appropriate. 31. The FSA makes funding available for sampling and surveillance of food to help ensure risk-based, targeted checks at ports and LAs monitoring throughout the food chain. 1 In their response to the draft report the Competent Authority noted that all establishments, regardless of risk, are subject to an official control. However, local authorities are required to have a risk based approach to their programme therefore higher risk establishments are prioritised over lower risk. 7

32. The results of laboratory analysis on official sampling are stored in the UK Food Surveillance System (UKFSS). UKFSS is a database which stores food and feed control sampling data as part of enforcement activities undertaken by LAs across the UK. The system is used on a voluntary basis; it is used by all LAs in Scotland and Northern Ireland, and is being used by 231 and 21 LAs in England and Wales, respectively. Data stored can be used by enforcement authorities and the FSA to provide evidence to inform risk-based sampling, to allow early identification of trends in food hazards and to provide a mechanism for reporting LA data to the FSA and the EC. It also provides the FSA and enforcement authorities with a means of co-ordinating sampling across authorities. The FSA will continue to support the sharing of sampling results and ensure coordination of enforcement activity of LAs. Reports on this programme can be found at: http://www.food.gov.uk/enforcement/monitoring/samplingresources/samplingandsurveil lance 33. The FSA and FSS are working on developing a risk based prioritisation tool for primary producers. Conclusions on Organisation and Scope of Official Controls 34. The system of the organisation of official controls is in place and satisfactory. The planning of official controls is risk-based. 5.2.4 Procedures for Performance and Reporting of Control Activities Legal requirements Articles 8 and 9 of Regulation (EC) No 882/2004. Findings 35. For the trading standards official controls, the LAs are supported by checklists developed by LAs based on guidance from the FSA and FSS. The audit team noted that the checklists observed do not sufficiently cover issues of importance for chemical contaminants in raw materials, finished products and processes (for example, no section on possible chemical contaminants in food and adequate risk assessment of the chemical contaminants in the FBOs' HACCP-plans; no section on whether the FBOs' own checks adequately address possible chemical contaminants in relevant products). 36. After every official control, an inspection report is prepared and provided to the FBO. If urgent action is necessary, this information is given to the FBO during the inspection. In the UK, via the respective Food Law Codes of Practice, LAs are required to provide specified information (including stipulated by Article 9(2) of Regulation (EC) No 882/2004) to FBOs at the conclusion of the intervention. There is no central database on the official control reports in the UK. The audit team observed some official control reports and noted that, contrary to the provisions of the UK Food Law Codes of Practice, not all information specified by Article 9(2) of Regulation (EC) No 882/2004 was in the reports observed (for example, information on the control methods applied, and, where appropriate, action that the business operator concerned is to take, was missing). 8

37. The FSA has in place verification procedures as regards official controls carried out by LAs. The audit team was presented with two supervision reports carried out in 2015 and 2016. The FSA stated that in case of deficiencies in performance of controls by LAs, the situation is followed up within six months from the date of the original supervision. A detailed description of the verification and review of official controls is presented in the Country Profile, as applied in accordance with the requirements laid down by Article 8(3)(a) of Regulation (EC) No 882/2004. Conclusions on Procedures for Performance and Reporting of Control Activities 38. The procedures to carry out official controls are in place and in general adequate. There is a sufficient system of verification of the effectiveness of the official controls. The official controls are documented in a proper manner and the reports are provided to the establishments inspected. Deficiencies in addressing chemical contaminants in checklists, and incompleteness of the written reports, may inhibit the functioning of the system of official controls. 5.2.5 Implementation of Official Controls on Food Business Operators' Obligations Legal requirements Articles 10(2)(b)(i), (c), (d) and 15(1) of Regulation (EC) No 882/2004, Articles 4(1), 7 and 8 and Part B of Annex I to Regulation (EC) 852/2004, Article 18 of Regulation (EC) No 178/2002, Regulation (EC) No 401/2006, Regulation (EC) No 1882/2006 and Regulation (EC) No 333/2007. Findings 39. In general, the official controls are implemented in line with the requirements laid down by Article 3(2-3) of Regulation (EC) No 882/2004. 