Denmark

Rekommendationen anger som grund

Official veterinarians use a standard template to carry out the epidemiological enquiries of outbreaks. The template follows the requirements of Article 57 of Regulation (EU) 2016/429. Most outbreak investigations concluded that wild birds were the source of infection. However, they did not consistently include detail on possible biosecurity breaches. For instance, in one case, minks that had got into the hen house had killed laying hens, but this was not further examined or linked to biosecurity breaches.

The DVFA did not document the tracing back and forward of all possible sources (people, vehicles or otherwise) that may have led to the introduction of the virus to the outbreak or to other possible epidemiological links of the outbreaks. The DVFA acknowledged the lack of records and expressed that there are plans to develop a system to record the steps and conclusions of the investigations of all possible epidemiological links to the cases. This was not in line with Article 57(2) and Article 62 of Regulation (EU) 2016/429. 13

In one outbreak, a private veterinarian, not delegated to take official samples, took swabs under the active surveillance program. These tested positive to H5N1. On the day the authorities got the notification from the laboratory, an official veterinarian took official samples, which tested negative. As this was a flock of ducks, where symptoms of H5N1 are usually limited or absent, the competent authority confirmed the outbreak based on the positive laboratory results of the samples taken by the private veterinarian. However, the authorities did not investigate further. They did not trace the private veterinarian's movements, check for cross contamination or carry out serological testing. This was not in line with Article 62 of Regulation (EU) 2016/429 and Article 17 of Delegated Regulation (EU) 2020/687.

The DK-VET carried out epidemiological analysis for each outbreak and each epidemic year. They include findings from the epidemiological investigations and laboratory results, including whole genome sequencing. These reports include information gathered on the outbreaks from the DVFA officials and explore links between viruses detected in poultry and wild birds.

Rekommendationen anger som grund

Private veterinarians take samples for active and sometimes also for passive surveillance. The DVFA has not delegated official activities, such as sampling within this active surveillance program, to private veterinarians (see also finding 66). This is not in line with Articles 28, 30 and 31 of Regulation (EU) 2017/625.

In one outbreak, a private veterinarian, not delegated to take official samples, took swabs under the active surveillance program. These tested positive to H5N1. On the day the authorities got the notification from the laboratory, an official veterinarian took official samples, which tested negative. As this was a flock of ducks, where symptoms of H5N1 are usually limited or absent, the competent authority confirmed the outbreak based on the positive laboratory results of the samples taken by the private veterinarian. However, the authorities did not investigate further. They did not trace the private veterinarian's movements, check for cross contamination or carry out serological testing. This was not in line with Article 62 of Regulation (EU) 2016/429 and Article 17 of Delegated Regulation (EU) 2020/687.

The DVFA consistently documented certain verification steps, such as movement derogation reviews and disinfection validations. At the time of the audit, the authorities did not have plans to establish regular biosecurity controls, enforce these proportionally, or delegate and supervise private veterinarians tasked with carrying out official activities. 16 4.2.2 Conclusions on control and eradication

Rekommendationen anger som grund

Regarding DAA, for many years the official controls on fruit and vegetables throughout Denmark have been carried out by the same team of two inspectors (30 official inspec- tions/year). The two inspectors informed the audit team that they had not received any specific training on official controls at primary production relating to microbiological risks.

Employees are introduced to relevant legislation and educated in practical skills and aca- demic subjects to the extent required. Some inspectors participate in other relevant courses outside DVFA such as Better Training for Safer Food (BTSF). However, there was no evidence of any cascade course activity in order to share experiences and knowledge from BTSF courses with other colleagues.

During official inspections observed by the audit team, both DAA and DVFA inspectors and their supervisors demonstrated clear deficiencies in identifying non-compliances. Only when the audit team highlighted evidence of several very critical deficiencies over- looked during official inspections (e.g. cross-contamination and staff misconduct) did inspectors and supervisors acknowledge that these deficiencies represent a real risk for food safety (see paragraphs 19, 20, 35, 48, 49, 52, 57, 65, 66 and 67). Audit / Supervision

However, to date, supervisors were unable to identify and guide inspectors in assessing key deficiencies related to microbiological risks, including those that could lead to direct contamination of food products. Furthermore, the audit team identified that non- conformities are evaluated differently by DVFA inspectors.

A training system has been implemented. However, to date, inspectors and supervisors are unable to assess key deficiencies related to microbiological risks, highlighting a lack of knowledge and experience which, combined with inconsistency in the implementation and enforcement of EU legislation, has a negative impact on the implementation of offi- 5 cial controls.

There is a supervisory system in place for official controls. However, the system fails to identify and correct all shortcomings in official controls. This highlights deficiencies in both the knowledge and experience of CAs and subsequent consistency and effective- ness of the implementation and enforcement of EU legislation.

Medlemsstatens svar

me time, it will be discussed with the inspectors what is a 30 April 2025 necessary level of hygiene and what is not acceptable in a given production. DVFA will give instructions to the official inspectors in order to make sure that sprout companies will test all seeds for

Rekommendationen anger som grund

The inspectors correctly assessed the certification requirements for mung beans imported from a specific third country as seeds for sprouting but neither the inspectors nor the su- pervisor identify that the FBO also sprout imported seeds from other third countries. Therefore, the requirements (e.g. health certificates) for other seeds imported from other third countries have never been verified meaning, DVFA has never fully enforced this requirement.

Although sprout-producing establishments are identified by DVFA as representing higher risk, the fact that verification of compliance with Regulation (EC) No 208/2013 on certi- fication requirements of imported seeds for sprouting has been overlooked during official controls, in conjunction with the fact that Regulation (EC) No 2073/2005 on preliminary tests for seeds and analyses of sprouts has been misinterpreted by the CAs, undermines the effectiveness of such official controls.

Medlemsstatens svar

sprouts regardless of the country of origin. All batches of 1 October 2024 seeds for sprouts must be tested. A will perform official sampling and testing of sprouts January 2025

Rekommendationen anger som grund

Regarding the FBO obligation to carry out preliminary tests on each individual batch of seeds for sprouting purchased, DVFA informed the audit team that, according to their of- ficial procedures, this obligation only concerns seeds from third countries. In the case of seeds intended for sprouting produced in the EU, the official procedures of DVFA pro- vide that inspectors can accept from FBOs a declaration provided by the supplier regard- ing the analytical results of the dried seeds supplied.

This is not in line with Regulation (EC) No 2073/2005 which establishes that FBO must test each batch of seeds for sprouting received, before starting the batch for production. 10 DVFA acknowledged this deficiency and confirmed they would start the process to amend their official guidelines for sprout-producing establishments.

Although sprout-producing establishments are identified by DVFA as representing higher risk, the fact that verification of compliance with Regulation (EC) No 208/2013 on certi- fication requirements of imported seeds for sprouting has been overlooked during official controls, in conjunction with the fact that Regulation (EC) No 2073/2005 on preliminary tests for seeds and analyses of sprouts has been misinterpreted by the CAs, undermines the effectiveness of such official controls.

Medlemsstatens svar

gation water in order to verify own check systems of the various companies. Page: 1 Ref. Ares(2024)4579070 - 25/06/2024 s of report ref. DG(SANTE) 2024-8040 of the audit carried out from 21 May 2024 lating to microbial safety of food of non-animal origin Action Proposed by the competent authority Proposed action plan deadline The DVFA conducts follow-up inspections whenever a non- compliance is detected. These follow-up inspections take place no later than two months following the inspection where the non-compliance has been detected. When exactly the follow-up inspections take place depends on the severity and amount of non-compliances detected. If the non-compliance regards a food-safety issue, the follow-up inspections will be prioritised and can take place as soon as er the same day or the following day. non-compliance does not pose a risk to food-safety, the follow up inspection will be conducted no later than two months following the inspection where the non-compliance

Rekommendationen anger som grund

During official inspections observed by the audit team, both DAA and DVFA inspectors and their supervisors demonstrated clear deficiencies in identifying non-compliances. Only when the audit team highlighted evidence of several very critical deficiencies over- looked during official inspections (e.g. cross-contamination and staff misconduct) did inspectors and supervisors acknowledge that these deficiencies represent a real risk for food safety (see paragraphs 19, 20, 35, 48, 49, 52, 57, 65, 66 and 67). Audit / Supervision

However, to date, supervisors were unable to identify and guide inspectors in assessing key deficiencies related to microbiological risks, including those that could lead to direct contamination of food products. Furthermore, the audit team identified that non- conformities are evaluated differently by DVFA inspectors.

A training system has been implemented. However, to date, inspectors and supervisors are unable to assess key deficiencies related to microbiological risks, highlighting a lack of knowledge and experience which, combined with inconsistency in the implementation and enforcement of EU legislation, has a negative impact on the implementation of offi- 5 cial controls.

There is a supervisory system in place for official controls. However, the system fails to identify and correct all shortcomings in official controls. This highlights deficiencies in both the knowledge and experience of CAs and subsequent consistency and effective- ness of the implementation and enforcement of EU legislation.

If premises, hygiene requirements and HACCP program are in compliance, the final approval will be granted. However, as noted by the audit team during the official inspec- tion, several long-term structural and maintenance shortcomings were observed (see paragraph 57), questioning the requirement of Article 148(5) of Regulation (EU) 2017/625, according to which CAs shall keep the approval of establishments under re- view when carrying out official controls. Importers that are importing and distributing risky produce

DVFA informed the audit team that official inspections are carried out in accordance with written procedures and instructions that include the requirement of Article 14 of Regulation (EU) 2017/625. However, there are no official written procedures regarding the classification, management and evaluation of non-conformities detected during offi- cial inspections and there are no instructions on the actions that inspectors must take to- wards the product when cases of direct contamination of foods are found during an offi- cial inspection.

In addition, in case of reported non-compliances, CAs do not require the FBO to provide an action plan.

Inspectors confirmed to the audit team that there are no specific procedures regarding classification/management of non-compliances including an absence of instructions on the actions to be taken when cases of potential direct contamination of food are found during an official inspection. 5.3.3 Implementation of official controls Legal requirements Articles 9(1)(d), 14 and 44(1) of Regulation (EU) 2017/625; Articles 4 to 8 and Annexes I and II of Regulation (EC) No 852/2004; Article 18 of Regulation (EC) No 178/2002; Article 3 of Regulation (EU) No 208/2013; Regulation (EU) 2020/2235; Article 1 and Annex I of Regulation (EC) No 2073/2005. Findings Primary production

Presence of birds inside greenhouses and food handling areas, inconsistent use of gloves by harvesting personnel, potential cross-contamination of harvested strawberries either by placing crates directly on the ground or by unhygienic worker practices, excessive presence of cobwebs in all packaging/storage areas of a primary FBO, were not identi- fied as non-compliance by either the inspectors or their supervisor.

There was no evidence that the above-mentioned non-compliances had ever been identi- fied and registered in previous inspections, including during the annual screening on "marketing standards" (see paragraph 38). In addition, the review of previous inspection reports detected very few non-compliances which gives the impression of compliance throughout the country, when this was not the case.

DVFA informed the audit team that their procedures provide inspectors with the power to require corrective action to be taken. However, during the audit, neither the FBOs, nor the inspectors, nor the supervisor took/requested immediate action to segregate potential- ly contaminated products observed by the audit team. Sprout-producing establishment

The inspectors correctly assessed the certification requirements for mung beans imported from a specific third country as seeds for sprouting but neither the inspectors nor the su- pervisor identify that the FBO also sprout imported seeds from other third countries. Therefore, the requirements (e.g. health certificates) for other seeds imported from other third countries have never been verified meaning, DVFA has never fully enforced this requirement.

However, other major non-compliances such as other areas of the establishment where condensation water was dropping on the food, a worker touching dirty surfaces and then product without changing gloves and rust/oil on two electric motors above the produc- tion line were overlooked by inspectors and their supervisor. In the last three years (six official inspections) there was no evidence that the inspectors or supervisor had detected or recorded similar non-compliances requiring correction.

During the debriefing with the FBO the inspectors, with the agreement of the supervisor, concluded that the official report will contain a minor non-conformity related to rusty pipes. Non-conformities regarding direct contamination of the product, even when de- tected by inspectors (condensation falling on the product), were not planned to be docu- mented in the official report. Processors (cut fruits and vegetables)

The audit team observed inspections being carried out at the high care areas (cutting, slicing, chopping area). In one case, even though several non-compliances were detected and recorded by the inspectors, condensation issues (even over the exposed foodstuff), rust and dirt on the ceiling and flaking paint on some equipment was overlooked.

When guided by the audit team, the inspectors identified inadequate cleaning of machin- ery, which the FBO considered as clean and ready to be used. However, there was no ev- idence (official reports) that similar non-compliances had ever been identified in previ- ous inspections as in other processors of FNAO.

The inspectors and the supervisor confirmed to the audit team that, during inspection, non-compliances, like the one observed by the audit team during these official inspec- tions, are usually communicated orally to the FBOs rather than documented in the offi- cial reports.

CAs did not keep the approval of sprout-producing establishments under review when carrying out official controls, contrary to the requirements of Article 148(5) of Regulation (EU) 2017/625. 12

Regarding farms, sprout-production establishments and processors, non-compliances are rarely detected and even when detected their severity is underestimated and rarely record- ed. In addition, in case of on-the-spot detected risks, no immediate action is taken on the affected product. This calls into question both the ability of the inspectors to evaluate the non-conformities in a consistent manner and the effectiveness of their supervision. Con- sequently, the deficiencies in the enforcement measures undermines the effectiveness of the system of official controls.

Although sprout-producing establishments are identified by DVFA as representing higher risk, the fact that verification of compliance with Regulation (EC) No 208/2013 on certi- fication requirements of imported seeds for sprouting has been overlooked during official controls, in conjunction with the fact that Regulation (EC) No 2073/2005 on preliminary tests for seeds and analyses of sprouts has been misinterpreted by the CAs, undermines the effectiveness of such official controls.

Medlemsstatens svar

tected. The food businesses are obliged to comply with the rules and in case a non-compliance is not remedied, the DVFA will impose a more severe sanction in order to make the business comply. If necessary, the DVFA can also choose to intensify the number of inspections for a limited amount of time (e.g. 5-6 inspections within 3 months). Action plan: The DVFA has taken the recommendation into consideration. The DVFA is currently revising the DVFA general inspection manual and will in that process emphasize the importance of Page: 2 Ref. Ares(2024)4579070 - 25/06/2024 s of report ref. DG(SANTE) 2024-8040 of the audit carried out from 21 May 2024 lating to microbial safety of food of non-animal origin Action Proposed by the competent authority Proposed action plan deadline conducting follow-up inspections as quickly as possible in case of non-compliances with a potential risk to food safety. Furthermore, the process of evaluating the DVFA general inspection manual will focus on raising awareness among the DVFA inspection personnel to carefully consider using administrative fines when coming across serious non- compliances during inspections. Furthermore, the DVFA will ensure that the

Rekommendationen anger som grund

During official inspections observed by the audit team, both DAA and DVFA inspectors and their supervisors demonstrated clear deficiencies in identifying non-compliances. Only when the audit team highlighted evidence of several very critical deficiencies over- looked during official inspections (e.g. cross-contamination and staff misconduct) did inspectors and supervisors acknowledge that these deficiencies represent a real risk for food safety (see paragraphs 19, 20, 35, 48, 49, 52, 57, 65, 66 and 67). Audit / Supervision

However, to date, supervisors were unable to identify and guide inspectors in assessing key deficiencies related to microbiological risks, including those that could lead to direct contamination of food products. Furthermore, the audit team identified that non- conformities are evaluated differently by DVFA inspectors.

There is a supervisory system in place for official controls. However, the system fails to identify and correct all shortcomings in official controls. This highlights deficiencies in both the knowledge and experience of CAs and subsequent consistency and effective- ness of the implementation and enforcement of EU legislation.

DVFA informed the audit team that their procedures provide inspectors with the power to require corrective action to be taken. However, during the audit, neither the FBOs, nor the inspectors, nor the supervisor took/requested immediate action to segregate potential- ly contaminated products observed by the audit team. Sprout-producing establishment

The inspectors and the supervisor confirmed to the audit team that, during inspection, non-compliances, like the one observed by the audit team during these official inspec- tions, are usually communicated orally to the FBOs rather than documented in the offi- cial reports.

CAs did not keep the approval of sprout-producing establishments under review when carrying out official controls, contrary to the requirements of Article 148(5) of Regulation (EU) 2017/625. 12

Regarding farms, sprout-production establishments and processors, non-compliances are rarely detected and even when detected their severity is underestimated and rarely record- ed. In addition, in case of on-the-spot detected risks, no immediate action is taken on the affected product. This calls into question both the ability of the inspectors to evaluate the non-conformities in a consistent manner and the effectiveness of their supervision. Con- sequently, the deficiencies in the enforcement measures undermines the effectiveness of the system of official controls.

Although sprout-producing establishments are identified by DVFA as representing higher risk, the fact that verification of compliance with Regulation (EC) No 208/2013 on certi- fication requirements of imported seeds for sprouting has been overlooked during official controls, in conjunction with the fact that Regulation (EC) No 2073/2005 on preliminary tests for seeds and analyses of sprouts has been misinterpreted by the CAs, undermines the effectiveness of such official controls.

Medlemsstatens svar

end. Furthermore, the inspection units have a thorough training all inspection personnel must complete in their first few months of appointment to the DVFA. OCR and Danish legislation: All DVFA staff, including staff conducting inspections, are instructed to comply with OCR, the Danish Public Administration Act as well as the DVFA sanctioning practice Page: 3 s of report ref. DG(SANTE) 2024-8040 of the audit carried out from 21 May 2024 lating to microbial safety of food of non-animal origin Proposed action plan deadline undation for k. The law ith certain remote) in a rinciple is in ely. ualified from at inspector in question. To pectors don't s in a row. ate with the apply FA deem the A to apply a etected, if he non- 024)4579070 - 25/06/2024 s of report ref. DG(SANTE) 2024-8040 of the audit carried out from 21 May 2024 lating to microbial safety of food of non-animal origin Proposed action plan deadline is principle be considered of non- with a formal e non- e batch of e cleaning re is what kind of public? there mitigate og ajeure n history into n- s of report ref. DG(SANTE) 2024-8040 of the audit carried out from 21 May 2024 lating to microbial safety of food of non-animal origi

Rekommendationen anger som grund

In two compound feed manufacturers visited, the audit team reviewed the most recent inspection reports and interviewed the inspectors on their assessment of the operators' HACCP plans. In one establishment, the hazard identification section of the operators' HACCP plan and the HACCP principles on the critical control point were correct. The audit team could identify some minor shortcomings, and during the audit interview the feed inspector correctly assessed these shortcomings. In the other establishment, the hazard identification section of the operators' HACCP plan was very generic. Several hazards were grouped and assessed together, although the severity and the probability of these hazards were very different (for example several different types of cereals and relevant hazards were assessed as a single group). The operator's monitoring procedure described in its own HACCP system did not reflect the procedure implemented by the operator. The above shortcomings in the implementation of Article 6(2) of Regulation (EC) 183/2005 identified by the audit team were overlooked during previous official inspections.

In the two food establishments manufacturing and suppling a by-product to the feed chain, the feed part of operators' HACCP systems were deficient. HACCP documents did not consider any production steps relating to feed production or the procedures for the collection and storage of retained samples. These are not in line with the requirements of Article 7(1) and point 4 of the section "Quality control" of Annex II to Regulation (EC) No 183/2005. HACCP documents did not contain descriptions of the requirements for labelling of feed materials manufactured and several of these requirements were not communicated to the buyers (feed material, compulsory declarations) which is not in line with the requirements of Articles 15 and 16 of Regulations (EC) 767/2009. These deficiencies were not identified in previous inspection reports.

While the planning of inspections is based on appropriate risk-related factors and official inspections are carried out on all operators on a risk basis and with an appropriate frequency, the effectiveness of these controls is weakened by some shortcomings in the inspectors assessments of (i) feed business operators' hazard analyses and critical control point plans, and (ii) operators' tests on minimisation of cross contamination and on production of suitably homogeneous mixture of feed. 5.2.2 Sampling and analysis 5.2.2.1 Planning of official sampling

Rekommendationen anger som grund

Verification of establishments' tests on homogeneity of mixing (required by point 3 of section "Facilities and equipment" of Annex II to Regulation (EC) No 183/2005) is included in the scope of the inspections, when applicable. In three compound feed manufacturers visited, the audit team reviewed the operators' documentation on the related homogeneity testing and interviewed the inspectors who usually control those establishments. In one establishment visited, several results of mixing homogeneity tests were unacceptable in the previous 9 months, but the FeBO did not implement any corrective actions. In another establishment, the number of the samples collected was lower than necessary to implement an adequate test. The feed inspectors met indicated 7 that they had experienced difficulties in comprehensively assessing the procedures and the results of the mixing homogeneity tests.

Verification of establishments' tests to assess the measures taken to minimise cross- contamination, required by point 2 of section "Facilities and equipment" and point 3 of section "Production" of Annex II to Regulation (EC) No 183/2005, forms part of the inspections, when applicable. In one establishment visited, the analytical results of the testing laboratory were higher than the maximum limit described in Section VII, Annex I to Directive 2002/32 in the last 3 years. However, the FeBO did not implement any corrective actions. In two establishments FeBOs collected only one sample from the first mix after cleaning (flushing) for testing of cross-contamination. In one establishment, several samples were collected from the first mix after flushing, however these samples were pooled and sent to the laboratory as one sample. Similarly, as for verification of operators test on homogeneity of mixing, the feed inspectors met indicated that they had experienced difficulties in comprehensively assessing the procedures and the results of the cross-contamination tests.

While the planning of inspections is based on appropriate risk-related factors and official inspections are carried out on all operators on a risk basis and with an appropriate frequency, the effectiveness of these controls is weakened by some shortcomings in the inspectors assessments of (i) feed business operators' hazard analyses and critical control point plans, and (ii) operators' tests on minimisation of cross contamination and on production of suitably homogeneous mixture of feed. 5.2.2 Sampling and analysis 5.2.2.1 Planning of official sampling

Rekommendationen anger som grund

The audit team reviewed the records on seven sampling and one inspection-related non-compliances. The feed inspectors could provide the official reports on the follow- up of these non-compliances. In three cases, these reports provided sufficient information indicating that the FeBOs had carried out investigations to address the deficiency, and/or identify the root cause, and/or prevent the reoccurrence of the identified non-compliances. This is in line with the requirements of Article 138 of Regulation (EU) 2017/625. However, in four of the sampling-related cases, the competent authority could not provide evidence that the FeBOs carried out adequate investigations to identify the root cause and took actions to prevent the reoccurrence of the identified non-compliances (e.g. evidence on the destruction of contaminated feeds, correction of recipes/labels, modification of HACCP procedures) which is not in line with the above-mentioned legislative requirements.

The inspection related non-compliance arose from a problem with the operator's traceability system. The feed inspector identified this problem in July 2024 and issued a warning to the operator. The feed inspector carried out an on-the-spot follow up inspection at the end of September 2024. The problem still remained during the follow-up inspection. The feed inspector prepared the second warning to the operator and took steps to inform the police. At the time of the audit the second warning was still under legislative check by the relevant unit of the DVFA. The competent authority could not provide evidence that the FeBO took actions to correct the traceability problem and carried out an investigation to prevent the reoccurrence of the problem in its traceability system. The competent authority did not impose any other sanctions (e.g. restrict and/or prohibit the placing on the market of products) relating to the untraceable products. The approach taken by the competent authority is not in line with the requirements of Article 138 of Regulation (EU) 2017/625.

Feed inspectors stated that they do not have the right to request action plans with deadlines from FeBOs relating to identified non-compliances. This is not in line with the requirements of Article 13 (1)(d) of Regulation (EU) 2017/625. Conclusion on the actions taken on non-compliances

Notwithstanding the fact that the competent authority took actions when non- compliances were identified in the course of official controls, the competent authority did not ensure the quality and consistency of official controls in relation to (i) the determination of the origin and extent of the non-compliances, (ii) operator remediation of the non-compliance, and (iii) prevention of further occurrences. Collectively these shortcomings weaken the effectiveness of the system of official controls on feed. 13

Rekommendationen anger som grund

As noted in section 5.1.4 above, there are no documented procedures for performing of- ficial checks on consignments of animals entering the EU. A local work instruction has been issued by the BCP Copenhagen airport, which does provide guidance for perform- ing such checks, quarantine and where necessary, euthanasia of consignments of live fish. The audit team noted based on their evaluation of the CHEDs for two consignments of live fish that had been rejected, that the fish were euthanised by the operators respon- sible for the consignments, in both cases using methods of euthanasia that were not in line with the local work instruction at the BCP, and not under supervision by the DVFA. 5.2.4.2 Checks on consignments of goods of animal origin

The audit team noted however, that in cases where non-compliances with the rules in Ar- ticle 1(2) of the OCR were identified, the DVFA had not systematically finalised the CHED, as soon as the official checks including any sampling, had been completed, and a decision that the consignment was not in compliance had been made, which is not in line with Articles 55(1) or 56(5) of the OCR. The audit team also noted that the consign- ments were not systematically placed under official detention as required by Article 66 (1) of the OCR.

Nine of the CHEDs selected and evaluated by the audit team concerned non-compliant consignments. The team noted that the lack of official detention had enabled four of the consignments concerned to be exported or destroyed by the operator without the knowledge or supervision of the DVFA, and prior to the decision being recorded in the CHED.

The DVFA informed the audit team that the operator responsible for a non-compliant consignment is invited to specify which of the measures provided in Article 66(3) of the 11 OCR, would be taken. The DVFA confirmed that, commonly, no deadline for a response is specified for the response and therefore non-compliant goods may remain at the BCP for extended periods of time, which is not in line with the same Article of the OCR. Conclusions on the implementation of official controls on animals and goods entering the EU

However, the significant shortcomings in the decision-making process and follow-up of measures to be taken in the case of non-compliant consignments of animals and goods have the potential to undermine the overall effectiveness and appropriateness of official controls to ensure that only compliant consignments of animals and goods enter the EU.

Rekommendationen anger som grund

The DVFA has issued documented procedures in the form of a manual, for the perfor- mance of official controls on goods of animal origin entering the EU. The DVFA in- formed the audit team that there are no documented procedures for the official controls on animals entering the EU, other than a locally issued work instruction at the BCP Co- penhagen airport, and no specific documented procedures for the implementation of measures in case of suspicion of non-compliance, confirmed non-compliance or risk, re- ferred to in Articles 65, 66 and 67 of the OCR.

The audit team noted therefore, that the documented procedures do not include all of the subject areas set out in Chapter II of Annex II to the OCR, which is not in line with Arti- cle 12(1) of that regulation. The audit team found related shortcomings in the implemen- tation of measures in case of non-compliance, as detailed in section 5.2.5 (findings 36 and 44 to 46) below.

The audit team noted that the outcome of the control verifications may be discussed within experience exchange groups held between BCPs and the central International Trade Division which is responsible for documented procedures. However, they are not systematically shared with that Division, which thereby risks compromising the review of documented procedures required by Article 12(3) of the OCR.

However, shortcomings in the documented procedures and in the access to information necessary for the competent authority to confirm that consignments subject to mandatory control at BCPs had been correctly notified undermine the overall effectiveness of the organisation of the official controls. 5.2 IMPLEMENTATION OF OFFICIAL CONTROLS ON ANIMALS AND GOODS ENTERING THE EU The audit team evaluated the implementation of official controls on consignments of animals and goods entering the EU through BCPs in Denmark against the applicable EU requirements detailed in Annex 2 to this report, based on an initial selection by the audit team and competent authorities of 39 valid CHEDs recorded in TRACES-NT, and further evidence obtained during the audit, including the on-the-spot visits to three BCPs and two control points. 5.2.1 Prior notification of arrival of consignments

Rekommendationen anger som grund

In the case of the inspection centre Samskip A/S at the port BCP Aarhus, the audit team noted a number of significant shortcomings with respect to the minimum requirements laid down in Article 64(3) of the OCR and Articles 3 and 4 of Implementing Regulation (EU) 2019/1014, in particular: · the unloading area was not covered by a roof; · there was no dedicated storage area or room; · the inspection room was not fitted with walls, floors and ceilings that are easy to clean and disinfect, adequate drainage nor adequate natural or artificial light; · the inspection room was also not equipped with a table with smooth washable sur- face easy to clean and disinfect.

The audit team noted that the inspection centre had been subject of an annual check by the Food Inspection Unit prior to the amendment of the checklist, which had identified similar shortcomings concerning the inspection room. Corrective action had been re- quired and implemented by the inspection centre, but as noted above, had not ensured compliance with the minimum requirements. In response to the audit findings, the DVFA withdrew the listing of the inspection centre shortly after completion of the audit.

The audit team visited two control points designated by the competent authority for the performance of identity and physical checks on consignments of non-animal origin and noted that neither control points were compliant with the minimum requirements for BCPs, which is not in line with Articles 53(1)(a) of the OCR. In particular: · at one control point, part of the storage area is also used for the inspection of goods and did not have a supply of hot and cold running water and facilities for washing and drying hands which is not in line with Article 3(1)(b) of Implementing Regula- tion (EU) 2019/1014. Neither did it have a table for performing inspections (these 13 were carried out on a pile of wooden pallets), nor cleaning and disinfection equip- ment. · at the second control point, the audit team noted that the designated storage area was not adequate for the number of consignments received, and that consignments were stored outside of the designated storage area. The inspection room itself is also used for other activities, including the storing of administrative documents, and was not easy to disinfect or clean, which is not in line with Articles 3(2) and 4(1) of Imple- menting Regulation (EU) 2019/1014.

The audit team noted that the DVFA checks were carried out prior to the amendment of the checklists and had not therefore identified the above issues. Conclusions on the compliance of border control posts with minimum requirements

Despite those measures, the audit identified shortcomings in one inspection centre at one BCP, and at both control points evaluated. The competent authority has strengthened their verification checks and took action during the audit to delist the non-compliant inspection centre.

Rekommendationen anger som grund

BACKGROUND 4.1. Legal framework for controls on residues in animals and food of animal origin Member States are legally obliged to implement control plans for residues of pharmacologically active substances in animals and animal products (2). These comprise three plans: a national risk-based control plan for production in the Member States, a national randomised surveillance plan for production in the Member States and a national risk-based control plan for third-country imports into the EU. The plans are subject to annual Commission evaluation (3). Member States must establish multiannual national control programmes for pesticides residues (4) in food and feed of both animal and non-animal origin and they must also participate in the multiannual Union control programme (5). Member States must control the presence of certain contaminants in food of both animal and non-animal origin. The foodstuffs of animal origin to be tested and the categories of contaminants to be tested are specified (6) as is the sampling frequency (7). This control plan is also subject to annual Commission evaluation (8). The scope of this report is limited to the competent authorities' implementation of control plans for residues of pharmacologically active substances, pesticides and contaminants in animals and animal products. It does not cover the controls on residues in food and feed of non-animal origin.

The control and surveillance plans for residues of pharmacologically active sub- stances, pesticides and contaminants are generally implemented in line with planned arrangements, EU legal requirements and are effective, underpinning assurances on the residues status of food of animal origin. Notwithstanding this positive assess- ment, the practice of randomly sampling animals in slaughterhouses for the risk- based national control plan for residues of pharmacologically active substances, un- dermines the effectiveness of this particular control programme.

Medlemsstatens svar

The Danish Veterinary and Food Administration is in the process of developing a guide on targeted sampling to the official samplers at the slaughterhouses. It is the intention to have the uideline ready for 2025.

Rekommendationen anger som grund

With regard to the testing of follow-up samples taken by DVFA officials in cases where it was already confirmed that the illegal content of malachite green in aqua- culture was due to environmental contamination, the audit team was informed that the operator had the right to choose any laboratory for this testing, regardless of whether it had been designated or not by the DVFA. The use of undesignated labora- tories for official follow-up investigation for malachite green is not in line with EU rules (45). Conclusions on follow-up of non-compliant results

For the most part the actions taken in case of non-compliant results are effective and underpin assurances on the chemical safety of food of animal origin. Nevertheless the practice of permitting operators to select laboratories (which have not been designat- ed for conducting official controls) for the testing of official samples taken in the context of official follow-up investigation for malachite green, undermines the effec- tiveness of the follow-up process. 5.7. Laboratories, methods of analysis, method validation and internal quality control

The audit team noted that, while no formal designation was needed for the DVFA laboratory, and a co-operation agreement was in place between the DVFA and the (national) subcontracted laboratory, the co-operation agreement was made between the (national) subcontracted laboratory and the foreign sub-contracted laboratory, and thus the foreign laboratory was not appropriately designated, which is not in line with pertinent EU requirements (48).

The DVFA did not audit both national laboratories since they were subject to audits by the Danish national accreditation body every 15 months, which is in line with EU legislation49. The arrangements in place do not ensure adequate co-ordination and collaboration with the foreign laboratory, as there is no direct relationship between the competent authority and the foreign laboratory since the co-operation agreement is not directly between them, which is not in line with EU legislation (50).

Notwithstanding the fact that all of the laboratories testing samples for residues of pharmacologically active substances, pesticides and contaminants taken under the various control and surveillance plans are accredited to EN/ISO/IEC 17025, operate appropriately validated methods and have well-functioning quality control systems in place, the use of non-accredited methods for pesticides in certain commodities and questions on the designation of one of the laboratories involved, weakens the effec- (52) Article 7 of Regulation (EU) 2021/808 (53) Regulation (EU) 2021/808 (54) Chapter 3 of Annex I to Regulation (EU) 2021/808 (55) Article 4 of Regulation (EU) 2021/808 and Article 38(2) of Regulation (EU) 2017/625 10 tiveness of the residues control system.

Medlemsstatens svar

The practice mentioned in conclusion 37 of permitting operators to select laboratories has been used only in cases with illegal findings of malachite green in aquaculture, where it has s been confirmed that the content of malachite green was due to environmental contamination and not illegal treatment. However, the Danish Veterinary and Food Administration acknowledges that this practice is problematic and therefore, it has been effected immediately, at in future cases these samples shall be analysed at the Danish Veterinary and Food dministration laboratory in Ringsted. The Danish Veterinary and Food Administration will start the designation of the foreign laboratory used for verification of a minor number of substances as soon as possible.

Rekommendationen anger som grund

With respect to pesticide analysis, the DVFA laboratory the Liquid Chromatography ­ (tandem) Mass Spectrometry (LC-MS/MS) method for 207 pesticides in honey is accredited as is the LC-MS/MS method for three pesticides in eggs. Neither the LC- MS/MS methods for 62 pesticides tested in meat, milk / milk products, eggs / egg products, aquaculture / products of aquaculture nor that for 120 pesticides tested in imported fish products is accredited. This is not in line with EU legislation (47).

Notwithstanding the fact that all of the laboratories testing samples for residues of pharmacologically active substances, pesticides and contaminants taken under the various control and surveillance plans are accredited to EN/ISO/IEC 17025, operate appropriately validated methods and have well-functioning quality control systems in place, the use of non-accredited methods for pesticides in certain commodities and questions on the designation of one of the laboratories involved, weakens the effec- (52) Article 7 of Regulation (EU) 2021/808 (53) Regulation (EU) 2021/808 (54) Chapter 3 of Annex I to Regulation (EU) 2021/808 (55) Article 4 of Regulation (EU) 2021/808 and Article 38(2) of Regulation (EU) 2017/625 10 tiveness of the residues control system.

Medlemsstatens svar

Analysis of residues of pharmacologically active substances and pesticides are within the scope of accreditation. However, a few methods are not in the scope of accreditation. These ter methods are planned to be validated and accredited during 2024-2026. The quality control for methods outside accreditation is the same as for methods within accreditation. Please see Attachment I for more details. Page: 1 enmark to the recommendations of report ref. DG(SANTE) 2024-8080 of the audit carried out from 05 March controls on residues of pharmacologically active substances, pesticides and contaminants in animals and animal ction Proposed by the competent authority and deadline by which it will be completed Page: 2

Rekommendationen anger som grund

The DAA inspectors provide a summary written record of all official controls to the operator as required by Article 13(1) of Regulation (EU) 2017/625. This is usually in paper format following completion of the inspection report upon returning to the office.

Deficiencies in the VAKS system for recording official controls is resulting in some inefficiencies in work practices.

Medlemsstatens svar

The DAA will take steps aiming to further digitize inspection notes and documentation and will investigate the possibilities for 5. developing an interface, making the IT-system VAKS for the recording of controls compatible with tablets and enabling the inspectors to put the information directly in VAKS using their hand-held device.

Rekommendationen anger som grund

The competent authority has a contingency plan on how to contain and manage outbreaks of listed or emerging Category A diseases in aquaculture animals. The contingency plan is available on the intranet of the DVFA. However, the last version of the contingency plan is dated 2010 and the DVFA is planning to update it in 2024, together with a new format for the epidemiological inquiry that the Team Aqua has to use to investigate aquatic diseases.

The competent authorities need to update the contingency plan, together with the epidemiological inquiry to be used for the investigation of aquatic diseases, taking into account the new provisions set by the AHL. This would allow the competent authorities to act, in line with the EU legislation, if a Category A disease is occurring in the country.

Medlemsstatens svar

al Being prepared is an important precautionary principle to enable a rapid and effective iseases, response to any outbreak of a notifiable disease. Due to the resent development of EU) food and veterinary crises in Europe, the Danish Veterinary and Food Administration (DVFA) has an increased focus on contingency planning throughout the organization. As a result, a contingency policy and a general contingency plan have been developed for the entire organization. The DVFA is currently in the process of updating and developing new contingency plans for category A diseases. The Danish contingency plan for category A diseases consists of a general plan ("partial plan" in attachment 1) and relevant action plans in relation to e.g. culling, disposal, cleaning and disinfection, and supplemented by disease specific action cards and templates (see attachment 1). The Veterinary partial plan regarding category A diseases includes category A diseases of aquatic animals. The plan is based on AHL and is expected to be completed in the 3rd quarter of 2024.

Rekommendationen anger som grund

The audit team visited one trout establishment approved and ranked as medium risk by the Team Aqua, since aquatic animals are sent to put and take lakes only. However, the competent authority didn't consider the high risk of infection due to the close proximity to an angling lake next to it, the recurrence of the IHN infection with two outbreaks occurred respectively in 2021 and 2022 and the high number of movements in and out the establishment. This is not in line with Chapter 2 a) and b), Part I of Annex VI to Regulation (EU) 2020/689, as regards the risk ranking to be applied in certain approved aquaculture establishments in this regard.

The audit team visited another trout establishment approved and ranked as medium risk, which has been infected with IHN in September 2022. Fish is kept for approximately three months, from 200 to 500 grams. The operator is consigning to more than 50 establishments, of which approximately 70 % are put and take lakes, and it is receiving from other 20 establishments. Nevertheless, the competent authority didn't consider the high risk of infection due to the high number of movements going in and out the establishment. This is not in line with Regulation (EU) 2020/689, as mentioned in the previous finding. The competent authority agreed with the audit team on the need to review the risk assessment for ranking approved establishments, together with the DTU Aqua.

The competent authority have implemented effectively the new process of registration and approval of the aquaculture establishments, according to the AHL, with clear guidelines and instructions for the correct implementation from the VIU. However, in some cases the Team Aqua has assessed the risk of the approved establishments not taking into account all the criteria listed in Annex VI to Delegated Regulation (EU) 2020/689. This may increase the risk of not detecting the IHN during the regular health visits carried out by the Team Aqua.

The official veterinarians of the Team Aqua are implementing, in particular, the risk-based health surveillance in all approved aquaculture establishments, as foreseen by Chapter 1, Part III, Annex VI of Regulation (EU) 2020/689. The frequency of the health visits and possible sampling depends on the risk of the approved aquaculture establishment, determined by the DVFA as high, medium or low. However, the DVFA ranked some of the establishments visited by the audit team, as explained in previous findings 36 and 37, with a risk ranking (i.e. "medium" instead of "high) which is resulting in a frequency of health visits too low.

Medlemsstatens svar

Between November 2023 and March 2024, the DVFA and DTU-Aqua have on developed a new model for the risk assessment of approved aquaculture establishments (see attachment 2). The risk ranking is done for each category C ion disease, and summarised in an overall risk for the assessed establishment. All risk factors listed in EU legislation has been considered in connection with the risk ranking for each disease, and if considered relevant they are included in the risk ranking. 38 The proposed action is therefore already completed. In connection with the next risk-based health visits, the VIU will reassess the risk ranking of each approved aquaculture establishment in accordance with the new model. The result of the risk ranking will be included in the control report for each control visit. Page: 1 ort, received 10 March 2024 e recommendations of report ref. DG(SANTE) 2023-7760 of the audit carried out from 11 September mal health controls on live aquatic animals Action Proposed by the competent authority and deadline by which it will be completed

Rekommendationen anger som grund

The competent authorities in Finland notified an IHN outbreak in June 2022, following a consignment of aquatic animals received from Denmark. The DVFA traced and tested, with negative results, the two establishments delivering respectively eggs and fishes to the confirmed outbreak in Finland, explaining to the audit team that a possible transmission of the disease was 24 due to the transit through one of the two water exchange points registered in Denmark. As a consequence, the DVFA established a specific procedure for the cleaning and disinfection of one of the two water exchange points, with all transporters requested to transit that specific water exchange point when delivering fish for intra-EU trade to other free IHN compartments. However, no specific procedures for the cleaning and disinfection have been approved for the other water exchange point registered in Denmark. 5.5.3 Certification

The competent authorities implemented the certification for the trade of aquatic animals from Denmark to other Member States as required by EU legislation. However, the competent authority deemed transportation of aquatic animals, transiting through a water exchange point in Denmark without proper disinfection procedures, as the possible cause of the transmission of IHN to another Member State This is not in line with the EU legislation, that requires that the competent authorities must certify that animals are not transported under conditions that jeopardize their health status.

Medlemsstatens svar

As described in finding 107, only two water exchange points are registered in ensure Denmark, and the DVFA has approved the cleaning and disinfection procedures of one of the two water exchange points. The DVFA has informed the operator, as well as the transporters and aquaculture industry organisation, that the other water exchange point cannot be used in connection with any transport of fish before the cleaning and disinfection procedures of this water exchange point have been approved by the DVFA. The proposed action is therefore already completed.

Rekommendationen anger som grund

The VIUs are carrying out spot-checks of transporters according to Regulation (EU) 1/2005, for animal welfare. However, they are not carrying out the official controls of the transporters registered in the Country. This is not in line with the provisions of paragraph 1 of Articles 9, 10 and 20 of Regulation (EU) 2017/625 (Official Controls Regulation - OCR), regarding the regular risk based official controls on all operators, including transporters, at any stage of production, processing and distribution of animals, products of animal origin and germinal products. Those official controls would need to address, as well, the activities under the control of operators, i.e., the general obligation on operators as regards biosecurity requirements for the transport of aquatic animals, foreseen by Article 3 of Regulation (EU) 2020/990.

The competent authorities have approved the procedures for the cleaning and disinfection of the transporters, one water exchange point, and for all the other documents foreseen by the EU legislation. However, they are not applying regular official controls on transporters and water exchange points, on a risk basis, to verify compliance with the approved procedures and the activities under the control of operators. This is not in line with the EU legislation, and it is not assuring the absence of risk of transmission of IHN or other aquatic diseases with the 25 movement of aquatic animals.

Medlemsstatens svar

s on In accordance with article 269 of the AHL, the DVFA is currently developing a ge national order and thereby adopting national measures. This order will include the legislative basis for approving water exchange point, and specify the regular risk based official control of approved water exchange point. This order is expected to enter into force on 1st January 2025. In 2025, the DVFA will carry out a prioritised action on transporters, which will include inspection of the transporter's compliance with the general obligation on operators as regards biosecurity requirements for the transport of aquatic animals, foreseen by Article 3 of Regulation (EU) 2020/990. Prioritised action is a specific type of in-depth inspection in the Danish veterinary inspection system. Page: 2

Rekommendationen anger som grund

The audit team obtained evidence of past training courses (since 2019 till 2022, included) provided to the DVFA staff carrying out official controls and noted that, except for rules on handling SRM, none of the training courses provided covered ABP requirements applicable to operators acting along the ABP chain. Moreover, all of the DVFA inspectors met stated that have not ever received a specific ABP training, except for rules on SRM, and they had to self-educate on the ABP rules, what was in their opinion challenging due to complexity of the ABP legislation. The lack of appropriate training on ABP, to enable staff performing official controls to undertake their duties competently and to preform official controls in a consistent manner, is not in line with the requirements of Article 5(4) of Regulation (EU) 2017/625.

DVFA did not provide, in the last four years, the ABP specific training to the staff verifying ABP requirements along the ABP chain. Moreover, existing operating control procedures, manuals and guides do not provide sufficient instructions to staff to enable them to competently and consistently carry out their inspection duties on operators acting along the chain of ABP and DP. The lack of appropriate support, in terms of training and documented procedures, to the staff performing official controls along the ABP chain undermines the overall quality and effectiveness of these controls.