40. To assess implementation of the requirements laid down by Article 10(2) of Regulation (EC) No 882/2004, the audit team observed nine official controls, including two at primary producers (the profiles of the FBOs visited are given in Table 1). During these inspections, the audit team observed LA TSOs' assessment of the FBOs' Hazard Analysis and Critical Control Points (HACCP) plans (risk assessments at the two primary producers) related to chemical contaminants. At four of the nine FBOs, the LA TSOs made a correct assessment of the establishment's risk analysis. In five of the nine FBOs, some deficiencies were noted by the audit team: At a manufacturer of smoked fish, risk for occurrence of PAHs was not considered in the HACCP plan. At a manufacturer of nut, cereal and fruit based snacks and two manufacturers of ready to eat cereals and related foods, the FBOs HACCP plans did not address occurrence of chemical hazards in some ingredients (e.g. vegetable oil, salt) and, for one of manufacturers of ready to eat cereals and related foods, occurrence of process contaminants (acrylamide, PAHs). At a manufacturer of baby food, a team of two LA TSOs did not assess the FBO s HACCP plan in relation chemical contaminants. The audit team was informed that 9

routine trading standards inspections do not cover assessment of FBOs chemical hazards identification. This is not in line with the requirements laid down by Article 3(1)(a) and Article 10(2)(d) of Regulation (EC) No 882/2004. Without checking HACCP plans, the inspectors cannot assess whether the identification of chemical hazards is appropriate. 41. The deficiencies described in paragraph 40 were not spotted by the LA TSOs during the official controls observed. 42. The LA TSOs met during the nine official controls were interviewed on the procedure to make a (non-)compliance decision based on the analytical result of official samples. At all nine visits, the TSOs met did not correctly identify all parameters necessary for a valid (non-)compliance decision, i.e. that expanded measurement uncertainty (MU) and correction for analytical recovery shall be reported by the OCLs and shall be taken into account for the purpose of making a (non-)compliance decision in accordance with respective provisions laid down by Regulation (EC) No 401/2006, part B.4 of Regulation (EC) No 1882/2006 and part D.2 of Regulation (EC) No 333/2007. Since compliance/non-compliance decisions are to be made by LAs, lack of knowledge on the above mentioned legislative provisions may lead to legally contestable decisions. Conclusions on Implementation of Official Controls on Food Business Operators' Obligations 43. The system of implementation of official controls is in place and in general adequate. 44. Official controls do not adequately verify FBOs' risk assessments and thus overlook gaps in HACCP-based procedures, in particular in relation to chemical contaminants in ingredients and process contaminants. In one LA, absence of adequate verification of FBOs' risk assessment inhibits the functioning of the system of official controls. 45. The competencies of LA TSOs on the interpretation of analytical results of official samples present some deficiencies. This might lead to incorrect compliance or noncompliance decisions by the LAs. 5.2.6 Official Sampling Procedure Legal requirements Article 11 of Regulation (EC) No 882/2004, Regulation (EC) No 401/2006, Regulation (EC) No 1882/2006 and Regulation (EC) No 333/2007. Findings 46. In accordance with the requirements laid down by Article 10(1) of Regulation (EC) No 882/2004, sampling is undertaken as a part of the system of official controls. 47. The official samples comprise informal (taken without taking account of the relevant legislative provisions) and formal samples (taken in full compliance with the relevant legislative provisions). The audit team observed that a large proportion of samples reported through UKFSS are informal samples. The audit team noted that every LA has 10

its own enforcement policy. The enforcement policies observed contained no description of the procedures to follow when an informal sample happens to be non-compliant due to high levels of contaminant(-s); with the exception of the procedure provided by the FSS. The lack of clear procedure when a non-compliant informal sample is found makes the legal status of informal samples somewhat unclear for the purposes of the official controls 2. 