Official controls in slaughterhouses and meat cutting plants cover ABP requirements with very narrow focus, often omitting or misinterpreting basic requirements for categorisation of ABP (which in turn determines their further destination), their use and disposal. When comparing inspection results recorded in the inspection reports and the operators' routine daily practices in visited meat establishments, the audit team noted that official controls did not notice that certain operators mixed various categories of ABP without downgrading, as required, the resulting ABP to the highest risk category. The controls did not identify the incorrect categorisation of ABP and their subsequent dispatch to destinations that are not authorised under EU rules. For instance, Category 2 (e.g. tissues of animals presenting signs of transmissible diseases or collected from the establishments floors) and Category 1 materials (e.g. ABP containing residues of SRM) were dispatched to biogas plants without prior processing and/or to processing plants approved for processing of Category 3 material (from which derived products can be used for feeding of farmed animals).

The VIUs' inspectors interviewed by the audit team demonstrated a high level of awareness of ABP requirements. However, in the absence of ABP specific control guides, the evidence of the inspections carried out showed that these were neither conducted consistently nor systematically checked the relevant aspects. Consequently, some of the applicable requirements had been checked multiple times and some not at all.

Medlemsstatens svar

r The DVFA will initiate new training programs, which will ensure that proper introduction and training is given to the staff performing official controls on ABP. This is carried out through a more systematic peer-to-peer training and an intro training program, which is developed and approved in a cooperation between the staff in the Head Office and in the local inspection Units. This is expected to be in place by the beginning of 2024.

FA will initiate - new training programs, which will ensure that proper introduction and training is given to the staff performing official controls on ABP. This is ied out through a more systematic peer-to- peer training and an intro training program, which is developed and approved in a cooperation between the staff in the Head Office and in the local inspection Units. This is expected to be in place by the beginning of 2024.

Rekommendationen anger som grund

The DVFA has put in place numerous written procedures, manuals, guidelines and checklists supporting the staff during their official control duties, including inspections and sampling. The audit team examined these documents and noted the following: a. The guides, manuals and procedures for general organisation of official controls apply across the food and the feed areas of the DVFA controls and, except for SRM, do not contain specific references or requirements for carrying out ABP controls. b. The guides, manuals, procedures and checklist covering specific areas and types of operators, establishments and plants are entirely food and feed orientated, and refer solely to requirements of the food, the feed, the animal health and the animal welfare legislation. Except for SRM and sampling on DP at APB operators, they cover none of the ABP requirements applicable to operators acting along the ABP chain, even though numerous operators in the country, including these from the food and the feed sector, generate and/or handle ABP and/or DP. 5 Therefore, the documented control procedures are not adequate and sufficiently specific to support the staff performing official controls along the ABP chain and to ensure the effectiveness, appropriateness, quality and consistency of those controls. This is not in line with requirements of Article 5(1)(a) and (b) and Article 12(1) of Regulation (EU) 2017/625.

DVFA did not provide, in the last four years, the ABP specific training to the staff verifying ABP requirements along the ABP chain. Moreover, existing operating control procedures, manuals and guides do not provide sufficient instructions to staff to enable them to competently and consistently carry out their inspection duties on operators acting along the chain of ABP and DP. The lack of appropriate support, in terms of training and documented procedures, to the staff performing official controls along the ABP chain undermines the overall quality and effectiveness of these controls.

The existing guidelines and manuals for official controls in the meat producing sector (applying country-wide) do not cover ABP requirements, except these related to the removal of SRM (see finding 10).

The audit team analysed existing control guidelines, manuals and instructions and a selection of inspection reports provided by DVFA and noted that: a. Official controls in dairy establishments are feed-orientated and, while focusing on products of milk origin destined for feeding purposes (mostly a whey), do not cover other ABP destined for disposal. Moreover, within a limited selection of requirements applicable to ABP destined for feeding purposes, inspection reports lack information on verification of testing requirements applicable to feed materials 11 of animal origin laid down in Chapter 1 of Annex X to Regulation (EU) No 142/2011. b. Official controls on operators running kitchen activities and generating catering waste focus on requirements and rules derived from the food hygiene and animal health legislation, omitting requirements of the EU ABP legislation. c. Official controls at farms (carried out by the DVFA and the DAA), cover different requirements; namely: i) these by DVFA are limited to checks on conditions for temporary storing of fallen animals and the time of their collection (requirement derived from the animal health legislation), and ii) those by the DAA cover feed hygiene and feed ban (the requirements derived from the feed and TSE legislation). Neither of the controls cover requirements of ABP legislation applicable to farms (e.g. the conditions for the use of ABP of milk origin by farms depending on the type of ABP they receive, and these applicable to ABP originating at farms either due to on-farm slaughter or for fallen animals, for example: collection, transport, disposal, documentation accompanying ABP consignments, etc.). 5.2.3.2 Official controls in ABP plants

The VIUs' inspectors interviewed by the audit team demonstrated a high level of awareness of ABP requirements. However, in the absence of ABP specific control guides, the evidence of the inspections carried out showed that these were neither conducted consistently nor systematically checked the relevant aspects. Consequently, some of the applicable requirements had been checked multiple times and some not at all.

Medlemsstatens svar

The DVFA will create control supporting guides for the control of ABP, which will ensure consistency and documentation. The guides will be directed to the various situation where control of ABP is done. This is expected to be carried out by the end of 2023.

FA will create - l supporting for the control of Page: 1 19 June 2023 y the competent authorities of Denmark in response to Report ref. DG(SANTE) 2022-7416 of the audit carried out from 14 22 in order to evaluate the implementation of official controls on animal by-products (ABP) and derived products (DP) tion Proposed by the DK COMMENTS TO ­ mpetent authority Commission services' assessment of the response of competent authority P, which will ensure nsistency and cumentation. The ides will be directed the various situation ere control of ABP is ne. This is expected to carried out by the end 2023.

Rekommendationen anger som grund

The CCA has adopted several legal acts and guidelines, relevant for the scope of this audit, in recent years. According to the CCA, the changes introduced by these are mainly updates to reflect changes made to EU legislation. The CAs and food business operators (FBOs) are implementing the updated legislation and guidance documents. However, · Guidance No 9504/2021 (on meat inspection), contains certain provisions for post- mortem inspection (PMI) in red meat slaughterhouses that are not in line with Regulation (EU) 2019/627 (see paragraphs 49 and 50). · Guideline No 9853/2017 (on microbiological criteria for foodstuffs) contains certain provisions that are not in line with Regulation (EC) No 2073/2005 (see paragraphs 69 and 70).

The official control system is supported by updated national legislation and guidelines. Certain provisions in the guidance documents are contrary to the requirements of Regulations (EU) 2019/627 and (EC) No 2073/2005 which impacts the application of related EU requirements.

There are cooperation and coordination mechanisms in place within DVFA's departments and a system of notification of other CAs when the non-compliances identified need follow-up by other CAs. The OVs at slaughterhouse level notify welfare and animal identification issues that may concern the transport or farm level. Nonetheless, the outcome of on the farm/database controls is not considered when making decisions on animal identification (see paragraphs 40 and 41). 4 5.2.2 Impartiality, freedom from conflict of interest

The CA considers that, the recommendation on enforcing provisions that animals are clean prior to slaughter, has been addressed. Nonetheless, the findings at the time of the current audit suggest that the recommendation was not adequately addressed. AMI in cattle did not include checking animal cleanliness (see paragraph 39). This was replaced by CA audits of FBOs' systems for dealing with dirty cattle. 5.2.6 Enforcement

In large slaughterhouses the OVs permanently present during slaughter perform AMI. In the small slaughterhouses, private veterinarians employed on a contract basis for AMI perform this task. AMI includes controls on most of the relevant aspects (e.g. health status, identification, food chain information and animal welfare). However, AMI does not include verification of FBOs compliance with their obligation to ensure that animals have a clean hide, skin or fleece, so as to avoid any unacceptable risk of contamination of the fresh meat during slaughter. This is not in line with Article 11(4) of Regulation (EU) 2019/627.

Animal health and animal welfare findings of AMI are reported to the relevant services 9 and on-farm investigations are conducted when considered necessary. However, the OVs allow slaughter for human consumption of unidentified or incorrectly identified bovine animals in the absence of an adequate investigation to ascertain the identity of the animal.

When a bovine is presented for slaughter with eartags of an animal that appears as belonging to the category of stolen/dead on large grazing areas (and not found) in the central bovine database or is without eartags, the owner is allowed to make a sworn declaration for a change of details in the central bovine database. The CA allows slaughter for human consumption if the breed, sex and age of the animal matches the (new details) notified by the farmer and accepted by the database. The CA does not receive the outcome of a(possible) investigation before making the decision on the animal. (2)

In the cases assessed, the OVs were aware that red meat animals need to be inspected less than 24 hours after arrival and less than 24 hours before slaughter and had agreed arrangements with the FBOs to ensure this happens, including when animals were delivered during the weekend. AMI of all animals is documented. Nonetheless, should the agreed arrangements for the FBOs (to call the OV for AMI when new animals are presented for slaughter) fail to deliver, there is no reconciliation carried out between the number of animals passing AMI with the number passing PMI, nor are the OVs required to do so.

The CA has put provisions in place for AMI to be carried out at the holding of provenance. This option is used for emergency slaughter of animals and for animals that are difficult to handle. Approximately 350 private veterinarians have been designated as OVs for AMI on farms. The designation was not precluded by training or a system of measures to safeguard potential conflict of interest. The CA organised an information session and plans for more such sessions to be delivered to reach all designated OVs. Ante-mortem inspection - poultry

In the cases assessed there were sufficient OVs/OAs to cover all inspection posts during production and for all carcases to receive PMI. There are arrangements in place for PMI including official controls for Trichinella in pigs, tuberculosis and cysticercosis in bovines and pigs. The "in principle" standard are Articles 18-24 of Regulation (EU) 2019/627, while in certain circumstances (e.g. for outdoor pigs) the inspection maintains the additional incisions and palpations as per previous Regulation (EC) No 854/2004. Nonetheless, certain provisions of Guidance 9504/2021 differ from the provisions of Regulation (EU) 2019/627 · The guidance allows carcases with certain pathology and contamination, including abscess and pus contamination, to pass PMI. The FBO is required to remove such PMI findings (that are considered as not severe by the CA) and health-mark the carcasses. The CA carries out controls on a limited number of carcases and a zero tolerance system is not in place. This is contrary to the requirements of Article 45 (o) and (r) and 48 of Regulation (EU) 2019/627. · The guidance allows carcases of bovine animals more than 2 years old, where less than 10 Cysticercus cysts are found at PMI, to be passed as fit for consumption, without freezing, after removal of body parts where the cysts were found. In the bovine slaughterhouse visited, the limit was lowered to 5 cysts, but the carcases were still passed as fit for human consumption, without freezing, contrary to the 11 requirements of Articles 30 and 45(h) of Regulation (EU) 2019/627.

In addition, PMI does not check for the presence/safe removal of SRM, contrary to the requirements of Article 17(d)(ii) of Regulation (EU) 2017/625. Carcases with SRM are health-marked, contrary to the requirements of Article 45(q) of Regulation (EU) 2019/627. The OVs do not subsequently check if the FBO has removed SRM from all health-marked carcases either. The OVs check FBOs' procedures for SRM removal, during audits on related FBOs procedures, a few times a year.

The cows' udders were not inspected unless they went for human consumption and/or pet food in the case assessed. The OV explained this is in line with the clarification received from central level. However, this approach is contrary to the requirements of Article 19(1)(k) of Regulation (EU) 2019/627.

The audit team identified different additional gaps in PMI in the sheep and bovine slaughterhouses visited, contrary to the requirements of Articles 12(3), 17 and 45(r) of Regulation (EU) 2019/627: · In the large bovine slaughterhouse, the inadequate presentation of the spleen by the FBO (long highlighted by the OV, but not addressed) resulted in the spleen not being routinely inspected;. · In the small slaughterhouses, several gaps in PMI had been noted by the audit team, but not by the supervisor present that day [e.g. in the sheep slaughterhouse, the OV was not checking the lower part of the carcase (and therefore missing several carcases with contamination in the lower part), not palpating the lungs and livers, inspecting green offal only at random; for bovine carcases the OV could not inspect the upper part of the carcase (as was not using a stand); the kidneys were not presented decapsulated neither for sheep nor for cattle and were not adequately inspected]; · The OVs performing PMI had stamped carcases with some "minor" (in terms of size) contamination in both small slaughterhouses and to a larger extent in the sheep slaughterhouse. The supervising OV identified only a small number thereof.

The CA has arrangements in place for poultry PMI to be carried out by OVs and OAs as well as in some cases by authorised slaughterhouse staff under the supervision of the OV. The slaughterhouse staff is authorised to perform certain PMI tasks only after undergoing relevant training and passing an assessment by the CA. The arrangements for poultry PMI were found to be largely satisfactory in the case assessed, where slaughterhouse staff were performing PMI, apart from the absence of a system for the OV/OA to check viscera and body cavities on a representative sample for each flock, contrary to the requirements of Article 25 of Regulation (EU) 2019/627. 5.3.2 Official controls over food business operators' obligations Food Chain Information and animal identification Legal requirements Article 3 of Regulation (EC) No 853/2004 in conjunction with its Annex II, Section III and Article 9 and 10, 40, 41 42 and 43.1 of Regulation (EU) 2019/627. Findings

The OVs check FBO procedures to ensure that only animals/batches of animals that are properly identified and traceable to origin are slaughtered and these checks have been satisfactory in the cases assessed, with FBOs bringing inadequately identified and unidentified animals to the attention of the OVs. Nonetheless, the action taken by the CA in these cases is not fully satisfactory (see paragraphs 40 and 41). General and specific hygiene requirements, including HACCP based procedures Legal requirements Articles 4 and 5 of Regulation (EC) No 852/2004, Article 3 of Regulation (EC) No 853/2004 and Article 46 of Regulation (EU) 2019/627. 13 Findings

Health-marking is applied either by the OV or by the FBO under the supervision of the OV. In the case assessed, the health mark was kept under official control outside of working hours. Controls on the application of identification marks are included in the audit plan and are carried out at least once a year. However, contrary to the requirements of Article 48(2)(a) of Regulation (EU) 2019/627, the health mark is also applied in certain cases when there are grounds for declaring the meat unfit for human consumption (see paragraphs 49-52).

There is a system in place for meat inspection on bovine, ovine, porcine and poultry that covers most relevant aspects. However, the effectiveness of AMI and, to a larger extent, PMI is impacted by a number of systemic shortcomings.

Medlemsstatens svar

Regarding section 5.1, finding no. 1: Regarding section 5.1, finding no. 1: The official veterinarian shall declare Until now, it has been the fresh meat unfit for human understanding that the Danish system consumption according to article 45 is in accordance with the EU in Regulation (EU) no 2019/627. This legislation. The comments from the is in line with the guideline no Commission do leave the DVFA to 9504/2021 that states that all reconsider the practice in question. carcasses are to be PM inspected, and The DVFA will return with a final all abnormalities and contaminations position ultimo 2023. are to be marked by the CA for later removal because they are unfit for Regarding section 5.3, finding no. 40, human consumption. The meat 41 and 57: inspection handles severe findings As for now a farmer can report an that requires further attention from animal lost by making a notification the CA (red code), while the FBO on a special "holding-number" handles the rest (blue codes). 105000/3965. Later if the farmer finds out that the animal is still alive, it is possible for the farmer to contact According to Article 7 (3)(a)(b) of the people working with the central Regulation (EU) 2

Rekommendationen anger som grund

In accordance with the same procedure, the CA notified the national derogation on increased (slaughterhouse capacity) thresholds for performing delayed PMI compared to those provided in point (a), (i) and (ii) of Article 13(1) of Regulation (EU) 2019/627. However, the notification concerns 11 slaughterhouses whilst, at the time of the audit, delayed PMI was performed in most of the small slaughterhouses in the scope of the audit. Contrary to the requirements of Article 13(2) of Regulation (EU) 2019/627, the CA did not take into consideration, in the above notification, the combined annual production of the establishments where delayed PMI is performed. Conclusions on legislation and implementing measures

The CA notified the national derogations and adaptations. However, given that the notification of national measures on delayed PMI, carried out in accordance with the procedure laid down in Directive (EU) 2015/1535, was incomplete, the Commission has not granted the (already implemented) derogation.

Medlemsstatens svar

The DVFA notified the Commission The notification has been uploaded in on 8. November 2019, the TRIS on the 25th of April 2023, notification was published under please see Annex IV. number 2019/552/DK. There was a standstill period ending on 10 February 2020, during which no national steps could be taken. As no Page: 24 tions, updated 2023/05/12 Commission services' assessment of the action plan submitted by the competent authorities of E) 2022-7434 of the audit carried out from 07 June 2022 to 21 June 2022 in order to evaluate the control systems in place ction requirements Action Proposed by the competent Updated action proposed by the authority competent authority detailed opinions were noted during the standstill period, the provision could enter into force at the expiry of the standstill period according to the rules on the notification procedure. Even though Denmark has notified the derogation on delayed post mortem and the notification is still into force we will re-notify by the first quarter of 2023. This notification will include the combined animal production of the slaughterhouses of which the delayed post mortem inspection is applied.

Rekommendationen anger som grund

The CA provides training to the official control staff, including OVs and official auxiliaries (OAs). In the absence of a national system to deliver OV training in line with Article 13 and Chapter I of Annex II to Regulation (EU) 2019/624, the supervisors assess the training needs upon recruitment of an OV and deliver the training considered needed. The CA highlighted that a centrally developed/harmonised training is in the process of being developed. The training needs for OVs already active are identified during routine supervision and then custom made for the respective OV, sometimes also delivered by the supervisor. When the supervisor is not able to adequately detect knowledge gaps and poor performance ­ as seen during the audit for small slaughterhouses ­ the OV does not receive the adequate training.

The training topics for a new OV (developed by one of the four supervisors of OVs in small slaughterhouses) did not cover aspects on which the OV is required to perform controls such as animal welfare and animal identification. In addition, no control procedure is available for the latter. The OV performing AMI, PMI and some animal welfare controls in one of the small slaughterhouses visited did not have a satisfactory knowledge of PMI requirements (see paragraph 52) and of certain animal welfare 5 requirements (e.g. 0.7- 0.8A for 15 seconds was considered better than the 1A for effective electric stun).

The OVs benefit from the technical support of the supervisors. The supervisors for the teams in large slaughterhouse met by the audit team were knowledgeable in the areas under review. Nonetheless, the audit team identified gaps in the knowledge of the supervisor for small slaughterhouses, met during the audit (e.g. as regards prevention of cross-contamination and PMI ­ see paragraphs 52, 60-62, 64). The supervisor also had an incorrect understanding of certain animal welfare requirements (e.g. considering captive bolt a stun-to-kill method for which effectiveness of stunning did not need to be checked).

As the supervisor is the only CA member of staff performing controls on FBOs' hygiene requirements, as well as supervision of OVs contracted for AMI, PMI and some animal welfare controls, these knowledge gaps have impacted the effectiveness of the controls in the small establishments visited during the current audit (see paragraphs 60-62, 64, 66). The same supervisor was in charge of controls in 25 small slaughterhouses, assessing the training needs, training and supervising the contracted OVs in these establishments. This questions the effectiveness of the controls in the remaining small slaughterhouses.

The CA structure, organisation and resources, including support, are adequate, overall, to deliver effective controls in large slaughterhouses. However, this is not the case for small slaughterhouses, where the system relies heavily on a small number of staff to perform audits, act as supervisors, determine training needs and deliver the training of OVs. This system has proved inadequate in terms of detecting poor hygiene practices as well as poor staff performance/knowledge gaps and has resulted in ineffective controls.

Medlemsstatens svar

f Training activities: E-learning courses for OVs in AMI and PMI for ill develop new e-learning cattle/pigs/sheep/poultry/horses/deer ourses including video material will be developed, focusing on lege covering AMI and PMI for artis performance of AMI and PMI, cattle/pigs/sheep/poultry and considering the needs identified professional exercise of authority. during the audit. A working group The e-learning courses will be used will prepare a manuscript for each course with deadline August 2023, r training new employees the educational material is produced including substitute OVs) and for the with a set of deadline October 2023, tinuous training of OV's when and OVs are required to complete the training needs are identified. All OVs relevant courses before the end of 2023. Subsequently, the manuscript Page: 25 tions, updated 2023/05/12 Commission services' assessment of the action plan submitted by the competent authorities of E) 2022-7434 of the audit carried out from 07 June 2022 to 21 June 2022 in order to evaluate the control systems in place ction requirements Proposed by the competent Updated action proposed by the ity competent authority ming control in small and course mater

Rekommendationen anger som grund

Half of the active production areas (around 30 production areas in 2021) have a permanent classification. However, the remaining active production areas (or parts of them) are classified on a weekly basis (see Appendix 1 for further details) based on single results instead of on data from a review period, which is contrary to Article 52(3) of Regulation (EU) 2019/627. The same system (permanent and non-permanent classification) applies to aquaculture farms.

Regarding the production areas and aquaculture farms with permanent classification: · The competent authority has fixed a review period of three years for sampling data in order to determine compliance with the requirements for class A, B and C areas, as required in Article 52(3) of Regulation (EU) 2019/627. · According to national legislation, the DVFA can classify production areas and aquaculture farms only if there are at least 24 analytical results in the previous three years, as foreseen in the community guide 7 and in the guidance document of the European Union Reference Laboratory (EURL) for monitoring bacteriological and viral contamination of bivalve molluscs 8. · Every year (except in two occasions in the last 15 years), the DTU carries out the assessment of the results of the previous three years and proposes the classification of each production area and aquaculture farm. 6 In their response to the draft report, the DVFA added that this structure has existed for several years and is latest confirmed by the main agreement between DTU and DVFA, dated l5th of October 2021, and by the document "COMMUNITY REFERENCE LABORATORIES/NATIONAL REFERENCE LABORATORIES (EU 625/2017), Member State: DENMARK, Rev. October 2021". 7 Community guide to the principles of good practice for the microbiological classification and monitoring of bivalve mollusc production areas (Issue 3, January 2017), available at https://ec.europa.eu/food/system/files/2018- 12/biosafety_fh_guidance_community_guide_bivalve_mollusc_monitoring_en.pdf. 8 The guidance document of the EU Reference Laboratory for monitoring bacteriological and viral contamination of bivalve molluscs ("Good Practice Guide; issue 7", December 2018). 5 · Once per year, the competent authority updates the list of production areas and aquaculture farms with permanent classification at the DVFA's website (available at https://www.foedevarestyrelsen.dk/Leksikon/Sider/Produktionsområder%20med%20 permanent%20mikrobiologisk%20klassificering.aspx), as required in Article 66(a) of Regulation (EU) 2019/627. The audit team noted that: · Following an agreement between the DVFA and the DTU, the classification of a production area applies to all molluscs' species available in that production area. Results from different molluscs' species are used for achieving the required 24 results. However, results from hand-picked samples that are collected closer to the shore, are excluded due to the higher level of contamination observed (see paragraph 12). Sanitary surveys highlight this different pattern in contamination and recommend the segregation of the corresponding (closer to the shore) parts, but the DVFA has not followed this recommendation. · The majority of the results considered for the classification of production areas are from harvesters' own-check plans. In the case of "opening samples" (five samples taken on the same day from the same harvesting ground, see Appendix 1) all five results are taken into account for making up the 24 results. This may give a non- adequate view of the E. coli variation in time, affecting therefore the accuracy of classification. · National legislation establishes that an appropriate number of samples have to be taken continuously in the previous 12 months. The DVFA considers that one analytical result in the previous 12 months is sufficient to keep the permanent classification of a production area and aquaculture farm9, which diverges from the recommendations of the above mentioned Community and EURL guides. · National legislation establishes a minimum frequency of sampling for keeping certain classification: every four weeks for class A areas, every thirteen weeks for class B and every six months for C, which for class B and C diverges from the recommendations of the above mentioned Community and EURL guides. · Contrary to Article 52(2) and (3) of Regulation (EU) 2019/627, a production area can have more than one classification within it. Namely, if a sample ta

The DVFA has a yearly project to verify the reliability of harvesters' results. This project consists of a reduced number of samples (30 per year for biotoxins and toxin- producing plankton; 60 for E. coli 12) that officials take at the same location and moment than harvesters, following the corresponding DVFA sampling protocols. The audit team noted that: · The audit evaluated a case with discrepancies between the two results, with official results higher than operators' results. The DVFA disregarded the results from its official samples and used the harvesters lower results for action or decision (see paragraph 19). This makes the entire verification project of no value. In addition, the DVFA does not carry out an assessment of the discrepant results to check, for example, a correspondence with particular samplers. 12 According to the data provided, harvesters analysed around 950 samples E. coli in 2020, 470 toxin-producing plankton and 670 for biotoxins. 10 · According to the project, officials take five samples for E. coli testing from each production area to mirror the harvesters' sampling (i.e. five samples every time), which reduces the temporal official data available. · Results from official testing for biotoxins show that two or three samples from the same location are often taken, resulting also in a reduction of the temporal official data available.

The temporary classification of production areas (or parts of them) and the permanent classification of production areas are not in line with EU requirements. Therefore, the system does not ensure proper classification of production areas (or their parts) and there is a risk of live bivalve molluscs with an inadequate level of E. coli contamination for direct consumption reaching the market, which is a potential threat for food safety.

The DVFA does not follow the sanitary surveys' recommendations for sampling points and has not established a monitoring programme following them. Consequently, the validity of the sampling points for microbiological quality and their representativeness for production areas is questionable and so is the classification of production areas.

Medlemsstatens svar

These recommendations have caused the DVFA to initialize work to reconsider the use of Sanitary Survey's, including the use of fixed sampling points and the set-up of monitoring of production areas used for classifications. This is done in collaboration with DTU and AU. The DVFA expects this process to complete within 2022.

Rekommendationen anger som grund

Contrary to Article 57 of Regulation (EU) 2019/627, the DVFA has not established a monitoring programme for production areas based on the examination of sanitary surveys (the aim of this programme is to ensure that the results are representative of the area in question, as it establishes the number of samples, geographical distribution of sampling points and sampling frequencies). This is the second recommendation of the previous audit report that is still pending. 5.1.2 Monitoring of classified production areas

The DVFA does not follow the sanitary surveys' recommendations for sampling points and has not established a monitoring programme following them. Consequently, the validity of the sampling points for microbiological quality and their representativeness for production areas is questionable and so is the classification of production areas.

Medlemsstatens svar

se recommendations have caused the DVFA to initialize work to reconsider the use of the Sanitary Survey's, including the establishment of monitoring programs. This is done in collaboration with DTU and AU. The DVFA expects this process to complete within 2022.

Rekommendationen anger som grund

The national legislation, which is the basis for the control system of live bivalve molluscs (Order No 1793 of 2 December 2020 and Guideline No 9042 of 29 January 2021 concerning food hygiene), deviates from almost all the EU requirements for bivalve molluscs (this report describes those deviations in detail).

Regarding the production areas and aquaculture farms with permanent classification: · The competent authority has fixed a review period of three years for sampling data in order to determine compliance with the requirements for class A, B and C areas, as required in Article 52(3) of Regulation (EU) 2019/627. · According to national legislation, the DVFA can classify production areas and aquaculture farms only if there are at least 24 analytical results in the previous three years, as foreseen in the community guide 7 and in the guidance document of the European Union Reference Laboratory (EURL) for monitoring bacteriological and viral contamination of bivalve molluscs 8. · Every year (except in two occasions in the last 15 years), the DTU carries out the assessment of the results of the previous three years and proposes the classification of each production area and aquaculture farm. 6 In their response to the draft report, the DVFA added that this structure has existed for several years and is latest confirmed by the main agreement between DTU and DVFA, dated l5th of October 2021, and by the document "COMMUNITY REFERENCE LABORATORIES/NATIONAL REFERENCE LABORATORIES (EU 625/2017), Member State: DENMARK, Rev. October 2021". 7 Community guide to the principles of good practice for the microbiological classification and monitoring of bivalve mollusc production areas (Issue 3, January 2017), available at https://ec.europa.eu/food/system/files/2018- 12/biosafety_fh_guidance_community_guide_bivalve_mollusc_monitoring_en.pdf. 8 The guidance document of the EU Reference Laboratory for monitoring bacteriological and viral contamination of bivalve molluscs ("Good Practice Guide; issue 7", December 2018). 5 · Once per year, the competent authority updates the list of production areas and aquaculture farms with permanent classification at the DVFA's website (available at https://www.foedevarestyrelsen.dk/Leksikon/Sider/Produktionsområder%20med%20 permanent%20mikrobiologisk%20klassificering.aspx), as required in Article 66(a) of Regulation (EU) 2019/627. The audit team noted that: · Following an agreement between the DVFA and the DTU, the classification of a production area applies to all molluscs' species available in that production area. Results from different molluscs' species are used for achieving the required 24 results. However, results from hand-picked samples that are collected closer to the shore, are excluded due to the higher level of contamination observed (see paragraph 12). Sanitary surveys highlight this different pattern in contamination and recommend the segregation of the corresponding (closer to the shore) parts, but the DVFA has not followed this recommendation. · The majority of the results considered for the classification of production areas are from harvesters' own-check plans. In the case of "opening samples" (five samples taken on the same day from the same harvesting ground, see Appendix 1) all five results are taken into account for making up the 24 results. This may give a non- adequate view of the E. coli variation in time, affecting therefore the accuracy of classification. · National legislation establishes that an appropriate number of samples have to be taken continuously in the previous 12 months. The DVFA considers that one analytical result in the previous 12 months is sufficient to keep the permanent classification of a production area and aquaculture farm9, which diverges from the recommendations of the above mentioned Community and EURL guides. · National legislation establishes a minimum frequency of sampling for keeping certain classification: every four weeks for class A areas, every thirteen weeks for class B and every six months for C, which for class B and C diverges from the recommendations of the above mentioned Community and EURL guides. · Contrary to Article 52(2) and (3) of Regulation (EU) 2019/627, a production area can have more than one classification within it. Namely, if a sample ta

As required by Article 56(2) of Regulation (EU) 2019/627, the DVFA has completed sanitary surveys for all of the production areas (this is one of the two recommendations of the previous audit report still pending). By request of the DVFA, the DTU has carried out sanitary surveys, grouping all the production areas according to the water basins where they are located (fourteen sanitary surveys in total), and has published them on its website. The DTU plans to revise all sanitary surveys in a 5-year period (2021-2025). The audit team noted that: · The DTU informed that the revision will not include all the content but most of it and that it does not plan to carry out annual reviews (recommended by the EURL guidance document to ensure that the environmental conditions have not changed). · The DTU expects certain changes (inclusion or exclusion of areas/new areas) during the revision of some of the sanitary surveys to ensure that sanitary surveys are representative of the area considered, as required in Article 58 of Regulation (EU) 2019/627. However, there are no clear procedures in place to ensure that a sanitary survey is completed before new areas are classified (see also paragraph 10 regarding temporary classification of production areas), contrary to what is required by Article 56(1) of Regulation (EU) 2019/627. · The audit team reviewed a few sanitary surveys and noted that they are well structured and contain, in general, the elements required by legislation. However, the 7 DTU confirmed that it does not carry out any sampling to confirm the significance of the points selected, which can diminish the accuracy of the surveys. · The DVFA does not follow the recommendations from sanitary surveys regarding the location of sampling points for E. coli sanitary surveys; therefore the DVFA cannot ensure that sampling results are representative of the production area concerned. · The DVFA does not follow other recommendations from sanitary surveys either (e.g. the segregation or exclusion of a production area due to the different microbiological patterns observed). This can have an impact as, according to national legislation, harvesting should not take place near dumping areas, wastewater discharges, harbours, marinas or in areas affected by pollutants that can cause a deterioration of the sanitary quality of shellfish; however, it does not provide specific details on what size the exclusion zone must.

Contrary to Article 59 of Regulation (EU) 2019/627, the DVFA does not monitor production areas to check the microbiological quality of live bivalve molluscs, the presence of toxin-producing plankton in water and the biotoxins and chemical contaminants in live bivalve molluscs.

Consequently, contrary to Article 61(1), the DVFA does not draw up sampling plans for the purpose of the checks mentioned in the previous point.

Very significant deficiencies of the official control system control weaken the ability of the authorities to ensure that bivalve molluscs harvested in Denmark are safe for the consumers.

The DVFA does not monitor production areas (and consequently has not drawn up sampling plans) to check the microbiological quality of bivalve molluscs, the presence of toxin-producing plankton in seawater, and the presence of biotoxins and heavy metals in bivalve molluscs. This poses a risk that live bivalve with high levels of E. coli contamination, or with biotoxins and heavy metals content above the regulated limits, might be placed on the market.

Medlemsstatens svar

The DVFA disagrees with the statements in No 14 and 32 regarding the lack of monitoring of production areas. The DVFA does monitor production areas for the presence of E. coli. However, recognizes that a sampling plan with fixed, sampling points has not been carried out, as SANTE F asks for, see response to recommendation 2. The DVFA direct the attention to article 57 in 627/2019, which stipulates that the competent authority shall ensure that the number of samples, the geographical distribution of sampling points and the sampling frequency for the programme shall ensure that the results of the analysis are representative of the area in question, not fixed. However, the DVFA believes that the Danish system delivers safe bivalve molluscs to the market. The DVFA will reconsider the use of Sanitary Survey's, the use of data and the set-up of monitoring of production areas; see the response to recommendations 1 and 2. This is done in collaboration with DTU and AU. Page: 1 f report ref. DG (SANTE) 2021-7253 of the audit carried out from 28 April 2021 bivalve molluscs Action Proposed by the competent authority The DVFA expects this process to complete within 2022. The national legislatio

Rekommendationen anger som grund

The national legislation, which is the basis for the control system of live bivalve molluscs (Order No 1793 of 2 December 2020 and Guideline No 9042 of 29 January 2021 concerning food hygiene), deviates from almost all the EU requirements for bivalve molluscs (this report describes those deviations in detail).

Contrary to Article 59 of Regulation (EU) 2019/627, the DVFA does not monitor production areas to check the microbiological quality of live bivalve molluscs, the presence of toxin-producing plankton in water and the biotoxins and chemical contaminants in live bivalve molluscs.

Consequently, contrary to Article 61(1), the DVFA does not draw up sampling plans for the purpose of the checks mentioned in the previous point.

Very significant deficiencies of the official control system control weaken the ability of the authorities to ensure that bivalve molluscs harvested in Denmark are safe for the consumers.

The DVFA does not monitor production areas (and consequently has not drawn up sampling plans) to check the microbiological quality of bivalve molluscs, the presence of toxin-producing plankton in seawater, and the presence of biotoxins and heavy metals in bivalve molluscs. This poses a risk that live bivalve with high levels of E. coli contamination, or with biotoxins and heavy metals content above the regulated limits, might be placed on the market.

Medlemsstatens svar

of The DVFA disagrees with the statement in No 14 regarding the on lack of monitoring of production areas. DVFA monitors h production areas for toxic, producing plankton but recognizes that a sampling plan with fixed sampling points has not been carried out, as SANTE F asks for. The DVFA direct the attention to article 57 in 627/2019, which stipulates that the competent authority shall ensure that the number of samples, the geographical distribution of sampling points, and the sampling frequency for the programme shall ensure that the results of the analysis are representative of the area in question, not fixed. However, the DVFA believes that the Danish system delivers safe bivalve molluscs to the market. The DVFA will reconsider the set-up of monitoring of production areas according to toxin- producing plankton. This is done in collaboration with DTU and AU. The DVFA expects this process to complete within 2022. The national legislation, Order No 1793 of 2 December 2020, and if relevant, Guideline No, 9042 of 29 January 2021 will be adjusted depending on the described above. The DVFA expects this process to complete within the first mid- 2023.

Rekommendationen anger som grund

The national legislation, which is the basis for the control system of live bivalve molluscs (Order No 1793 of 2 December 2020 and Guideline No 9042 of 29 January 2021 concerning food hygiene), deviates from almost all the EU requirements for bivalve molluscs (this report describes those deviations in detail).

Contrary to Article 59 of Regulation (EU) 2019/627, the DVFA does not monitor production areas to check the microbiological quality of live bivalve molluscs, the presence of toxin-producing plankton in water and the biotoxins and chemical contaminants in live bivalve molluscs.

Consequently, contrary to Article 61(1), the DVFA does not draw up sampling plans for the purpose of the checks mentioned in the previous point.

Very significant deficiencies of the official control system control weaken the ability of the authorities to ensure that bivalve molluscs harvested in Denmark are safe for the consumers.

The DVFA does not monitor production areas (and consequently has not drawn up sampling plans) to check the microbiological quality of bivalve molluscs, the presence of toxin-producing plankton in seawater, and the presence of biotoxins and heavy metals in bivalve molluscs. This poses a risk that live bivalve with high levels of E. coli contamination, or with biotoxins and heavy metals content above the regulated limits, might be placed on the market.

Medlemsstatens svar

of The DVFA disagrees with the statement in No 14 regarding the lack of monitoring of production areas. The DVFA does monitor Page: 2 f report ref. DG (SANTE) 2021-7253 of the audit carried out from 28 April 2021 bivalve molluscs Proposed by the competent authority icles production areas for biotoxins but recognizes that a sampling th fixed, sampling points has not been carried out, as asks for. The DVFA direct the attention to article 57 2019, which stipulates that the competent authority shall that the number of samples, the geographical distribution ling points, and the sampling frequency for the me shall ensure that the results of the analysis are ntative of the area in question, not fixed. However, The lieves that the Danish system delivers safe bivalve s to the market. A will reconsider the set-up of monitoring of ion areas according to biotoxins. This is done in ration with DTU and AU. A expects this process to complete within 2022. ional legislation, Order No 1793 of 2 December 2020 relevant, Guideline No, 9042 of 29 January 2021, will be d to comply with EU legislation for live bivalve molluscs. A expects this process to complete within the first of 2023.

Rekommendationen anger som grund

Article 65 of Regulation (EU) 2019/627 allows competent authorities to take into account the results of checks carried out by food business operators or organisations when deciding on classification, reclassification, opening and closures if, among other requirements, the competent authority designates the laboratories carrying out the tests. However, contrary to this Article, the DVFA has not designated the laboratories carrying out the analysis for harvesters. In Denmark, national legislation requires that harvesters carry out sampling and testing where they intend to harvest and the DVFA bases all its decisions mostly on the harvesters' results (although the DVFA verifications samples described in paragraph 18 are also included for permanent classification). Harvesters can choose any of the private laboratories with a contract with the DVFA. These private laboratories operate under the same conditions described in paragraph 5. The audit team intended to review the laboratory that carried out all testing for harvesters in the years 2019 and 2020, but this laboratory refused the meeting and the DVFA did not have access to it either. This laboratory reported 4,300 biotoxins results for the regulated toxin groups (from approximately 670 samples), and 950 E. coli results. According to some of the laboratory reports available, E. coli testing was subcontracted to a laboratory that used an accredited method but there was no further information available regarding this laboratory (e.g. name, location, analytical method used, etc.). In 2021, harvesters decided to change the laboratory.

As envisaged in national legislation, the DVFA bases the control system of bivalve molluscs on harvesters' own-checks results (see Appendix 1 for further details). The audit team noted the following regarding this part of national legislation: · The conditions for sampling and transport of samples for E. coli testing are not in line with ISO 6887-3, which may compromise the quality of samples (i.e. storage of samples at up to 15ºC instead of 10 ºC, 48 hours instead of 24 to start the analyses of samples, etc.), and samples are not labelled at the point of collection. As a consequence, their legal, scientific and technical validity cannot be ensured, contrary to Article 34(5) of Regulation (EU) 2017/625. · Regarding the conditions for sampling for toxin-producing plankton testing: For qualitative purposes, it consists of the pulling of a 20 microns mesh diameter net, three times vertically from the bottom to the surface. This method is adequate for qualitative purposes as it filters larger amounts of water and, therefore, concentrates the number of plankton cells. However, the 8 fact that the vessels are not in movement when samples are collected may diminish the performance of this method. For quantitative purposes, it requires the use of Niskin-like bottles, which might not allow the collection of a representative sample of the water column required in to Article 61(7) of Regulation (EU) 2019/627. · It requires that samples for toxin-producing plankton are taken at the same location as samples for biotoxins, being therefore of limited value as early warnings suggesting an accumulation of biotoxins, and not representative of the areas concerned, contrary to Article 61(1) of Regulation (EU) 2019/627. · It groups production areas for toxin-producing plankton sampling so harvesters have to take only one sample in any of the production areas constituting the group. However, the audit team noted some cases of very different results of samples taken on the same day in different production areas within the same group, which demonstrates that areas are not correctly grouped. · It requires weekly sampling of seawater for toxin-producing plankton testing except from November to March, when it can be done fortnightly. The weekly periodicity is adequate but the evidence presented to support fortnightly sampling does not provide a robust basis for this decision: the data evaluated are from 1991 to 2007 therefore the situation may have changed since then; the DVFA submitted a document that highlights the abundance of cysts of PSP producing plankton spp10. Therefore, peaks of plankton preceding biotoxin episodes can go unnoticed in between two samplings if they are taken 14 days apart. · It indicates that samples are normally heat-treated11 before being tested. This heat treatment, which takes place at organisations' premises, can lead to an unknown underestimation in the quantity of biotoxins present in the original sample and undermines their legal, scientific and technical validity, contrary to Article 34(5) of Regulation (EU) 2017/625. · It requires weekly sampling of bivalve molluscs for biotoxins testing except for PSP from October to March, when the sampling frequency can be lower depending on DVFA decision. However, the DVFA did not provide a risk assessment with robust basis for this possibility to reduce the frequency of PSP testing. 5.1.3 Decisions after monitoring/decisions by the competent authority

As the DVFA does not carry out official monitoring of production areas, it does not take any decision in line with Article 62 of Regulation (EU) 2019/627. All the decisions the DVFA took were based on harvesters' own-checks results (harvesters take samples in 10 In their response to the draft report, the DVFA confirmed that the possibility of fortnightly sampling has never been used. 11 In their response to the draft report, the DVFA clarified that is not the case of samples from line cultures and one of the ones from primary producers. 9 production areas and in aquaculture farms and, based on the analytical results, the DVFA opens production areas or aquaculture farms for harvesting with certain classification (see Appendix 1 for further details)). Article 65(2) of Regulation (EU) 2019/627 allows the use of food business operators' results for the purpose of deciding on the classification, reclassification, opening or closure of production areas under certain conditions, but the system in Denmark does not meet them. The audit team noted the following that limit the reliability of harvesters' results: · The DVFA does not carry out training, inspection or supervision of harvesters taking samples (other than a visit to five samplers per year focussing on the ones from which samples are suspected to have hygiene problems or with remarks on the reports from the different laboratories) and could not provide a figure on the number of samplers in Denmark because any harvester can be a sampler. According to the information provided by the DVFA, harvesters organisations tend to use always the same samplers. · Results are not representative of the whole production area as harvesters take samples where they want to harvest and not at designated points. · The protocols for harvesters included in national legislation are not sufficiently detailed. The practices used during sampling and transport of samples affect the quality of samples (see Appendix 1 for further details). · Harvesters cook samples before sending them to laboratories (national legislation indicate that samples are normally heat-treated before analysis, either by the immersion in boiling water for 30 seconds o by steaming). This may lead to an unknown underestimation in the quantity of biotoxins present in the original sample (see paragraphs 44, 45 and 46).

The absence of designated official and private laboratories, and the lack of official supervision and verification of the laboratories, undermines the ability of Danish authority to back up the decision it takes regarding harvesting of molluscs.

The DVFA takes decisions based on results of checks carried out by harvesters, but without taking all necessary steps to ensure their reliability. The heat treatment of most of the samples, the lack of training of samplers, absence of control and supervision of samplers, sampling points not representative of the area, etc. diminish the trustworthiness of these results. This poses a risk that live bivalve molluscs with high levels of E. coli contamination, or with biotoxins and heavy metals content above the regulated limits, might be placed on the market.