48. The FSA monitoring program on nitrates is subcontracted to a private company and representatives of the company take all official samples within the programme. The audit team met representatives from the company. The standard operation procedure (SOP) applied for the official sampling on nitrates was presented. The audit team noted that there were references to two different procedures, one based on Directive 2002/63/EC (on methods of sampling for the official control of pesticide residues in and on products of plant and animal origin) and another based on Regulation (EC) No 1882/2006. This leads to ambiguity on what sampling procedure is applied by the company. The audit team was presented with a video record of an official sampling (nitrates in leafy green vegetables in field). The sampling was performed in accordance with the requirements laid down by Regulation (EC) No 1882/2006. Adequate measures are undertaken to ensure that the sample is transferred to the laboratory and the sample preparation takes place within 24 hours of sampling, and that the sample is kept cool during transport (if this is not possible the sample is deep-frozen within 24 hours of sampling and kept frozen up to a maximum of six weeks). This is in accordance with the requirements laid down by part A.3.7 of the Annex to Regulation (EC) No 1882/2006. 49. During the nine official controls observed, the audit team interviewed LA TSOs on sampling procedures as regards official samples for control of mycotoxins, heavy metals, nitrates and PAHs. In four cases of the nine, the LA TSOs correctly referred to relevant legislative provisions. In five cases of the nine, some deficiencies were observed: In three cases, the TSOs were not able to refer to and apply relevant legislative provisions but stated that advice would be sought from the respective public analyst; no autonomous action would be performed without that. In one case (sampling for mycotoxins), the TSO correctly identified the lot size but had difficulties in applying relevant legislative provisions laid down by Regulation (EC) No 401/2006, partly due to errors in the FSA Guide on Mycotoxins Sampling (incorrect size of aggregate samples for sampling on nuts). The TSO stated that the LA only has sampling equipment suitable for sampling of small lots but no sampling spears etc. for sampling, for example, bulk materials. In one case (sampling for mycotoxins from a store where sampling throughout the lot is not feasible), the TSOs in general correctly applied relevant legislative provisions laid down by Regulation (EC) No 401/2006 but did not verify that the quantity of the sampled part was at least 10% of the lot to be sampled. 2 In their response to the draft report the Competent Authority noted that an enforcement policy will detail in general terms the criteria for taking formal action (notices, prosecutions, seizures etc.). They are intended as public facing documents so that the public and businesses can understand the framework in which LAs operate and would initiate formal action they are unlikely to include specific directions with regards to sampling for this purpose. However, LAs in accordance with the Framework Agreement and the Food Law Code of Practice are required to have in place sampling procedures which would include direction to officers in the event of a non-compliant sample. 11

Conclusions on Official Sampling Procedure 50. The system of taking official samples is in place and adequate. 51. The knowledge of the LA TSOs of the requirements for representative sampling, procedures and techniques to apply is insufficiently robust to ensure representative sampling. This might weaken the validity of the official samples taken and compromise the analytical results. 5.2.7 Laboratory Performance Legal requirements Articles 4(2)(c), 11(1), 12(1) (2) and (3) and 33 of Regulation (EC) No 882/2004, Annex II to Regulation (EC) No 401/2006, part C-D of Annex to Regulation (EC) No 1882/2006 and part C-D of Annex to Regulation (EC) No 333/2007. Findings 52. Official samples of foodstuffs can be analysed by a number of OCLs, both public and private, designated in line with the requirements laid down by Article 12(1) of Regulation (EC) No 882/2004. The list of designated OCLs is available on the FSA's public website. 53. ISO 17025 accreditation audits, performed by United Kingdom Accreditation Service (UKAS), comprise verification of fulfilment of respective legislative provisions by the OCLs. The audit team was informed that a new agreement between the FSA and UKAS is currently being negotiated. 54. The audit team was presented with information on NRL (relevant to the scope of the audit) activities. All undertakings required by Article 33(2) of Regulation (EC) No 882/2004 were included in the activities presented by the NRLs and were assessed by the audit team as satisfactory. 55. The audit team was informed about one borderline sample (PAHs in smoked fish). The audit team noted that the interpretation of the analytical data was not in line with the requirements laid down by part D.2 of Annex to Regulation (EC) No 333/2007. 56. The audit team visited two OCLs and met representatives from a third OCL (undertaking nitrate monitoring). The OCLs visited have sufficient and trained technical staff. The premises and the equipment of the OCLs are fit for purpose. OCL 1 57. The first laboratory visited was a public OCL accredited according to ISO 17025. The audit team examined analyses of mycotoxins and heavy metals. 58. Sample preparation for mycotoxins analyses is almost exclusively done by wet homogenisation (slurry sample preparation). Ability of the process to achieve complete homogenisation was performed in cooperation with LGC (government reference laboratory) on peanuts only. The results are however extrapolated to other commodities. No verifying studies on the efficiency of the homogenisation, or for other matrices, have 12