Medlemsstatens svar

cks These recommendations have caused the DVFA to reconsider ng the set-up regarding sampling. of The DVFA expects this process to complete within 2022. ng laboratories see comments to recommendations 10 to

Rekommendationen anger som grund

As envisaged in national legislation, the DVFA bases the control system of bivalve molluscs on harvesters' own-checks results (see Appendix 1 for further details). The audit team noted the following regarding this part of national legislation: · The conditions for sampling and transport of samples for E. coli testing are not in line with ISO 6887-3, which may compromise the quality of samples (i.e. storage of samples at up to 15ºC instead of 10 ºC, 48 hours instead of 24 to start the analyses of samples, etc.), and samples are not labelled at the point of collection. As a consequence, their legal, scientific and technical validity cannot be ensured, contrary to Article 34(5) of Regulation (EU) 2017/625. · Regarding the conditions for sampling for toxin-producing plankton testing: For qualitative purposes, it consists of the pulling of a 20 microns mesh diameter net, three times vertically from the bottom to the surface. This method is adequate for qualitative purposes as it filters larger amounts of water and, therefore, concentrates the number of plankton cells. However, the 8 fact that the vessels are not in movement when samples are collected may diminish the performance of this method. For quantitative purposes, it requires the use of Niskin-like bottles, which might not allow the collection of a representative sample of the water column required in to Article 61(7) of Regulation (EU) 2019/627. · It requires that samples for toxin-producing plankton are taken at the same location as samples for biotoxins, being therefore of limited value as early warnings suggesting an accumulation of biotoxins, and not representative of the areas concerned, contrary to Article 61(1) of Regulation (EU) 2019/627. · It groups production areas for toxin-producing plankton sampling so harvesters have to take only one sample in any of the production areas constituting the group. However, the audit team noted some cases of very different results of samples taken on the same day in different production areas within the same group, which demonstrates that areas are not correctly grouped. · It requires weekly sampling of seawater for toxin-producing plankton testing except from November to March, when it can be done fortnightly. The weekly periodicity is adequate but the evidence presented to support fortnightly sampling does not provide a robust basis for this decision: the data evaluated are from 1991 to 2007 therefore the situation may have changed since then; the DVFA submitted a document that highlights the abundance of cysts of PSP producing plankton spp10. Therefore, peaks of plankton preceding biotoxin episodes can go unnoticed in between two samplings if they are taken 14 days apart. · It indicates that samples are normally heat-treated11 before being tested. This heat treatment, which takes place at organisations' premises, can lead to an unknown underestimation in the quantity of biotoxins present in the original sample and undermines their legal, scientific and technical validity, contrary to Article 34(5) of Regulation (EU) 2017/625. · It requires weekly sampling of bivalve molluscs for biotoxins testing except for PSP from October to March, when the sampling frequency can be lower depending on DVFA decision. However, the DVFA did not provide a risk assessment with robust basis for this possibility to reduce the frequency of PSP testing. 5.1.3 Decisions after monitoring/decisions by the competent authority

The DVFA takes decisions based on results of checks carried out by harvesters, but without taking all necessary steps to ensure their reliability. The heat treatment of most of the samples, the lack of training of samplers, absence of control and supervision of samplers, sampling points not representative of the area, etc. diminish the trustworthiness of these results. This poses a risk that live bivalve molluscs with high levels of E. coli contamination, or with biotoxins and heavy metals content above the regulated limits, might be placed on the market.

Medlemsstatens svar

ntrols, These recommendations have caused the DVFA to reconsider ns the set-up of monitoring by food business operators or led organisations representing food business operators. alidity, The DVFA expects this process to complete within 2022. ing set-up of monitoring of toxin-producing plankton in ion areas, see recommendation 4. Page: 3 f report ref. DG (SANTE) 2021-7253 of the audit carried out from 28 April 2021 bivalve molluscs Action Proposed by the competent authority Regarding temperature for sampling and transport of E. coli samples, see recommendation 12

Rekommendationen anger som grund

As the DVFA does not carry out official monitoring of production areas, it does not take any decision in line with Article 62 of Regulation (EU) 2019/627. All the decisions the DVFA took were based on harvesters' own-checks results (harvesters take samples in 10 In their response to the draft report, the DVFA confirmed that the possibility of fortnightly sampling has never been used. 11 In their response to the draft report, the DVFA clarified that is not the case of samples from line cultures and one of the ones from primary producers. 9 production areas and in aquaculture farms and, based on the analytical results, the DVFA opens production areas or aquaculture farms for harvesting with certain classification (see Appendix 1 for further details)). Article 65(2) of Regulation (EU) 2019/627 allows the use of food business operators' results for the purpose of deciding on the classification, reclassification, opening or closure of production areas under certain conditions, but the system in Denmark does not meet them. The audit team noted the following that limit the reliability of harvesters' results: · The DVFA does not carry out training, inspection or supervision of harvesters taking samples (other than a visit to five samplers per year focussing on the ones from which samples are suspected to have hygiene problems or with remarks on the reports from the different laboratories) and could not provide a figure on the number of samplers in Denmark because any harvester can be a sampler. According to the information provided by the DVFA, harvesters organisations tend to use always the same samplers. · Results are not representative of the whole production area as harvesters take samples where they want to harvest and not at designated points. · The protocols for harvesters included in national legislation are not sufficiently detailed. The practices used during sampling and transport of samples affect the quality of samples (see Appendix 1 for further details). · Harvesters cook samples before sending them to laboratories (national legislation indicate that samples are normally heat-treated before analysis, either by the immersion in boiling water for 30 seconds o by steaming). This may lead to an unknown underestimation in the quantity of biotoxins present in the original sample (see paragraphs 44, 45 and 46).

The DVFA has a yearly project to verify the reliability of harvesters' results. This project consists of a reduced number of samples (30 per year for biotoxins and toxin- producing plankton; 60 for E. coli 12) that officials take at the same location and moment than harvesters, following the corresponding DVFA sampling protocols. The audit team noted that: · The audit evaluated a case with discrepancies between the two results, with official results higher than operators' results. The DVFA disregarded the results from its official samples and used the harvesters lower results for action or decision (see paragraph 19). This makes the entire verification project of no value. In addition, the DVFA does not carry out an assessment of the discrepant results to check, for example, a correspondence with particular samplers. 12 According to the data provided, harvesters analysed around 950 samples E. coli in 2020, 470 toxin-producing plankton and 670 for biotoxins. 10 · According to the project, officials take five samples for E. coli testing from each production area to mirror the harvesters' sampling (i.e. five samples every time), which reduces the temporal official data available. · Results from official testing for biotoxins show that two or three samples from the same location are often taken, resulting also in a reduction of the temporal official data available.

The DVFA takes the following decisions after harvesters' own-checks results: a) For microbiological quality of live bivalve molluscs · The DVFA temporarily downgrades the classification of production areas following results over the expected ones for a given classification. However, permanent A production areas temporarily downgraded to B, go back to A following acceptable results of five samples taken on the same day, which is contrary to Article 52(3) of Regulation (EU) 2019/627 (this reversion should only take place if the evaluation of the results available for the review period determines compliance with the corresponding standards). The DVFA permanently downgrades production areas if 5% of results in the previous three months are above the expected ones, which is also not in line with the same article. · In permanently classified production areas, the DVFA grants a higher or lower classification to smaller zones within the production area, based on the E. coli testing results (i.e. a three nautical miles square with A status within a class B production area), which is contrary to Article 52(2) and (3) of Regulation (EU) 2019/627. · The number of dilutions made during the analyses of samples was not enough to provide an exact number when results were >18,000 (see paragraph 11), therefore the DVFA cannot determine if these areas exceeded the limit of 46,000 established in Articles 54(3) and 55(2) of Regulation (EU) 2019/627 for class B and C. Consequently, in these cases, it could not take adequate decisions for reclassification and/or closure of production areas. b) For the presence of toxin-producing plankton in water · National legislation contains a list of toxin-producing plankton species considered as relevant for each of the biotoxin groups in the Danish waters, and establishes "action limits" for each species. These action limits trigger the individual sampling of all the individual production areas within the "algae surveillance area", but harvesting continues if biotoxin testing results are below the regulatory limits. This is contrary to Articles 61(3)(a) and 61(7) of Regulation (EU) 2019/627, which require, respectively, an intensive sampling following results suggesting an accumulation of toxins in live bivalve mollusc flesh and the precautionary closures of production areas until results of toxin analysis are obtained. c) For the presence of biotoxins in live bivalve molluscs 11 · As required in Article 62(1) of Regulation (EU) 2019/627, the DVFA keeps production areas closed and prevents the harvesting of live bivalve molluscs when biotoxins results exceed the regulatory limits established. However, the DVFA considers that biotoxins results start to be valid one week after the sample is taken, and are valid for seven days from that moment (see Appendix 1 for further details). 5.1.4 Recording and exchange of information

The DVFA takes decisions based on results of checks carried out by harvesters, but without taking all necessary steps to ensure their reliability. The heat treatment of most of the samples, the lack of training of samplers, absence of control and supervision of samplers, sampling points not representative of the area, etc. diminish the trustworthiness of these results. This poses a risk that live bivalve molluscs with high levels of E. coli contamination, or with biotoxins and heavy metals content above the regulated limits, might be placed on the market.

Medlemsstatens svar

her These recommendations have caused the DVFA to reconsider o the set-up of monitoring by food business operators or rs or organisations representing food business operators and the use of th data. The DVFA expects this process to complete within 2022.

Rekommendationen anger som grund

The only relevant species harvested in Denmark is common whelk (Buccinum undatum). The DVFA does not carry out the official controls required in Article 11 of Regulation (EU) 2019/624, in particular aimed at verifying the health standards and the specific requirements set out respectively in Chapters V and IX of Section VII of Annex III to Regulation (EC) No 853/2004.

According to national legislation, food business operators in establishments have to take samples to analyse E. coli, biotoxins and heavy metals. Based on a document from DTU, the DVFA had decided to reduce the frequency of gastropods sampling. Currently, sampling is limited to three months of the year (from April to June) with one sample in the first fishing week in each production area, and an additional sample every four weeks during that period. However, the document does not provide robust basis for this decision. It intends to find a justification for a continuous sampling instead of presenting a solid risk assessment scientifically justifying the decision to reduce frequencies (the document states that, based on the available scientific basis and data, it is not possible to provide very clear justifications and arguments for continuous sampling for the presence of biotoxins in gastropods in Denmark).

The DVFA has a yearly project to verify the reliability of food business operators' results similar to the one described in paragraph 18. 12 5.1.6 Movement of live bivalve molluscs and registration document accompanying batches

The fact that the DVFA does not carry out official controls on marine gastropods and that it requires from operators a very reduced number of testing, poses the risk that whelks with biotoxins content above the regulated limits might be placed on the market.

Medlemsstatens svar

The DVFA has two sampling projects a year (verification) on ired in non-filtering sea snails. These deal with the control of biotoxins and heavy metals, respectively. The samples are taken in auctions, cold stores, dispatch centres and processing plants. Based on the recommendation from SANTE F, the DVFA will look into whether these verification-sampling projects should be adjusted.

Rekommendationen anger som grund

Contrary to Article 37 of Regulation (EU) 2017/625, the DVFA has not designated official laboratories5. At the time of the audit, the DVFA laboratory tests official 5 In their response to the draft report, the DVFA added that, from June 2021, it has designated a number of private laboratories as official laboratories according to the criteria stated in Regulation (UE) 2017/625. The mentioned designation replaces the contracts that were signed earlier. The private laboratories perform the 3 samples for E. coli but it uses private laboratories for samples for toxin-producing plankton and biotoxins. The DVFA has a contract with these private laboratories (from 2010 and 2012 respectively). In the contracts, the DVFA requires laboratories to be EN ISO/IEC 17025 accredited by the accreditation body of the Member State where laboratories are located. It does not carry out audits of laboratories as it considers them redundant with the accreditation assessment, which is allowed by Article 39(1) of Regulation (EU) 2017/625. However, the DVFA does not carry out any type of control, supervision or verification of the conditions established in the contract. In addition, the contract allows laboratories to subcontract testing to other laboratories, under the same rules that in the original contract. Nevertheless, according to the DVFA, the compliance with the rules is exclusively the responsibility of the first laboratory and this responsibility is limited to confirming that the subcontracted laboratory is accredited by the accreditation body of the Member State where it is located.

Article 65 of Regulation (EU) 2019/627 allows competent authorities to take into account the results of checks carried out by food business operators or organisations when deciding on classification, reclassification, opening and closures if, among other requirements, the competent authority designates the laboratories carrying out the tests. However, contrary to this Article, the DVFA has not designated the laboratories carrying out the analysis for harvesters. In Denmark, national legislation requires that harvesters carry out sampling and testing where they intend to harvest and the DVFA bases all its decisions mostly on the harvesters' results (although the DVFA verifications samples described in paragraph 18 are also included for permanent classification). Harvesters can choose any of the private laboratories with a contract with the DVFA. These private laboratories operate under the same conditions described in paragraph 5. The audit team intended to review the laboratory that carried out all testing for harvesters in the years 2019 and 2020, but this laboratory refused the meeting and the DVFA did not have access to it either. This laboratory reported 4,300 biotoxins results for the regulated toxin groups (from approximately 670 samples), and 950 E. coli results. According to some of the laboratory reports available, E. coli testing was subcontracted to a laboratory that used an accredited method but there was no further information available regarding this laboratory (e.g. name, location, analytical method used, etc.). In 2021, harvesters decided to change the laboratory.

The absence of designated official and private laboratories, and the lack of official supervision and verification of the laboratories, undermines the ability of Danish authority to back up the decision it takes regarding harvesting of molluscs.

Medlemsstatens svar

quired Laboratories related to the official control of E. coli, biotoxins, and potentially toxic algae related to molluscs in Denmark. For the parameter E. coli, the following private laboratories are designated: · June 22nd 2021, * · June 7th 2021, For the parameter biotoxins, the following private laboratories are designated: Page: 4 f report ref. DG (SANTE) 2021-7253 of the audit carried out from 28 April 2021 bivalve molluscs Action Proposed by the competent authority · June 7th 2021, * For the parameter potentially toxic algae, the following private laboratories are designated: · June 22nd 2021,* Furthermore, The Danish Veterinary and Food Administration (DVFA) laboratory performs official analysis for the parameters E. coli.* ).

Rekommendationen anger som grund

Contrary to Article 37 of Regulation (EU) 2017/625, the DVFA has not designated official laboratories5. At the time of the audit, the DVFA laboratory tests official 5 In their response to the draft report, the DVFA added that, from June 2021, it has designated a number of private laboratories as official laboratories according to the criteria stated in Regulation (UE) 2017/625. The mentioned designation replaces the contracts that were signed earlier. The private laboratories perform the 3 samples for E. coli but it uses private laboratories for samples for toxin-producing plankton and biotoxins. The DVFA has a contract with these private laboratories (from 2010 and 2012 respectively). In the contracts, the DVFA requires laboratories to be EN ISO/IEC 17025 accredited by the accreditation body of the Member State where laboratories are located. It does not carry out audits of laboratories as it considers them redundant with the accreditation assessment, which is allowed by Article 39(1) of Regulation (EU) 2017/625. However, the DVFA does not carry out any type of control, supervision or verification of the conditions established in the contract. In addition, the contract allows laboratories to subcontract testing to other laboratories, under the same rules that in the original contract. Nevertheless, according to the DVFA, the compliance with the rules is exclusively the responsibility of the first laboratory and this responsibility is limited to confirming that the subcontracted laboratory is accredited by the accreditation body of the Member State where it is located.

Article 65 of Regulation (EU) 2019/627 allows competent authorities to take into account the results of checks carried out by food business operators or organisations when deciding on classification, reclassification, opening and closures if, among other requirements, the competent authority designates the laboratories carrying out the tests. However, contrary to this Article, the DVFA has not designated the laboratories carrying out the analysis for harvesters. In Denmark, national legislation requires that harvesters carry out sampling and testing where they intend to harvest and the DVFA bases all its decisions mostly on the harvesters' results (although the DVFA verifications samples described in paragraph 18 are also included for permanent classification). Harvesters can choose any of the private laboratories with a contract with the DVFA. These private laboratories operate under the same conditions described in paragraph 5. The audit team intended to review the laboratory that carried out all testing for harvesters in the years 2019 and 2020, but this laboratory refused the meeting and the DVFA did not have access to it either. This laboratory reported 4,300 biotoxins results for the regulated toxin groups (from approximately 670 samples), and 950 E. coli results. According to some of the laboratory reports available, E. coli testing was subcontracted to a laboratory that used an accredited method but there was no further information available regarding this laboratory (e.g. name, location, analytical method used, etc.). In 2021, harvesters decided to change the laboratory.

The absence of designated official and private laboratories, and the lack of official supervision and verification of the laboratories, undermines the ability of Danish authority to back up the decision it takes regarding harvesting of molluscs.

The DVFA has access to official or private laboratories, and has a list of laboratories that harvesters can use for their own-checks. However, as described in paragraphs 5 and 6, it does not carry out any type of control, supervision or verification of the conditions established in the contract with them. Consequently, it does not use the tools that would have allowed it to detect the deficiencies noted by the audit team. 5.2.1 Laboratories for microbiology

The DVFA provided a list of private laboratories with a contract with the DVFA to carry out the analyses of harvesters' own-check samples. Harvesters can decide which of these laboratories to use. The audit team asked to meet with one of these laboratories to assess the testing of the lipophilic toxin group, but the laboratory refused to meet the audit team or to provide information to the DVFA. Therefore, the findings below for the lipophilic toxin group relate exclusively to a documentary revision of the laboratory mentioned in paragraph 41.

Regarding the ASP laboratory reviewed: · The laboratory carries out the qualitative and quantitative analyses using methods alternative to the EU reference ones (apparently due to the request of the competent authority of the Member State where this laboratory is located). The corresponding national accreditation body has accredited both methods. Although the quantitative method can be considered as an international recognised validated method, as required in Annex V, Chapter II(A) to Regulation (EU) 2019/627, this is not the case for the qualitative method (Annex V, Chapter II(B) of the same Regulation). All samples are tested with the qualitative method and only when results are above a certain threshold (1mg of domoic acid per kilogramme) the laboratory uses the quantitative method, which represents around 1% of the tests. · The laboratory validated the qualitative method in 2013, with material different from the certified reference material, and it has not renewed the validation since then. The laboratory did the validation using raw bivalve molluscs' matrix for blue mussel, oyster and cockle. However, testing uses heat-treated samples, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The laboratory validated the quantitative method in 2004 and it has not renewed the validation since then. Recently, in 2020, the laboratory carried out the verification of the use of new equipment for this method. The validation was done using raw bivalve molluscs matrix for blue mussel, oyster and cockle (it is in the process to include tunicates) However, testing uses heat-treated samples, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The participation in the last proficiency testing for the quantitative method shows satisfactory results, despite the fact that the bivalve molluscs species of the test samples were different that the ones validated. However, since for this participation, the laboratory used raw bivalve molluscs (as in the validation) instead of the normal practice that is the testing of heat-treated samples, the proficiency testing results lack reliability for the testing performed.

Regarding the PSP laboratory reviewed: · The laboratory has the EU reference method accredited by the corresponding national accreditation body. However, during the performance of the method, it deviates from it and from the corresponding EURL standard operational procedure in the heat treatment of the sample, which, consequently, may have an impact on the testing results. 17 · The laboratory validated the method in 2012; it has not renewed the validation since then. The validation report stated that it used raw bivalve molluscs' matrix for the validation, but did not provide information on the bivalve mollusc species used for it. · As mentioned before, biotoxin testing is mostly carried out with heat-treated bivalve molluscs, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The laboratory does not test all the relevant PSP compounds (i.e. GTX6 and C3,4 are missing). · The participation results of the last proficiency testing shows good results for total toxicity (sum of all the compounds) in two samples, but an unsatisfactory result of - 5.5 in the third sample. The assessment of results shows that two compounds (i.e. GTX1-4 and dcSXT) are not detected in any of the three samples and two other compounds (i.e. GTX2-3 and STX) are not detected in the third sample. The laboratory suspects of false negatives but did not find conclusive explanation for these results. It did not carry out any further investigation (it stated that this biotoxin is very unusual to find so they did not do any further investigation; in addition, there was no sufficient remaining sample for re-testing). · The lack of testing of certain compounds and the underreporting of other compounds poses a risk of underestimation of PSP if such an episode occurs.

Regarding lipophilic toxin laboratories (findings are limited to the assessment of the laboratory procedures and the last participation in a proficiency testing of the laboratory carrying out the analyses of official and the harvesters' samples since 1 January 2021): · This laboratory procedure is not in line with the last version of the EURL (which is dated in 2015) and it deviates from the EURL procedure in the extraction of the lipophilic toxin compounds. · The laboratory does not detect all the relevant compounds of the lipophilic toxin group. · As mentioned before, biotoxin testing is carried out with heat-treated bivalve molluscs. This deviates from the EU reference method for lipophilic toxins, therefore it may have an impact on the testing results. · The proficiency testing results shows unsatisfactory and questionable results for some of the compounds of this group. The laboratory suspects there is a problem with the procedure of extraction. If it is the case, there is a risk of underestimation of the lipophilic toxin group if such an episode occurs (the lipophilic toxin group is the most common biotoxin in Danish waters). Conclusions on laboratories

The DVFA has access to adequate laboratory capacity, but lacks the oversight of their 18 work; therefore, it cannot assure the proper performance of laboratory testing activities.

Medlemsstatens svar

Within the procedure for designation as an official laboratory, it mply is now in the DVFA specific stated that the analysis method e should be the specified reference method or an alternative method validated against the reference method according to an internationally recognized standard (e.g. ISO 16140). Further, the specified procedure for analysis shall be on the list of methods comprised by the accreditation for the actual laboratory. The Danish Veterinary and Food Administration (DVFA) has appointed the following National Reference Laboratories (NRL): 1. E. coli ­ since 1998: * 2. Biotoxins ­ as per September 1st 2021: * : Page: 5 f report ref. DG (SANTE) 2021-7253 of the audit carried out from 28 April 2021 bivalve molluscs Action Proposed by the competent authority An NRL for potentially toxic algae is not approved in Denmark, as this parameter is not covered by EU regulation for establishing national reference laboratories. We still find that additional audits of accredited laboratories carried out by the DVFA are redundant under article 39. 1.

Rekommendationen anger som grund

Regarding the production areas and aquaculture farms with permanent classification: · The competent authority has fixed a review period of three years for sampling data in order to determine compliance with the requirements for class A, B and C areas, as required in Article 52(3) of Regulation (EU) 2019/627. · According to national legislation, the DVFA can classify production areas and aquaculture farms only if there are at least 24 analytical results in the previous three years, as foreseen in the community guide 7 and in the guidance document of the European Union Reference Laboratory (EURL) for monitoring bacteriological and viral contamination of bivalve molluscs 8. · Every year (except in two occasions in the last 15 years), the DTU carries out the assessment of the results of the previous three years and proposes the classification of each production area and aquaculture farm. 6 In their response to the draft report, the DVFA added that this structure has existed for several years and is latest confirmed by the main agreement between DTU and DVFA, dated l5th of October 2021, and by the document "COMMUNITY REFERENCE LABORATORIES/NATIONAL REFERENCE LABORATORIES (EU 625/2017), Member State: DENMARK, Rev. October 2021". 7 Community guide to the principles of good practice for the microbiological classification and monitoring of bivalve mollusc production areas (Issue 3, January 2017), available at https://ec.europa.eu/food/system/files/2018- 12/biosafety_fh_guidance_community_guide_bivalve_mollusc_monitoring_en.pdf. 8 The guidance document of the EU Reference Laboratory for monitoring bacteriological and viral contamination of bivalve molluscs ("Good Practice Guide; issue 7", December 2018). 5 · Once per year, the competent authority updates the list of production areas and aquaculture farms with permanent classification at the DVFA's website (available at https://www.foedevarestyrelsen.dk/Leksikon/Sider/Produktionsområder%20med%20 permanent%20mikrobiologisk%20klassificering.aspx), as required in Article 66(a) of Regulation (EU) 2019/627. The audit team noted that: · Following an agreement between the DVFA and the DTU, the classification of a production area applies to all molluscs' species available in that production area. Results from different molluscs' species are used for achieving the required 24 results. However, results from hand-picked samples that are collected closer to the shore, are excluded due to the higher level of contamination observed (see paragraph 12). Sanitary surveys highlight this different pattern in contamination and recommend the segregation of the corresponding (closer to the shore) parts, but the DVFA has not followed this recommendation. · The majority of the results considered for the classification of production areas are from harvesters' own-check plans. In the case of "opening samples" (five samples taken on the same day from the same harvesting ground, see Appendix 1) all five results are taken into account for making up the 24 results. This may give a non- adequate view of the E. coli variation in time, affecting therefore the accuracy of classification. · National legislation establishes that an appropriate number of samples have to be taken continuously in the previous 12 months. The DVFA considers that one analytical result in the previous 12 months is sufficient to keep the permanent classification of a production area and aquaculture farm9, which diverges from the recommendations of the above mentioned Community and EURL guides. · National legislation establishes a minimum frequency of sampling for keeping certain classification: every four weeks for class A areas, every thirteen weeks for class B and every six months for C, which for class B and C diverges from the recommendations of the above mentioned Community and EURL guides. · Contrary to Article 52(2) and (3) of Regulation (EU) 2019/627, a production area can have more than one classification within it. Namely, if a sample ta

The laboratory for E coli operates to good standards and the results in proficiency testing indicate its results are reliable. The laboratory presents certain shortcomings that may affect the quality of results and their completeness regarding classification of class B and C production areas.

Medlemsstatens svar

uction Temperature f · Correct temperature options are now mentioned in method SOP ANA-05.1004. , · ISO 6887-3: 2017, section 7.2.5., NOTE: "Studies have d C shown that E. coli will not significantly increase in nd mussels or Pacific oysters at temperatures of 15 °C or less for up to 48 h" should be added in the project description. It will be implemented in 2021. · Analysis of temperature measurements in 2018-2021 shows that only 12% of temperature during transport does not comply with the ISO method. All temperature deviations occur during hot periods when water and air are hot, and sample materials are hot from the start. Therefore, we will try to find a solution on how to cool sample material from the start by using larger cooler bags with more cooling elements. It will be implemented in 2021. Dilutions · Method SOP ANA-05.1004 has been examined, and analysis procedures for areas A, B, and C were mentioned. · It is now mentioned in Method SOP ANA-05.1004 that extra dilutions must be made each time, regardless of production area classification A B or C. Page: 6 f report ref. DG (SANTE) 2021-7253 of the audit carried out from 28 April 2021 bivalve molluscs Action Proposed by the

Rekommendationen anger som grund

Regarding the ASP laboratory reviewed: · The laboratory carries out the qualitative and quantitative analyses using methods alternative to the EU reference ones (apparently due to the request of the competent authority of the Member State where this laboratory is located). The corresponding national accreditation body has accredited both methods. Although the quantitative method can be considered as an international recognised validated method, as required in Annex V, Chapter II(A) to Regulation (EU) 2019/627, this is not the case for the qualitative method (Annex V, Chapter II(B) of the same Regulation). All samples are tested with the qualitative method and only when results are above a certain threshold (1mg of domoic acid per kilogramme) the laboratory uses the quantitative method, which represents around 1% of the tests. · The laboratory validated the qualitative method in 2013, with material different from the certified reference material, and it has not renewed the validation since then. The laboratory did the validation using raw bivalve molluscs' matrix for blue mussel, oyster and cockle. However, testing uses heat-treated samples, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The laboratory validated the quantitative method in 2004 and it has not renewed the validation since then. Recently, in 2020, the laboratory carried out the verification of the use of new equipment for this method. The validation was done using raw bivalve molluscs matrix for blue mussel, oyster and cockle (it is in the process to include tunicates) However, testing uses heat-treated samples, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The participation in the last proficiency testing for the quantitative method shows satisfactory results, despite the fact that the bivalve molluscs species of the test samples were different that the ones validated. However, since for this participation, the laboratory used raw bivalve molluscs (as in the validation) instead of the normal practice that is the testing of heat-treated samples, the proficiency testing results lack reliability for the testing performed.

The use of a screening method, different from the one laid down in EU legislation and not internationally recognised validated for ASP testing, raises doubts regarding the accuracy of the results provided. Consequently, it might compromise the use of the quantitative method due to false negative results.

Medlemsstatens svar

The designated NRL will work to resolve the issues with the laboratories. n The DVFA expects this process to complete within mid-2022.

Rekommendationen anger som grund

Regarding the ASP laboratory reviewed: · The laboratory carries out the qualitative and quantitative analyses using methods alternative to the EU reference ones (apparently due to the request of the competent authority of the Member State where this laboratory is located). The corresponding national accreditation body has accredited both methods. Although the quantitative method can be considered as an international recognised validated method, as required in Annex V, Chapter II(A) to Regulation (EU) 2019/627, this is not the case for the qualitative method (Annex V, Chapter II(B) of the same Regulation). All samples are tested with the qualitative method and only when results are above a certain threshold (1mg of domoic acid per kilogramme) the laboratory uses the quantitative method, which represents around 1% of the tests. · The laboratory validated the qualitative method in 2013, with material different from the certified reference material, and it has not renewed the validation since then. The laboratory did the validation using raw bivalve molluscs' matrix for blue mussel, oyster and cockle. However, testing uses heat-treated samples, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The laboratory validated the quantitative method in 2004 and it has not renewed the validation since then. Recently, in 2020, the laboratory carried out the verification of the use of new equipment for this method. The validation was done using raw bivalve molluscs matrix for blue mussel, oyster and cockle (it is in the process to include tunicates) However, testing uses heat-treated samples, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The participation in the last proficiency testing for the quantitative method shows satisfactory results, despite the fact that the bivalve molluscs species of the test samples were different that the ones validated. However, since for this participation, the laboratory used raw bivalve molluscs (as in the validation) instead of the normal practice that is the testing of heat-treated samples, the proficiency testing results lack reliability for the testing performed.

Regarding the PSP laboratory reviewed: · The laboratory has the EU reference method accredited by the corresponding national accreditation body. However, during the performance of the method, it deviates from it and from the corresponding EURL standard operational procedure in the heat treatment of the sample, which, consequently, may have an impact on the testing results. 17 · The laboratory validated the method in 2012; it has not renewed the validation since then. The validation report stated that it used raw bivalve molluscs' matrix for the validation, but did not provide information on the bivalve mollusc species used for it. · As mentioned before, biotoxin testing is mostly carried out with heat-treated bivalve molluscs, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The laboratory does not test all the relevant PSP compounds (i.e. GTX6 and C3,4 are missing). · The participation results of the last proficiency testing shows good results for total toxicity (sum of all the compounds) in two samples, but an unsatisfactory result of - 5.5 in the third sample. The assessment of results shows that two compounds (i.e. GTX1-4 and dcSXT) are not detected in any of the three samples and two other compounds (i.e. GTX2-3 and STX) are not detected in the third sample. The laboratory suspects of false negatives but did not find conclusive explanation for these results. It did not carry out any further investigation (it stated that this biotoxin is very unusual to find so they did not do any further investigation; in addition, there was no sufficient remaining sample for re-testing). · The lack of testing of certain compounds and the underreporting of other compounds poses a risk of underestimation of PSP if such an episode occurs.

Regarding lipophilic toxin laboratories (findings are limited to the assessment of the laboratory procedures and the last participation in a proficiency testing of the laboratory carrying out the analyses of official and the harvesters' samples since 1 January 2021): · This laboratory procedure is not in line with the last version of the EURL (which is dated in 2015) and it deviates from the EURL procedure in the extraction of the lipophilic toxin compounds. · The laboratory does not detect all the relevant compounds of the lipophilic toxin group. · As mentioned before, biotoxin testing is carried out with heat-treated bivalve molluscs. This deviates from the EU reference method for lipophilic toxins, therefore it may have an impact on the testing results. · The proficiency testing results shows unsatisfactory and questionable results for some of the compounds of this group. The laboratory suspects there is a problem with the procedure of extraction. If it is the case, there is a risk of underestimation of the lipophilic toxin group if such an episode occurs (the lipophilic toxin group is the most common biotoxin in Danish waters). Conclusions on laboratories

The deviation from EU reference methods for biotoxins or their corresponding EURL standard operational procedures in the implementation of analytical methods undermines significantly the reliability of results.

Although laboratories have accredited and validated the analytical methods for biotoxins, the reliability of results is significantly weakened by the fact that the validation took place more than a decade ago and laboratories use matrices (heat- treated molluscs) different from the ones for which the methods were accredited and validated for.

Laboratories use raw samples when demonstrating their performance in biotoxins proficiency testing, but routine testing uses heat-treated samples, which lack proficiency assurances. Therefore, routine analytical results may be underestimating the content of biotoxins.

Medlemsstatens svar

These recommendations have caused the DVFA to look into the ed in set-up regarding analysis of cooked vs raw samples of bivalve as molluscs for biotoxins. The DVFA expects this process to complete within 2022. The designated NRL will work to resolve the remaining issues with the laboratories. The DVFA expects this process to complete within mid-2022.

Rekommendationen anger som grund

Regarding the PSP laboratory reviewed: · The laboratory has the EU reference method accredited by the corresponding national accreditation body. However, during the performance of the method, it deviates from it and from the corresponding EURL standard operational procedure in the heat treatment of the sample, which, consequently, may have an impact on the testing results. 17 · The laboratory validated the method in 2012; it has not renewed the validation since then. The validation report stated that it used raw bivalve molluscs' matrix for the validation, but did not provide information on the bivalve mollusc species used for it. · As mentioned before, biotoxin testing is mostly carried out with heat-treated bivalve molluscs, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The laboratory does not test all the relevant PSP compounds (i.e. GTX6 and C3,4 are missing). · The participation results of the last proficiency testing shows good results for total toxicity (sum of all the compounds) in two samples, but an unsatisfactory result of - 5.5 in the third sample. The assessment of results shows that two compounds (i.e. GTX1-4 and dcSXT) are not detected in any of the three samples and two other compounds (i.e. GTX2-3 and STX) are not detected in the third sample. The laboratory suspects of false negatives but did not find conclusive explanation for these results. It did not carry out any further investigation (it stated that this biotoxin is very unusual to find so they did not do any further investigation; in addition, there was no sufficient remaining sample for re-testing). · The lack of testing of certain compounds and the underreporting of other compounds poses a risk of underestimation of PSP if such an episode occurs.

Regarding lipophilic toxin laboratories (findings are limited to the assessment of the laboratory procedures and the last participation in a proficiency testing of the laboratory carrying out the analyses of official and the harvesters' samples since 1 January 2021): · This laboratory procedure is not in line with the last version of the EURL (which is dated in 2015) and it deviates from the EURL procedure in the extraction of the lipophilic toxin compounds. · The laboratory does not detect all the relevant compounds of the lipophilic toxin group. · As mentioned before, biotoxin testing is carried out with heat-treated bivalve molluscs. This deviates from the EU reference method for lipophilic toxins, therefore it may have an impact on the testing results. · The proficiency testing results shows unsatisfactory and questionable results for some of the compounds of this group. The laboratory suspects there is a problem with the procedure of extraction. If it is the case, there is a risk of underestimation of the lipophilic toxin group if such an episode occurs (the lipophilic toxin group is the most common biotoxin in Danish waters). Conclusions on laboratories

The fact that the laboratory for PSP cannot analyse part of the compounds of this group, that the results in proficiency testing show that it cannot detect a number of other compounds and laboratories do not find conclusive explanations for unsatisfactory and questionable results in proficiency tests, indicates a risk of underestimation of this biotoxin.

The results in proficiency testing show that the laboratory for lipophilic toxins cannot detect a number of other compounds, and the lack of conclusive explanations for questionable and unsatisfactory results in participation proficiency testing weakens the reliability of lipophilic toxins testing results.

Medlemsstatens svar

The designated NRL will work to resolve the issues with the nt laboratories. its The DVFA expects this process to complete within mid-2022. ards of ) Page: 7 f report ref. DG (SANTE) 2021-7253 of the audit carried out from 28 April 2021 bivalve molluscs Action Proposed by the competent authority

Rekommendationen anger som grund

Regarding the PSP laboratory reviewed: · The laboratory has the EU reference method accredited by the corresponding national accreditation body. However, during the performance of the method, it deviates from it and from the corresponding EURL standard operational procedure in the heat treatment of the sample, which, consequently, may have an impact on the testing results. 17 · The laboratory validated the method in 2012; it has not renewed the validation since then. The validation report stated that it used raw bivalve molluscs' matrix for the validation, but did not provide information on the bivalve mollusc species used for it. · As mentioned before, biotoxin testing is mostly carried out with heat-treated bivalve molluscs, which is not the matrix used in the validation of the method, and this may have an impact on the testing results due to the potential matrix effect. · The laboratory does not test all the relevant PSP compounds (i.e. GTX6 and C3,4 are missing). · The participation results of the last proficiency testing shows good results for total toxicity (sum of all the compounds) in two samples, but an unsatisfactory result of - 5.5 in the third sample. The assessment of results shows that two compounds (i.e. GTX1-4 and dcSXT) are not detected in any of the three samples and two other compounds (i.e. GTX2-3 and STX) are not detected in the third sample. The laboratory suspects of false negatives but did not find conclusive explanation for these results. It did not carry out any further investigation (it stated that this biotoxin is very unusual to find so they did not do any further investigation; in addition, there was no sufficient remaining sample for re-testing). · The lack of testing of certain compounds and the underreporting of other compounds poses a risk of underestimation of PSP if such an episode occurs.

Regarding lipophilic toxin laboratories (findings are limited to the assessment of the laboratory procedures and the last participation in a proficiency testing of the laboratory carrying out the analyses of official and the harvesters' samples since 1 January 2021): · This laboratory procedure is not in line with the last version of the EURL (which is dated in 2015) and it deviates from the EURL procedure in the extraction of the lipophilic toxin compounds. · The laboratory does not detect all the relevant compounds of the lipophilic toxin group. · As mentioned before, biotoxin testing is carried out with heat-treated bivalve molluscs. This deviates from the EU reference method for lipophilic toxins, therefore it may have an impact on the testing results. · The proficiency testing results shows unsatisfactory and questionable results for some of the compounds of this group. The laboratory suspects there is a problem with the procedure of extraction. If it is the case, there is a risk of underestimation of the lipophilic toxin group if such an episode occurs (the lipophilic toxin group is the most common biotoxin in Danish waters). Conclusions on laboratories

The fact that the laboratory for PSP cannot analyse part of the compounds of this group, that the results in proficiency testing show that it cannot detect a number of other compounds and laboratories do not find conclusive explanations for unsatisfactory and questionable results in proficiency tests, indicates a risk of underestimation of this biotoxin.

The results in proficiency testing show that the laboratory for lipophilic toxins cannot detect a number of other compounds, and the lack of conclusive explanations for questionable and unsatisfactory results in participation proficiency testing weakens the reliability of lipophilic toxins testing results.

Medlemsstatens svar

ivalve The designated NRL will work to resolve the issues with the laboratories. h The DVFA expects this process to complete within mid-2022.

Rekommendationen anger som grund

Two laboratories share the role of National Reference Laboratory (NRL) for E. coli, the DVFA and the DTU laboratories. The DVFA does not consider necessary to designate the DTU as NRL, which is not in line with Article 100(1) of Regulation (EU) 2017/625. The DVFA provided to the audit team the service contract that was in place between the DVFA and the DTU for the period 2017-2020, which included the list of tasks to be done. It also provided the list of NRLs (based on repealed EU legislation, Regulation (EC) No 882/2004), indicating the responsibilities of each one6. However, tasks required analysis for the parameters E. coli, biotoxins, and potentially toxic algae. 4 in Article 101(1)(b) and (c) of Regulation (EU) 2017/625 were not included, and the DVFA confirmed that neither of these laboratories carried them out.

Contrary to Article 100 of Regulation (EU) 2017/625, there is no NRL for biotoxins in Denmark. The DVFA explained that the DTU laboratory was the NRL until the end of 2020; the new one, the DVFA laboratory, was in the process of setting up the biotoxins methods. 5.1.1 Classification of production areas

The absence of an NRL for biotoxins and the fact that one of the NRLs for E. coli is not designated and that they do not fulfil all their responsibilities and tasks, is an important shortcoming that diminishes the reliability of results in an official system that is based on results from (a number of) private laboratories.

As described in paragraph 8, and contrary to Article 100 of Regulation (EU) 2017/625, there is no NRL for biotoxins. The NRL's tasks established in Article 101(1) of Regulation (EU) 2017/625 are not carried out. Taking into account the results in 16 participation in proficiency testing (described in paragraph 45 and 46 below), the implementation of Article 101(c) of Regulation (EU) 2017/625, regarding the appropriate follow-up of these tests by an NRL is of utmost importance.

The absence of a NRL for biotoxins and the existence of NRLs for E coli, which do not fulfil all the NRL responsibilities and tasks, negatively affect the harmonisation and improvement in the performance of testing methods by other official and private laboratories.

Medlemsstatens svar

100 The DVFAs laboratory, * has ­ through the main in agreement between DVFA and DTU -been designated as NRL for marine biotoxins by the 1st of September 2021. In line with the above mentioned, the DVFA laboratory, * , is also designated as NRL for the parameter E. coli as a shared responsibility with the DTU. Further, the DVFA laboratory performs an official analysis for the parameter E. coli. The designation of official laboratories implies that each laboratory fulfils the criteria described in the general accreditation standard ISO 17025 ­ including the obligation to inform the accreditation bodies in case of non-compliances regarding proficiency test or other deviation significant for the analytical result of a specified parameter. The strategy for DVFA in the area of laboratory quality is, therefore, that the accreditation system ­ including the third part inspection by the accreditation bodies - is sufficient to ensure the quality of the analytical results to be used as a part of the official activities mentioned in Regulation 2017/625. to respect the provisions of Article 339 of the Treaty on the functioning of the Page: 8

Rekommendationen anger som grund

The procedure applied in one seaport BIP in case of direct movement from the BIP to ship supply or to further storage facilities - one CVED for several consignments (and several veterinary certificates) in one means of transport - is not in line with the requirements of Directive 97/78/EC. On arrival at the BIP, consignments accompanied with several certificates, transported in the same means of transport were presented at the BIP by the operator with one CVED only. Consequently, not all certificates and not all commodities were mentioned on the CVED and thereafter recorded in TRACES.

The implementation of the import controls on animals, food and feed of animal origin and of goods subject to special import conditions is generally in line with EU legislation. The competent authorities ensure that documentary, identity and physical checks on animals and goods are carried out as required and that only compliant goods may be imported into the EU. Regular checks of airplane/shipping manifests contribute to ensuring that consignments requiring veterinary import controls do not escape such controls. Some weaknesses were noted in the documentary checks on re- imported consignments (for which the DVFA initiated corrective action) and those intended for ship supply, which weakens the requisite health guarantees and the traceability of consignments, respectively.

Medlemsstatens svar

With reference to the comments provided by the DVFA to finding icate number 40, the DVFA will stress to importers and the official ments inspectors at the BIPs, the importance of recording all health certificate numbers and commodities in the CVED. Furthermore, the DVFA will focus on ensuring receipt of confirmation from the captain of the ship within the time limit. Procedures for contacting the Customs authority in cases of missing reply will be added to the national guideline for control of ship supply. Page: 1

Rekommendationen anger som grund

At the DVFA local branches visited and during field visits, the audit team interviewed the inspectors on the criteria for the selection of targeted samples. The inspectors met stated that the choice of the farms to be sampled was mainly based on the farm profile and the size and even information on the use of veterinary medicinal products available via VetStat could be consulted. In many cases, the inspectors stated that the samples on the farms would be taken randomly thus underestimating the importance of targeted sampling, which is not in line with the requirements laid down in points 2.1 and 2.3.2.1 of the Annex to Decision 98/179/EC). This situation was also observed during the 2010 audit and led to a recommendation. The audit team noted that the DVFA procedure did contain adequate guidance on the criteria to be used for the selection of targeted samples and that the procedure was accessible to the inspectors (see finding 12), though not all of the inspectors interviewed by the audit team were aware of the procedure.

The residue monitoring plan is generally implemented in accordance with planned arrangements. However, its effectiveness is weakened by untargeted sampling on farms, sample clustering and the lack of tamper proof sealing of official samples. 5.1.4. Other residue monitoring programmes Legal Requirements Articles 9 and 11 of Directive 96/23/EC; Articles 8 and 10 of Regulation (EC) No 882/2004. Findings

Medlemsstatens svar

Competent authority response received on 11 July 2019: The risk criteria for targeted sampling and the financ-ing of the extra cost will be decided in a working group. The DVFA will implement targeted sampling of animals on farms from the year 2020.

Competent authority response received on 11 July 2019: The risk criteria for targeted sampling and the financ-ing of the extra cost will be decided in a working group. The DVFA will implement targeted sampling of animals on farms from the year 2020.

Rekommendationen anger som grund

The audit team noted that the same establishments (e.g. farms and slaughterhouses) had been sampled for several substance groups on the same day, in some cases even for the same substances. In some cases, this was because the number of samples to be taken was higher than the number of establishments to be sampled. However, in other cases, although the number of samples was limited and much smaller that the number of establishments to sample, several samples were taken from the same establishment. This contradicts the DVFA's own sampling procedure and is not in line with the sampling strategy and targeting criteria laid down in point 2.1, 2.2, 2.3.2 and 2.3.3 of the Annex to Decision 98/179/EC.