been performed. This is not in line with the requirements laid down by part 2 of Annex II to Regulation (EC) No 401/2006. 59. Division into three replicate samples (enforcement, defence and reference) from the homogenised material is routinely performed. This is in compliance with requirements laid down by part 3 of Annex II to Regulation (EC) No 401/2006. 60. MU was originally estimated by the bottom up procedure but after identification of the major sources of variation it was re-assessed by combining bias and reproducibility. The calculations were presented for the audit team. 61. The audit team noted that the laboratory used a single material (peanuts) for calculation of recovery and this recovery is normally used for all matrices, except spices where individual recovery determinations are made in case of suspicion of non-compliant results. The applying of the recovery obtained from the peanuts sample to other matrices may lead to legally questionable, or even erroneous, interpretation of the analytical results. This was demonstrated in one proficiency test (PT) where the z-score was outside the normal range and the reason for this was use of the wrong recovery level (the one from the quality control peanuts sample). When the recovery was determined by spiking of the PT matrix, the z-score was within the normal range. 62. One non-compliant sample (aflatoxins) was vertically audited. The analytical report did not contain information on recovery or correction for it, which is not in line with the requirements laid down by part 4.4.1 of Annex II to Regulation (EC) No 401/2006. The audit team was informed that this information is available in the internal documentation but not always presented for the LAs. This raises a question on consistency of reporting of the final analytical results for official samples. The audit team was informed that the same deficiency was identified during a recent ISO 17025 accreditation audit and was promptly rectified directly after that ISO 17025 audit. 63. The audit team audited heavy metal analyses with a focus on cadmium. The method audited was sufficient for performing the analyses for the purposes of official controls. MU was estimated in the same matter as for mycotoxins. Recovery estimation was made by spiking samples in every analytical batch. The analytical reports presented were in compliance with the requirements laid down by D.1 of Annex to Regulation (EC) No 333/2007. 64. The laboratory regularly participates in relevant PTs organised by the respective European Union Reference Laboratory (EURL) and other providers. The PT results observed by the audit team were in general satisfactory. Non-satisfactory PT results are appropriately addressed. 65. The normal turnaround time is 3-4 working days. The formal turnaround time, agreed with the LAs, is 5-15 working days depending on the complexity of the matrix/analysis and the urgency of the investigation. OCL 2 66. The laboratory is a private OCL. Some analyses for the purpose of official controls are sub-contracted to other laboratories in the same holding. Since the other sites belong to the same holding, the only prerequisite for subcontracting the analyses is the presence of a valid ISO 17025 accreditation and no verification or supervision is undertaken. 13

67. The audit team audited aflatoxin analyses which are to a small extent carried out at the site visited but for which the majority of the official samples are analysed at another site (not visited during the audit). The ability of the sample preparation (slurry homogenisation) to achieve complete homogenisation was studied once on one material only (peanuts). The results of the peanuts study are however applied to other commodities. No studies for other matrices have been performed. This is not in line with the requirements laid down by part 2 of Annex II to Regulation (EC) No 401/2006. 68. Division into three replicate samples (enforcement, defence and reference) from the homogenised materials is not routinely performed. This is not in compliance with the requirements laid down by part 3 of Annex II to Regulation (EC) No 401/2006, part A.3.6 of Annex to Regulation (EC) No 1882/2006 and part C.2.4 of Annex to Regulation (EC) No 333/2007. 69. As regards estimation of MU, the laboratory stated that this is done from the measurement scope and previous experience and no more details were presented at the time of the audit. 