The residue monitoring plan is generally implemented in accordance with planned arrangements. However, its effectiveness is weakened by untargeted sampling on farms, sample clustering and the lack of tamper proof sealing of official samples. 5.1.4. Other residue monitoring programmes Legal Requirements Articles 9 and 11 of Directive 96/23/EC; Articles 8 and 10 of Regulation (EC) No 882/2004. Findings

Medlemsstatens svar

Competent authority response received on 11 July 2019: The samples in all sub-groups in directive 96/23 in-cluding metals will be distributed in accordance with the production figures from each slaughterhouse. The procedure will be established from the start of 2020.

Competent authority response received on 11 July 2019: The samples in all sub-groups in directive 96/23 in-cluding metals will be distributed in accordance with the production figures from each slaughterhouse. The procedure will be established from the start of 2020.

Rekommendationen anger som grund

Although instructions and documents seen by the audit team unambiguously require official samples to be sealed, the audit team observed a number of cases where no official sealing had been performed, or performed in a way not ensuring the custody chain and such samples (e.g. from small slaughterhouses) had been kept/transported by a third party service provider. Moreover, during reception of the samples by the laboratories no sufficient control of the seals was carried out and non-sealed samples were accepted and processed. Thus, the requirements laid down in point 2.6 of the Annex to Decision 98/179/EC are not met7. This shortcoming was also observed during the 2010 audit.

The residue monitoring plan is generally implemented in accordance with planned arrangements. However, its effectiveness is weakened by untargeted sampling on farms, sample clustering and the lack of tamper proof sealing of official samples. 5.1.4. Other residue monitoring programmes Legal Requirements Articles 9 and 11 of Directive 96/23/EC; Articles 8 and 10 of Regulation (EC) No 882/2004. Findings

Medlemsstatens svar

ompetent authority response received on 11 July 2019: To ensure an adequate check of sample seals, the la-boratory will implement a procedure for sample re-ception in the laboratory quality manual describing a thorough inspection of sample seal condition.

ompetent authority response received on 11 July 2019: To ensure an adequate check of sample seals, the la-boratory will implement a procedure for sample re-ception in the laboratory quality manual describing a thorough inspection of sample seal condition. In September, The DVFA reminded all personnel involved in sampling to adhere strictly to the instructions present in the sample-project-manuals and in the laboratory quality-manual regarding the sealing of official samples. Furthermore, The laboratory updated the instructions regarding sealing of official samples and it was emphasized that samples must be sealed and that samples will be discarded if the samples are not properly sealed. Page: 1 k to the recommendations of report ref. DG(SANTE)/2019-6650-MR of the audit carried out from the control of residues and contaminants in live animals and animal products including controls on ion Proposed by the competent authority ling enforcement will take effect from ember 1st 2019. To ensure an adequate check of sample seals, in addition the oratory will implement a procedure for sample reception in the laboratory lity-manual describing a thorough inspection of sample seal condition. If

Rekommendationen anger som grund

The audit team visited all three laboratories and noted that: a. All the laboratories were accredited to ISO 17025:2005 by the Danish Accreditation Fund (DANAK); b. The laboratories' facilities, equipment and staffing levels were adequate and the staff met were competent; c. A Standard Operating Procedure (SOP) for reception of samples was in place, the criteria for accepting samples were recorded, and a procedure for sample rejection was available and followed; 7 d. The sensitivity of the analytical methods in use was sufficient to ensure the detection and quantification of residues of pharmacologically active substances at the Maximum Residue Limits (MRLs) listed in Table 1 of the Annex to Regulation (EU) No 37/2010, Minimum Required Performance Levels laid down in Article 4 and Annex II to Decision 2002/657/EC or, for unauthorised substances, the EURL recommended concentrations; e. At the DVFA laboratory in Ringsted and the DTU laboratory, the general SOP on validation of chemical methods of analysis included the assessment of all method performance parameters as required in Article 3 of Decision 2002/657/EC, with the exception of analyte stability. Whilst data on analyte stability in stock solutions were available, information regarding stability of analytes in relevant matrices had never been collected and evaluated as required in point 3.1.1.4 of the Annex to Decision 2002/657/EC. Reasonably short and largely adhered to turnaround times of 5 working days for antibiotics and 20 working days for the majority of pharmacologically active substances mitigate the possible effect of analyte instability in different matrices. f. Control charts (run for several parameters) were maintained and regularly checked (apart from the DVFA laboratory in Århus, see finding 34(b)) for long term trends in method performance. This fully satisfies the requirements of Article 5 of Decision 2002/657/EC. g. The laboratories regularly participated in relevant proficiency tests. Most of the results seen were satisfactory. When questionable results were obtained, adequate root-cause-analysis was carried out and appropriate corrective measures have been implemented.

Notwithstanding some minor shortcomings in method validation, the competent authority can have confidence in the laboratory performance and the reliability of analytical results.

Medlemsstatens svar

ompetent authority response received on 11 July 2019: In the Laboratory in Ringsted, the SOP for validation of methods (KVI- 03.0090) has been published in a new revision. The SOP now includes a modified par-agraph concerning stability. The former paragraph concerning Page: 1 k to the recommendations of report ref. DG(SANTE)/2019-6650-MR of the audit carried out from the control of residues and contaminants in live animals and animal products including controls on roposed by the competent authority y is updated and it is clearly mentioned that both the stability in solution atrix should be evaluated during the validation of a method. t authority response received on 16 August 2019: ted SOP for validation of methods (KVI-03.0090) was submitted. Page: 2

ompetent authority response received on 11 July 2019: the Laboratory in Ringsted, the SOP for validation of methods (KVI-03.0090) has been published in a new revision. The SOP now includes a modified par-agraph cerning stability. The former paragraph concerning stability is updated and it is arly mentioned that both the stability in solution and in matrix should be luated during the validation of a method. Page: 2 k to the recommendations of report ref. DG(SANTE)/2019-6650-MR of the audit carried out from the control of residues and contaminants in live animals and animal products including controls on ion Proposed by the competent authority petent authority response received on 16 August 2019: updated SOP for validation of methods (KVI-03.0090) was submitted. Page: 3

Rekommendationen anger som grund

At the time of the audit, the CA and CBs were in the process of defining the respective conditions and responsibilities which will eventually be part of the control plan. However, the control plan was in the early stages of preparation and the concept of supervision of the CBs by the CA had not been entirely worked out by the CA. Consequently, supervision of CBs by the CA was not yet in place. During inspections observed (see chapter 5.2.4), the audit team was informed by a CB inspector that the correct approach to controlling certain elements of the Product Specification was not 4 clear for the inspector. This lack of supervision by the CA was giving rise to uncertainty on what should be included in the official controls by CBs and risks non-compliant PGI being placed on the market. This is not in line with Article 5(2)(f) of Regulation (EC) No 882/2004.

There was evidence of lack of coordination between the CA, the CBs and Producers/Processors. As an example, during one site visit, a CB had not issued an inspection report for the most recent official control undertaken prior to this audit. According to the CB inspector this lack of reporting was due to unresolved contractual issues which prevented her from completing the official control. The CA contradicted the inspector. Nevertheless, the producer continued to sell the PGI using the protected name and Union symbol. It became clear to the audit team, that none of the key players were fully aware that this non-compliant situation existed until the time of the audit. This is not in line with Article 5(2)(f) of Regulation (EC) No 882/2004 which requires that there is efficient and effective coordination between the delegating CA and the CB. Conclusions on Control Bodies

At the time of the audit, supervision of CBs by CAs was not implemented and there was a lack of coordination between the key players involved in official controls. 5.2.3 Prioritisation of Official Controls Legal requirements Articles 37(1) and 38 of Regulation (EU) No 1151/2012. Article 3 of Regulation (EC) No 882/2004. Findings

At the time of the audit, a second CB had not issued any report for the official control of one PGI. Nevertheless, the producer continued to sell the PGI using the protected name and Union symbol. The CB inspector stated that a report had not been issued as there 6 was no written agreement with the Producer group in place at the time of the audit. The CB stated that a verbal agreement was in place and legally this was considered as a contract. The CA had granted the CB an approval to operate on the understanding that there was an agreement (whether verbal or written) between the Producer group and the CB. This situation highlighted a lack of coordination between the key players involved (see also paragraph 14). (b) Performance of Official Control Activities at Visited Premises

However, the following weaknesses were noted: Although the inspectors noted some labelling non-compliances, the full extent of the labelling requirements (Article 12(3) and 23(2) of Regulation (EU) No 1151/2012) were not fully implemented at the time of the audit. (see further details under this chapter, point (c) Labelling ­ correct use of Union symbol and presentation of the protected name). There was no verification of the of input/output calculations performed by the operator. This is not in line with Article 4 of Regulation (EU) No 668/2014.

The CB inspector met at one premises queried the exact meaning of some elements of the Product Specification. The CB inspector raised issues1 in relation to carrying out official controls in an effective manner. The preparation of the control plan by this CB had identified these issues as a challenge that needed clarification from the CA. (c) Labelling ­ correct use of Union symbol and presentation of the protected name

During the audit, there were numerous instances of non-compliances relating to the proper use of the Union symbol and the protected name. The ensuing discussion between the CB, CA, the producer and the audit team illustrated the challenge associated with labelling and the different interpretations that exists. It became clear that there is a need of guidance on labelling to be available to producers and CBs.

One of the inspections observed was the first official control of a particular PGI which constituted the verification of the Product Specification prior to placing the product on the market. In this instance, the operator had a model of the labelling intended to be used on the final product. The inspector briefly reviewed this information, but stressed that it was not the inspector's responsibility to approve labelling. In fact the labelling did not fully comply with the minimum requirements of the legislation, and in particular Article 12(3) of Regulation (EU) No 1151/2014. 1 In their response to the draft report the CA noted that issues relating to the exact geographical location had been resolved after completion of the audit. 7

It was noted by the audit team that the Union symbol and the protected name were not always sufficiently close together to clearly link these two elements in the same field of vision. There were also examples of the protected name being misspelt, the Union symbol being less than 15 mm as well as other descriptive information contained in the Product Specification being incorrectly presented on the label. The CB inspectors observed did not detect all of the labelling non-compliances. Conclusions on Procedures and Performances of Official Control Activities

Although the CB inspectors were knowledgeable, there was a lack of guidance relating to the interpretation of the PDO/PGI/TSG legislation and its application in practice. This resulted in a number of deficiencies being overlooked by the inspectors. 5.2.5 Enforcement Measures Legal Requirements Article 13(3) of Regulation (EU) No 1151/2012. Articles 54 and 55 of Regulation (EC) No 882/2004. Findings

Medlemsstatens svar

In order to ensure appropriate coordination between CA and CBs and to ensure sufficient coordination with other key players, the DVFA will: Assess the control plans submitted by the CBs in order to ensure that the control plans are fully in line with the requirements. Time frame: Autumn 2018 Establish a plan for audit of the CBs. Time frame: o Plan: June 2018 o End of audits: December 2018 Ensure coordination between the key players by arranging meetings between the key players Time frame: Autumn 2018

Rekommendationen anger som grund

The audit team noted several discrepancies between the activities indicated on the list of approved food establishments, the activities mentioned in the approval certificates and the activities actually carried out in the establishments visited.

In some of the establishments visited the audit team noted that: the establishments had not been approved for all the activities which were actually carried out (e.g. fishery products establishment producing fried chicken skin was not approved as a poultry meat products establishment; a catering establishment cooking fish (among other activities) was not approved as a fishery products establishment). the establishments were listed under several activities which were not being carried out, or for which they have not been approved. one of the establishments visited was approved as a meat products establishment, but was not on the published list of approved food establishments.

These findings are not in compliance with the requirements of Article 31(2)(c) and (e) of Regulation (EC) No 882/2004 which requires that the CA shall approve an establishment for the activities concerned and shall keep the approval under review when carrying out official controls.

Although the CA inspectors are required to verify the maintenance of approval conditions at least annually during inspection and the inspectors should update the relevant information in the KOR system (Control Objects Register ­the main database used by the CA for individual risk characterisation of each establishment), the audit team noted that it is not consistently carried out. As the published list of approved establishments obtains the information from KOR, the list is not up to date. Conclusion on registration/approval of food business operators

Procedures are in place for registration and approval of establishments handling or producing RTE food. However, there are shortcomings in their implementation which result in the list of approved establishments not being up-to-date and approval of establishments does not accurately reflect the actual activities carried out. 9

Medlemsstatens svar

pproval Guide" is in process to be updated on a number of topics. The updated version will be published in February 2019. levant to this recommendation are the following points: d The description of significant changes will emphasise that if a change in activities require a change in good hygiene practise of relevance to cross contamination or the HACCP-based procedures it is to be considered a significant change that must be reported to the DVFA. It will be emphasised in the guide that the list of examples of significant changes is not an exhaustive list. The inspectors are instructed to ask approved establishments for changes in activities ­ new activities as well as activities that are no longer relevant ­ at an inspection once a year ­ all approved establishments are inspected one or more times every year. This will be described in the Inspection Frequency Guidelines applicable from 1 January 2019. Furthermore, the inspectors must react if they observe activities that may be changed. The inspectors will be instructed only to approve establishments according to what type of product is delivered from the establishment, e.g. not to approve an establishment as cutting plant if it o

Medlemsstatens svar

The guideline for animal welfare controls in pig herds and/or the guideline on enrichment material will be updated to include liance criteria for air quality, feeding space, and drinkers. It will at the same time be considered, whether an update on the compliance criteria concerning mixing of pigs, which are already in the guideline, is appropriate. This update is expected to be finalised during the first half of 2018. An update of the guideline on animal welfare controls in pig herds to include compliance criteria for staff will have to wait the adoption of an order, which requires that the person responsible for a pig herd of a certain size shall have relevant training. This update is expected to be finalised during the second half of 2018.

Rekommendationen anger som grund

The CA has designed a template for holding register for sheep to be used by the farmer. Not all data required by EU legislation are included in the template, e.g. the result and date of last inventory, the name of transporter, the registration number of the vehicle or date of identification of new-born animals. On the sheep farm visited by the audit team those data (with exception of the result and date of last inventory) were not recorded by the keeper. This is not in line with Article 5 of Commission Regulation (EC) No 21/2004. 5.1.3 Correspondence with reality

On a sheep farm visited, all animals, including lambs for slaughter were tagged with two earmarks, one of them electronic. The flock register was updated in a timely manner by the owner. But not all data required by the EU legislation were included (for more see point 22). The sheep off movements were notified in a timely manner to the CHR (as batch movement). The result of the latest animal inventory was updated into the CHR in January 2016, but the number of animals recorded was significantly lower than the number of animals present on the farm at the time of audit team visit. The reason provided for the discrepancy was a mistake made by farmer's consultant when updating the CHR. 17 Conclusions on data quality

The incomplete official template for flock register on sheep farms has no big impact in the day-to-day business (small sheep population) but could delay traceability in the case of emergencies (e.g. Foot-and-Mouth disease outbreak).

Rekommendationen anger som grund

At the assembly centre visited, the audit team noted shortcomings related to the certification procedures, in particular that the official veterinarian accepted incomplete information or unsigned statements by the owners of the animals. Furthermore, the farmers' statements that animals stayed for 30 days in their holding (for pigs) and no 20 animals were imported during the same period (for pigs and bovines) were directly accepted as the basis for certification without any additional verification (e.g. validated against data in the CHR)1. This is not in line with point 2, Article 3 of Directive 96/93/EC. The same shortcomings were detected during an internal audit in 2012, but no actions have been implemented yet by the CA to address them. 5.2.2 Fitness for purpose

The central database is only partially used for certification purposes; instead the certification at assembly centres relies on a system of pre-certification statements, some of which are not verified as should be. As a result, the existing weaknesses negatively affect reliability of the certification process for intra-EU trade.

Rekommendationen anger som grund

The laboratory has adequate analytical scope and is accredited for the use of most of analytical methods, except for the screening tests. 5.3.8 Enforcement measures Legal requirements Articles 4(5) and 33 of Directive 2001/18/EC. Article 45 of Regulation (EC) No 1829/2003. Article 11 of Regulation (EC) No 1830/2003. Findings

Medlemsstatens svar

DAFA will continue to select crops and seed lots for official seed trols on a risk basis, ie. to focus the control on conventional crops and seed lots that have an increased risk of unintentional presence of GMOs. DAFA will extend the control to seed companies of all sizes. In 2017, particular emphasis will be put on seed control in smaller seed companies.

Medlemsstatens svar

To ensure that the audit universes of DAFA and DVFA cover all activities which the CA´s are responsible for, an agreed "DAFA- DVFA audit universe - document" will be produced. The document will be finished in 2017 and at least updated on an annual basis. At least one annual meeting between DAFA and DVFA´s internal audit units will be held, and one topic will be "sharing our audit plans and audit experiences" - especially in the areas that are covered by cooperation agreements. Page: 1

Rekommendationen anger som grund

There are eight DPEs in Denmark, six of which are local FCOs which are approved only for documentary control and they are not located at the specific points of entry in the Member State's territory which is in contradiction with the definition provided in Article 3 of Regulation (EC) No 669/2009.

In the list there are two private food and feed business operator facilities approved as DPEs. They are located in the port territory and they accept to store consignments for identity and physical checks, including laboratory analyses from other FBOs. At the same time they are approved as DPIs as well. However, documentary checks for these consignments are carried out at local FCOs instead of DPEs.

DPEs are not designated or used for documentary control in line with the legal requirements.

Medlemsstatens svar

Within three months the documentary checks will take place at points of entry for 669/2009 consignments.

Within three months the documentary checks will take place at points of entry for 669/2009 consignments. Updated action plan received on 15 May 2017 Before 1 January 2018 DPEs in line with Regulation (EC) No 669/2009, Article 3 (b) cf. Article 19 will be designated.

Rekommendationen anger som grund

Regarding both products of animal origin and products of non-animal origin, if the result shows a non-compliance, the consignment is rejected and can be destroyed or re- dispatched. In the case of re-dispatch, the DVFA requires confirmation from the country of destination that they agree to receive non-compliant consignments as required by Article 21 of Regulation (EC) No 882/2004. A 60-day deadline is given to food business operators to destroy or re-dispatch non-compliant consignments. The audit team noted that where the consignment of products of non-animal origin is re-dispatched, the importer is not required to submit evidence to ensure that the 60-day limit is respected as foreseen in Article 21(2) of Regulation 882/2004. The DVFA stated that in the case of destruction, the importer submits proof of the destruction of the rejected consignment in accordance with Article 20(2) of Regulation 882/2004.

Where a rejected consignment of products of non-animal origin is re-dispatched, the operator is not requested to submit any evidence that the consignment left Denmark for an approved destination. This does not ensure that the deadline established in EU legislation has been respected and opens up the possibility that the consignment may have been illegally placed on the EU market. 13

Medlemsstatens svar

he The orders have been revised to include the requirement to present h the documentation to the CA for re-dispatch or destruction of rejected consignments of products of non-animal origin. The orders come into force by January 2017.

he The orders have been revised to include the requirement to present h the documentation to the CA for re-dispatch or destruction of rejected consignments of products of non-animal origin. The orders come into force by January 2017.

Rekommendationen anger som grund

There is an adequate system in place for the registration of operators and overall it is effectively implemented. The CAs publish and keep an updated list of operators, in line with EU legislation. However, DAFA does not publish the documentary evidence and the documentary evidence issued by DVFA to the operators does not follow the model of Annex XII of Regulation (EC) No 899/2008.

Medlemsstatens svar

DVFA has already changed the documentary evidence so that the es are numbered in accordance with Annex XII and furthermore now includes box 7 referring to the "date of control(s)". A reference to www.findsmiley.dk has been inserted. On this site, which is continuously updated, the inspection dates as well as specific inspection reports from the 4 most recent inspections at each food company can be found. Please see the attached example which will be used for the future issued documentary evidences. The DAFA publishes a list of names of all organic farms and supplying companies on the internet. This list will be expanded in order to give the public access to Annex XII via a link.

DVFA has already changed the documentary evidence so that the boxes are numbered in accordance with Annex XII and furthermore now includes box 7 referring to the "date of control(s)". A reference to www.findsmiley.dk has been inserted. On this site, which is continuously updated, the inspection dates as well as specific inspection reports from the 4 most recent inspections at each food company can be found. Please see the attached example which will be used for the future issued documentary evidences. The DAFA publishes a list of names of all organic farms and supplying companies on the internet. This list will be expanded in order to give the public access to Annex XII via a link. Documentary evidence in accordance with Annex XII for organic farms and supplying ns (updated) received 12/02/2016 tent authorities of Denmark on N/A in response to Report ref. DG(SANTE)/2015-7358-MR of the der to evaluate the control systems for organic production and labelling of organic products companies is planned for July 2016.

Rekommendationen anger som grund

Controls on organic operators are in most cases annually carried out in line with Article 65(1) of Regulation (EC) No 889/2008. However, the frequency of control on a defined group of small farmers that do not place products on the market is not in line with the above-mentioned article.

Medlemsstatens svar

All organic operators will be inspected at least once a year in 2016 and onward. The approximately 100 organic operators which have had reduced control, are now active in DAFA IT control system and will be controlled once a year.

t All organic operators will be inspected at least once a year in 2016 and onward. The approximately 100 organic operators which : have had reduced control, are now active in DAFA IT control system and will be controlled once a year. The 100 organic operators have received a letter informing them that they will be subject to yearly control as of 1 January 2016. Please see Annex 1

Rekommendationen anger som grund

Additional visits are primarily unannounced and based in most cases on risk assessment as established in Article 65(4) of Regulation (EC) No 889/2008. However, the risk assessment carried out to select additional visits to organic farms is not in line with Article 65(4) of Regulation (EC) No 889/2008. 5.2.4 Controls of Operators Legal Requirements Titles II and III of Regulation (EC) No 834/2007. 9 Titles II and IV and Article 65(1) and 65(3) of Regulation (EC) No 889/2008. Articles 8 and 9 of Regulation (EC) No 882/2004. Findings

Medlemsstatens svar

nic DAFA has changed the procedure for additional visits. Operators will be selected at random, based on an overall assessment of the risk of non-compliance with the rules of organic production, where the results of previous inspections, the quantity of the products and the risk for exchange of products are taken into account.

DAFA has changed the procedure for additional visits. Operators will be selected at random, based on an overall assessment of the risk of non-compliance with the rules of : organic production, where the results of previous inspections, the quantity of the products and the risk for exchange of products are taken into account. Please see Annex 2-3

Rekommendationen anger som grund

Both DVFA and DAFA had planned for 2014 an adequate number of samples to be taken. As a general rule, sampling is risk based. However, the selection of operators (which does not include farmers) by DAFA is not based on the general evaluation of risk of non-compliance with the organic production rules at all stages of production, preparation and distribution, which is not in line with Article 65(2) of Regulation (EC) No 889/2008.