70. As regards recoveries, those are calculated based on spiking tests performed once a week by spiking an aflatoxins-free material. The audit team can therefore presume that this recovery is normally used for all matrices. Applying the recovery obtained from a single sample to all matrices may lead to using an erroneous recovery level and thus to an inaccurate interpretation of the analytical results. 71. Vertical audits were carried out on two samples (aflatoxins and heavy metals), both analysed at the site visited. For the aflatoxins sample, data on MU and correction for recovery was missing on the report. This is not in line with the requirements laid down by part 4.4.1 of Annex II to Regulation (EC) No 401/2006. For the heavy metal analysis, these data were present on the report observed. This is in line with the requirements laid down by part D.1 of Annex to Regulation (EC) No 333/2007. 72. The audit team observed a number of analytical reports for various official analyses and noted that reporting of analytical results to LAs is not always in compliance with the requirements laid down by part 4.4.1 of Annex II to Regulation (EC) No 401/2006, part D.4 of Annex to Regulation (EC) No 1882/2006 and part D.1 of Annex to Regulation (EC) No 333/2007. 73. The laboratory regularly participates in relevant PTs organised by the respective EURL and other providers. The PT results observed by the audit team were in general satisfactory. Non-satisfactory PT results are appropriately addressed. 74. The normal turnaround time, as agreed with LAs, is 4-6 weeks. Laboratory undertaking nitrate monitoring 75. The FSA monitoring program on nitrates (in response to the requirement in Commission Regulation (EC) No 1881/2006) is subcontracted to a private company, which in turn subcontracts the analytical determinations to a private laboratory (ISO 17025 accredited). The audit team met representatives from both companies but did not visit the physical site of the laboratory. 14

76. The nitrates SOP applied for the official samples was presented by the laboratory representatives. The audit team noted that the SOP did not include division of the homogenised material into three replicate samples (enforcement, defence and reference); the division into the replicate samples is required by part A.3.6 of Annex to Regulation (EC) No 1882/2006. The laboratory representatives stated that such division is undertaken in practice (could not be verified by the audit team due to the meeting taking place outside the laboratory) and that such provision was incorporated into Application form for an evidence gathering project with the Food Standards Agency Research. This document is not a part of the laboratory's quality assurance system and thus not available for the staff dealing with the analyses. During the meeting, the audit team was not provided with any evidence that the relevant provisions of part A.3.6 of Annex to Regulation (EC) No 1882/2006 are followed. 77. The ability of the homogenisation procedure to achieve complete homogenisation is assessed through a quality control sample, which is not suitable to prove that complete homogenisation is achieved for other matrices with different physical and chemical characteristics. This is not in line with the requirements laid down by part C(3) of Annex to Regulation (EC) No 1882/2006. 78. MU and recovery were accessed in earlier years (recovery in 2008) but have not been verified since then. 79. Reporting of MU and recovery level is undertaken on an annual basis only but not for every sample as required by part D.4 of Annex to Regulation (EC) No 1882/2006 since it has not been requested by the FSA. 80. The laboratory regularly participates in PTs. The PT results observed by the audit team were satisfactory. Conclusions on Laboratory Performance 81. The network of OCLs has the competence and capacity to analyse official samples in the scope of the audit. 82. The NRL duties are adequately undertaken. 83. Investigative, traceability and other legal activities may be compromised by the absence of replicate samples in the OCLs. 84. Not all legislative requirements as regards assurance of appropriateness of sample preparation (homogenisation) and reporting of the analytical results are met due to deficient quality procedures. This reduces the reliability of the analytical results used for the purpose of official controls. 5.2.8 National Enforcement Measures Legal requirements Article 54 and Article 55 of Regulation (EC) No 882/2004. 15

Findings 85. The primary objective of enforcement action taken by LAs is to achieve compliance in the most effective way and the approach is in line with the hierarchy of enforcement. The FSA presented procedure on national enforcement measures. In accordance with it, LAs will ensure that any enforcement action taken is reasonable, proportionate, riskbased, consistent and in line with their enforcement policy. An initial informal approach, such as offering advice or communicating by letter to achieve compliance, is a wellestablished practice and is understood by food businesses. This approach is encouraged where such an approach is likely to secure compliance with the requirements of food law within a reasonable timeframe based on the circumstances. 86. When determining the appropriate enforcement action, consideration is given to: Risks to consumer safety resulting from the non-compliance; Particular consumer sensitivities around an issue, leading to loss of consumer confidence or economic loss to industry; The potential for non-compliant foods being distributed widely with large numbers of consumers affected; The previous history and confidence in the FBO to undertake the work informally. 87. National enforcement measures reflect the requirements laid down by Articles 54 and 55 of Regulation (EC) No 882/2004. Further details on the national enforcement measures are given in the Country Profile. 88. The audit team studied two cases of non-compliances and one case of a borderline sample (for the latter, see paragraph 55). All the cases were from 2015-2016. 89. For the first non-compliant sample studied (PAHs in smoked fish, routine informal sample), the audit team noted that the gap between sampling and receiving the report from the OCL was three months. Three weeks after the OCL report the LA informed the FBO on the outcome of the analysis but no intervention measures were taken. The audit team was not provided with a reasonable explanation of the delay in informing the FBO. The report of the first routine trading standards inspection after the incident contained no reference to measures undertaken either by the LA or by the FBO. The audit team noted that adequate intervention measures were partly inhibited by the three months turnaround time by the OCL and short shelf life of the product. The audit team was not provided with information if any intervention measure(-s) was taken in this case. 90. For the second non-compliant sample studied (nitrates in leafy green vegetables, routine informal sample), the gap between sampling and receiving the report from the OCL was two months. The LA immediately after receiving the OCL report visited the producer but no intervention measure(-s) on the lot sampled could be undertaken due to the short shelf life of the product. The audit team was not provided with information if any intervention measure(-s) was taken in this case. 91. The handling of the above-mentioned cases was partly inhibited by long turnaround times in the OCL. The audit team noted that in many cases long turnaround times 4-6 weeks are agreed between LAs and OCLs. This is not in accordance with the requirements laid down by Article 4(2)(c) of Regulation (EC) No 882/2004 in terms of access to an adequate laboratory capacity for testing. 16

Conclusions on National Enforcement Measures 92. There is an adequate system of enforcement measures in place. Implementation of the system is undermined by lack of procedures for non-compliant informal samples. Long reporting times by the OCLs limit the effective actions available to LAs. 5.3 RAPID ALERT SYSTEM FOR FOOD AND FEED Legal requirements Article 50 of Regulation (EC) No 178/2002 and Article 19(3) of Regulation (EC) No 882/2004. Findings 93. The system of handling RASFF notifications is described in the Country Profile. 94. Prior to the audit, the audit team studied six notifications available through the RASFF portal. The audit team concluded that all the cases were handled in line with the requirements laid down by Article 50 of Regulation (EC) No 178/2002 and Article 19(3) of Regulation (EC) No 882/2004. Conclusions on Rapid Alert System for Food and Feed 95. There is an adequate system in place for dealing with incoming and outgoing RASFF notifications. 6 OVERALL CONCLUSIONS In the UK, there is an adequate risk-based system of official controls as regards chemical contaminants in foodstuffs through the whole production chain. It allows for satisfactory planning and implementation of the controls, identifies non-compliances and enforces necessary and proportionate corrective actions. There is a sufficient network of OCLs. Nonetheless, this system shows some weaknesses that can (a) result in incomplete assessment of FBO's HACCP-based procedures, (b) call into question the reliability of official sampling and analysis, in particular in terms of representativeness of the samples taken, methods used for sample preparation, availability of replicate samples in the OCLs and interpretation of analytical results, and (c) delay undertaking of effective actions due to lengthy turnaround times by the OCLs. 7 CLOSING MEETING A closing meeting was held on 21 October 2016 with representatives of the FSA and the NRLs. At this meeting, the audit team presented the main findings and preliminary conclusions of the audit. The FSA provided the audit team with additional information/clarifications. 17