Medlemsstatens svar

e DAFA will sample organic farmers in accordance with Article 65 (2) from the beginning of 2016.We will implement the change of sampling on farms in all stages of production, processing and Page: 1 15 port ref. DG(SANTE)/2015-7358-MR of the audit carried out from 26 May duction and labelling of organic products distribution by changing the instructions for the control and thereby expand the sampling from companies to include farms. Furthermore, we will set up criteria for risk evaluation of non- compliance with organic production rules on farms and companies in order to identify which samples preformed.

e DAFA will sample organic farmers in ns (updated) received 12/02/2016 tent authorities of Denmark on N/A in response to Report ref. DG(SANTE)/2015-7358-MR of the der to evaluate the control systems for organic production and labelling of organic products accordance with Article 65 (2) from the beginning of 2016.We will implement the ed change of sampling on farms in all stages of production, processing and distribution by changing the instructions for the control and thereby expand the sampling from companies to include farms. Furthermore, we will set up criteria for risk evaluation of non-compliance with organic production rules on farms and companies in order to identify which samples preformed. Please see Annex 4-6

Rekommendationen anger som grund

Generally, derogations granted comply with specific EU requirements. However, dehorning is carried out routinely and without prior authorisation by the CA which is not in line with Article 18 of Regulation (EC) No 899/2008. 5.2.8 Imports of Products from Organic Production Legal Requirements Articles 32 and 33 of Regulation (EC) No 834/2007. Commission Regulation (EC) No 1235/2008. In particular, Article 7, Chapter 3 of Title III, Article 19 and Annexes II, III, V and VI. Findings

Medlemsstatens svar

8 DAFA expects to receive about 3000 applications for dehorning per year. Therefore, we have started a working group to determine the extent of the task of developing a newIT system, where we can handle the applications on a case by case basis. We plan to develop a solution which is integrated with our Key self-service CAPtas. The farmer shall send a digital application for dehorning to us. We will handle the applications on a case by case basis in Captas and reply to the farmer digitally Page: 2

DAFA expects to receive about 3000 applications for dehorning per year. ior Therefore, we have started a working group to determine the extent of the task of developing a new IT system, where we can handle the applications on a case by case basis. We plan to develop a solution which is integrated with our Key self-service CAPtas. The farmer shall send a digital application for dehorning to us. We will handle the applications on a case by case basis in Captas and reply to the farmer digitally. We have not started the IT development in ns (updated) received 12/02/2016 tent authorities of Denmark on N/A in response to Report ref. DG(SANTE)/2015-7358-MR of the der to evaluate the control systems for organic production and labelling of organic products 2016 because of shortage of it funds. The development will be prioritised in 2017. Consequently the system will be ready for use in February 2018.

Medlemsstatens svar

is For 2016 it has been ensured that numbers of samples at the slaughterhouses is distributed according to throughput and evenly al distributed over each month of the year. To ensure the batches are sampled randomly, all slaughterhouses involved in sampling have .3 been provided with a randomized list of batches to be sampled based on the numbers of batches planned for slaughtering. CA has ensured that sampling has started from the beginning of January and stated in the project description that all month shall be represented in the sampling. Please see attached project descriptions

Rekommendationen anger som grund

Staff performing official controls in sprouting establishments and at primary producers of FNAO did not receive specific training in this area and were not adequately aware of any potential source of contamination at farm level. 5.2.2 Registration / Approval of Establishments Legal Requirements Article 6 of Regulation (EC) No 852/204, Article 2 of Regulation (EU) No 210/2013 and Article 31 of Regulation (EC) No 882/2004 Findings 5.2.2.1 Registration of Food Establishments

Medlemsstatens svar

udit Primary production excluding sprouting: C) · Written instructions and check-lists will be adjusted according to this audit. · A training course for staff at the DAFA will be held March or April 2016, regarding controls at primary production of food of non-animal origin (FNAO). Primary production in the form of sprouting and import, packaging etc. of seeds for sprouting: · A training course for staff of the DVFA involved in official controls of this type of activities will be held in the first half of 2016. · Two courses on Imports control are scheduled for the Autumn 2015.

Rekommendationen anger som grund

Procedures are in place for the registration of establishments. However, the retailers of seeds for sprouting are registered only if their activities regarding food are considered as a major part of their business. Registration of primary producers of FNAO does not include all relevant information to allow CAs to perform risk based official controls effectively.

Official food hygiene controls relating to microbial contamination of FNAO are carried out. There is an effective risk based system of official controls in place for sprouting establishments and the trade chain for seeds for sprouting. However, for other type of primary production operators, the lack of use of the information by the CA limits the ability to implement a risk based system of official controls to prevent microbiological contamination of FNAO. 5.2.4 Implementation of Official Controls over Food Business Operator's Obligations Legal Requirements Article 10(2)(b)(i), Article 10(2)(c), Article 10(2)(d) and Article 15(1) of Regulation (EC) No 882/2004, Article 4(1), Article 7, Article 8 and Part B of Annex I of Regulation (EC) No 852/2004, Article 18 of Regulation (EC) No 178/2002, Article 3 of Regulation (EU) No 208/2013, Regulation (EU) No 211/2013 and Annex I of Regulation (EC) No 2073/2005 Findings Cultivation

Medlemsstatens svar

When the primary producer of FNAO register, the registration rmed form has been changed so that during registration he can specify whether he produce "fruit and vegetable" except potatoes or "other vegetable" incl. potatoes. The changes can be viewed in the attached file (screen print from the DVFA website). We are looking into the possibility of using information from DAFA from applicants of EU funding regarding type and volume of production. This information would then be included when selecting producers of FNAO for official control.

Rekommendationen anger som grund

The CA did not ensure that approved sprouting establishments fully comply with the requirements of Regulation (EC) No 210/2013. 7 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

Medlemsstatens svar

The approved sprouts producing establishments have finished ongoing changes regarding infrastructure and equipment. The Page: 1 09/2015 Report ref. DG(SANTE)/2015-7454-MR of the audit carried out from 14 April y production of food of non-animal origin establishments regularly revise their procedures to ensure the protection against cross contamination. It is part of the official controls to control that the infrastructure, equipment and flow of products and personal in the establishments are in accordance with the requirements and ensure food safety. The DVFA is working on specific guidelines to be used during official control visits prior to approval of food business operators. The guidelines are expected to be finished in the first half of 2016.

Rekommendationen anger som grund

A general guide to GAP is available to primary producers. However, its dissemination and use was not encouraged by the CAs.

Medlemsstatens svar

DVFA will hold a training course for staff at DAFA in March or April 2016, regarding controls at primary production of food of non-animal origin (FNAO). The use of the Danish GAP guide will be included. DVFA has in mid-September 2015 sent an e­mail to The Danish Agriculture & Food Council about cooperation and ideas of how to re-inform the primary producers about the GAP guide in a good way.

Rekommendationen anger som grund

No major deficiencies were noted in visited establishments. All observed inspections correctly covered traceability, labelling requirements and cleanliness of establishment.

The official controls at farm level did not cover all food hygiene requirements to prevent microbiological contamination of FNAO and some were not assessed correctly.

The requirements on obligatory testing of seeds for sprouting and sprouts, laboratory results, personal hygiene and equipment were not assessed correctly by official controls.

Medlemsstatens svar

iven in Primary production: From 2016 a section will be added in the instruction manual for controls on FNAO, concerning open field control and how to review information on water sources used for irrigation. This information will also be explained at the course mentioned under rec. 4.

Rekommendationen anger som grund

There is in place system for follow-up of RASFF notifications however, the RASFF notifications concerning microbiological contamination of produce of Danish origin are not followed through to the producer level except for sprouting establishments.

Medlemsstatens svar

r of According to chapter 4 in the current cooperation agreement between DAFA and DVFA from December 2014, it is stressed that a mutual obligation to provide information and on the exchange of information of relevance to both parties exists. This includes the obligations mentioned in Article 2 of Regulation Page: 2 09/2015 Report ref. DG(SANTE)/2015-7454-MR of the audit carried out from 14 April y production of food of non-animal origin (EU) No 16/2011 concerning RASFF.

Rekommendationen anger som grund

According to national procedure, FBOs do not have to document good manufacturing practices (GMPs) if they are not linked to food safety. Official assessment of GMPs, as required by Article 10.2.d of Regulation (EC) No 882/2004, is restricted to the explanation of the modus operandi by FBOs to official inspectors. The audit team noted that in one of the establishments visited testing of ready-to-eat products was not documented. This lack of documentation had not been acted on by officials during previous control visits4.

In some cases, deficiencies noted by the audit team during the visits had already been identified during official controls or were detected by official inspectors during the Commission services visit (e.g. shortcomings on traceability or lack of temperature records). However, this was not always the case5, for example: Regarding general hygiene, in one establishment visited there was no physical separation (i.e. some space) between raw material and products ready to be smoked. Stagnant water on the floor after washing whole and gutted fish was also noted that could splash the products ready to be smoked which were kept nearby. In another two establishments visited protective clothes, including shoe protection, were used both inside and outside premises6. Regarding HACCP plans, in one establishment one process was not described in the plan; in another establishment the official inspector agreed with the audit team that the plan needed to be reviewed as it did not reflect the reality of production7. Regarding L. monocytogenes, in the establishments visited the audit team noted cases when the limit for L. monocytogenes in smoked products after their production was established at <10 cfu/g, the number of units comprising the sample was limited to one, testing of products was only carried out at the end of the shelf-life and never before the food left the immediate control of the FBO, frequencies described in the FBO procedures were not met and testing of equipment and surfaces was not included in the FBO's own-check plan8. 5 During the audit, the DVFA provided evidence of corrective actions already taken to correct some of these situations. 6 In their response to the draft report DVFA informed that this has also been followed up, so protective clothes are not used outdoors. This follow-up message was given orally. 7 In their response to the draft report DVFA noted that the HACCP plan shall according to the inspection frequency guidelines section 7 be inspected at least once every year (or every other year in elite establishments) at wholesalers, so the deficiencies would probably have been discovered during this control. The DVFA added that they will follow-up regarding the first establishment, so that all processes are included in the HACCP plan. For the second establishment, DVFA stated that they will ensure that the HACCP plan is reviewed to reflect the reality of the production. 8 In their response to the draft report DVFA noted that the limit for content of Listeria was corrected to "not detected in 25 g", that testing was started to be carried out before the food leaves the immediate control of the FBO, and the frequencies for control described in the FBO procedures were started to be met. Testing of equipment and surfaces were included in the FBOs own check plan. And, as the HACCP plan also had been effective in the other establishment, the DVFA had also accepted that the number of sample units could be 8 Regarding histamine, in one establishment visited the FBO did not carry out testing for histamine as it had requested the corresponding laboratory results from its supplier. However, these results comprised one sample instead of the nine required in Regulation (EC) No 2073/2005, and the testing method was not the method foreseen in the same Regulation9. Regarding food additives, in one establishment visited an additive non- authorised in fresh fish fillets (E 500) was used; in other establishments, a mix of additives which composition was not clear was used to prevent melanosis in Norwegian lobster. In addition, in this case the instructions provided by the supplier differed from the instructions on the product label and from the FBO procedures. The FBO did not follow a procedure when using it (e.g. specific quantity to be used, the immersion time, etc.).

The audit team noted that most of the above problems arose as official controls were not carried out as a verification (i.e. checking by examination and the consideration of objective evidence) of compliance with the relevant FBO obligations; instead, these official controls were based on simple declarations by the individual FBOs on whether or not there was compliance with the said requirements.

Official controls on production and placing on the market cover all stages of production and are in general implemented in accordance with EU requirements. However, for certain requirements (e.g. use of food additives), officials over-rely on statements and declarations provided by FBOs without actually verifying actual compliance. Thus, in those cases official controls do not provide guarantees that fishery products and FBO comply with relevant Union requirements.

Testing for parasites takes place randomly by visual inspection. In addition, during official controls, official inspectors focus on FBO own-checks and, in particular, if fishery products are intended to be consumed raw or subject to any other treatment that is insufficient to kill viable parasites, FBOs must have adequate procedures in place. These procedures shall include freezing of fishery products derived from wild caught fish or, in the case of farmed fishery products, a certificate to guarantee the requirements of Section VII, Chapter III.D.3.d of Regulation (EC) No 853/200411. However the audit team noted that in one of the establishments visited producing smoked wild salmon, the official inspector had not detected that salmon was not being subjected to freezing at any stage of production12. 11 According to the information provided by DVFA, Danish farmed trout is exempted from complying with this requirement until 1 January 2018 based on a study carried out by the Laboratory of Aquatic Pathobiology, Department of Veterinary Diseases Biology, Faculty of Health Science, University of Copenhagen named "Development of certification methodology for Danish mariculture for use by relief from new EU freezing requirements (2014)". 12 In their response to the draft report DVFA noted that a follow-up on the matter has been made, so that there now are procedures in place that ensure that all wild salmon is frozen if they are to be consumed raw or almost raw. 10

Official controls of fishery products are carried out according to EU requirements. The weaknesses detected regarding official checks on the operators' control of parasites indicates that fishery products reach the market with insufficient treatment to kill viable parasites.

At the opening meeting, DVFA provided information on the follow-up of four 2014 RASFF notifications, one related to one establishment and the other three related to another establishment. Regarding the first one, the L. monocytogenes strain was typed by whole-genome-sequencing and it was found that the L. monocytogenes type was identical to human isolates of listeria (involving mortality) and that this type had not, at that time, been seen in other samples from Danish establishments. Although the FBO revised the establishment's procedures, there was a new human outbreak in 2015 involving this specific L. monocytogenes type. Further measures have been taken by the FBO and, at the time of the audit, DVFA was monitoring them. Regarding the other three RASFF notifications, also involving L. monocytogenes, the follow-up focussed on the procedures in place by the FBO and the FBO's laboratory results. Advice concerning shelf-life labelling and monitoring was given by DVFA but no official samples to verify the FBO's laboratory results were taken (see paragraph 23 as part of the non- compliances described were seen in this establishment)14.

There is a system in place to investigate problems that occur and to take action in response to RASFF notifications. However, the over-reliance on statements and declarations provided by FBOs described in conclusion 27 has also had an adverse impact in the follow-up of RASFF notifications.

Medlemsstatens svar

od is defined in Article 2 of Regulation (EC) No 178/2002 and includes water "intentionally incorporated into the food during its manufacture, preparation or treatment." I.e., water that is not incorporated into the food is not itself food. Since ice used for chilling fish is not incorporated into the fish it is not food, and companies producing ice solely for this purpose are not a food businesses. Therefore Companies producing ice solely for chilling fish are not registered as food businesses and the Danish Veterinary and Food Administration does not perform official controls in these companies. In order to fulfil the requirements in Annex II, Chapter VII, point 4, of Regulation (EC) No 852/2004 the food business operator receiving ice for chilling fish must ensure that the ice is produced from potable water or clean water and is made, handled and stored before receipt in the food establishment under conditions that protect it from contamination. This is described in section 11.2 of the Danish guide on food hygiene (No 9236 of 29 April 2014). DVFA perform audits on the FBO's procedures for receiving ice. Companies producing ice solely for chilling fish have developed an industria

Rekommendationen anger som grund

Official controls on production and placing on the market cover all stages of production and are in general implemented in accordance with EU requirements. However, for certain requirements (e.g. use of food additives), officials over-rely on statements and declarations provided by FBOs without actually verifying actual compliance. Thus, in those cases official controls do not provide guarantees that fishery products and FBO comply with relevant Union requirements.

Medlemsstatens svar

he As stated in the DVFA comments to the draft report, the DVFA does not agree that there is a general overreliance by the DVFA inspectors on the statements made by FBOs. The DVFA will deal with verification in relation to food additives in the experience group for additives and the experience group for Page: 1 eport ref. DG(SANTE)/2015-7460-MR of the audit carried out from 23 June place governing the production and placing on the market of fishery products fish and fishery products. DVFA plans 6 two days courses for inspectors concerning control of own checks in 2016 and 2017, and verification of compliance by the FBO will be one of the items at the course. The first course is planned to take place in March 2016.

Rekommendationen anger som grund

In the slaughterhouse, in the processing plants and in the transporter visited, the audit team noted that, although checks on requirements of the ABP legislation were carried out during official controls, these had never included the specific requirements applicable to sourcing and transport of ABP to be used for the production of derogated PAP for aquaculture feeding. Moreover, it was noted that neither guidelines nor checklists were available to the inspectors to assist them in verifying compliance with these requirements. As a result, official controls overlooked a number of deficiencies (see points 11, 12 and 13). This is not in line with the requirements on official controls laid down in Section F, Chapter V of Annex IV to Regulation (EC) No 999/2001.

Official controls in manufacturers of feed for aquaculture cover aspects related to the feed ban in general. A specific checklist, which includes feed ban related issues, is available, 4 although it is officially no longer in use because it needs to be updated in order to include the specific requirements for the use of derogated PAP for aquaculture feeding. In this respect, the audit team noted that some requirements related to operators' own checks and their procedures for sourcing eligible PAP had been overlooked during official controls (see points 14 and 17). This is not fully in line with the requirements on official controls laid down in Section F, Chapter V of Annex IV to Regulation (EC) No 999/2001.

However, the audit team noted that the DVFA laboratory did not always use the Standard Operating Procedure describing the operational protocols for the combination of light microscopy and PCR, developed by the EURL-AP, although this procedure is a binding complement of the requirements laid down by Annex VI to Regulation (EC) No 152/2009. All samples were first submitted to screening by the light microscopy method irrespective of the nature of the concerned sample. It was noted that aquaculture feed samples containing blood meal were not submitted to PCR, contrary to the aforementioned Standard Operating Procedure, because the light-microscopy method did not detect any particles from terrestrial origin. Following the audit and during the drafting of the report, the competent authority informed the audit team that the laboratory would now follow the above-mentioned Standard Operating Procedure. 5 Conclusions on official control systems

A system of official controls on the requirements concerning the ABP legislation and the general feed ban is in place in slaughterhouses, ABP plants, transporters of ABP and PAP and in feed manufactures producing feed for aquaculture animals. This also includes a sampling plan to detect ruminant DNA in feed for aquaculture animals. However, no specific checks on the new requirements pertaining to the use of derogated PAP in aquaculture feed have been organised yet. As a consequence, although the competent authorities undertook some immediate actions to address some of the issues identified, they are not yet in position to ensure that only eligible PAP is used for the production of feed for aquaculture animals.

In one of the processing plants visited, the transport of ABP was outsourced to an external ABP transporter, who was also authorised to transport Category 2 and Category 3 ABP with no limitations in terms of animal species. This transporter was also providing the reusable containers. The audit team noted that the processing plant operator did not have any measures in place to ensure that the reusable containers were either dedicated exclusively for transport of eligible ABP or were cleaned by means of an authorised cleaning procedure prior to being used for the transport of ABP destined for production of 6 derogated PAP as required by Chapter IV, Section D, point (b) of Annex IV to Regulation (EC) No 999/2001. This was not identified during official controls.

The audit team noted that in one of the two processing plants visited consignments of PAP were dispatched without the required warning sentence in the commercial documents, as laid down by Chapter IV, Section D, point (e) of Annex IV to Regulation (EC) No 999/2001. This was not identified during official controls. 5.2.2 Use of PAP for producing compound feed

Both fish feed producers visited in Denmark were dedicated to the production of fish feed. According to the feed business operators, the only information required to their suppliers of PAP was an evidence of approval/registration in accordance with Regulation (EC) No 1069/2009 and a written declaration of the origin of PAP in terms of animal species. The audit team noted that no further information to confirm the origin and the eligibility of the PAP had ever been requested by the feed business operators. This issue was not checked during official controls.

The fish feed manufacturers visited used labels on the products containing information in several languages. In order to accommodate the information required in all languages, the operators made use of very small fonts difficult to read. Moreover, it was noted that in feed containing fish meal and other PAP, the operators used at the same time the warning sentence required for fish meal and the one required for PAP. As a result there was a risk of misinterpretation because fish meal can be fed to all farmed animals except ruminants, whereas PAP can only be fed to aquaculture farmed animals. This issue was not checked during official controls. 5.2.3 Actions in case of non-compliance

Official controls carried out in slaughterhouses, ABP plants, transporters of ABP and PAP and in feed manufactures producing feed for aquaculture animals, do not yet include regular checks on the new requirements pertaining to the use of derogated PAP in aquaculture feed. This limits the ability of the competent authorities to ensure that only eligible PAP is used for the production of feed for aquaculture animals.

Medlemsstatens svar

1) The central authority has informed the industry concerning the ure requirements in a general meeting in May 2015. 2) In May and June there have been conducted education days of the staff conducting inspections in the slaughterhouses, processing plants and transporters of ABP. 3) Guidelines for the inspections on requirements on slaughterhouses, processing plants, manufactures of feed for aquaculture and transporters of ABP have to be drafted and submitted to each part of the control. The work will be done in cooperation between the central authority and the inspectors. Deadline is in the end of 2015. Guidelines will be sent to the FVO as soon as possible. 4) The DVFA use the Standard Operating Procedure (SOP) "Operational protocols for the combination of light microscopy and PCR". The DVFA take samples dedicated to analyse for PCR detecting ruminant material. The Danish laboratory IT-system (LIMS) now "automatically lead" these samples to both Microscopy and PCR analysis, making sure that the PCR are done after the microscopy if relevant, or PCR are done directly in the case of content of animal feed material e.g. blood products. 5) The guideline for taking samples and analysing

Medlemsstatens svar

Regarding animal welfare assessment upon arrival (Annex III (1.1)) e DVFA has taken contact to the industry by letter of 14 November 2014 (Attached to the e-mail) with instruction that the industry guidelines must be updated, so that they specify, that all aughterhouses (including slaughterhouses which slaughter less than 1000 livestock unit per year) must perform animal welfare ssment of each consignment of animals upon the arrival at the slaughterhouse. The industry must no later than by the end of February 2015 submit a revised industry guideline for cattle and pigs to the DVFA for assessment. The industry guideline for the smaller slaughterhouses has since the FVO inspection been sent to the DVFA in a revised version for assessment. It is expected, that the revised guideline for the smaller slaughterhouses will be assessed before 1 February 2015. In the DVFA guideline concerning Regulation 1099/2009 the DVFA already refer directly to the requirement in Annex III (1.1) of Regulation 1099/2009 regarding animal assessment upon arrival. To make it more clear, the DVFA will in a revised DVFA guideline concerning Regulation 1099/2009 specify, that the requirements in Annex III, (1.1)

Medlemsstatens svar

Training courses to obtain certificate of competence with regard laughter operations Danish Veterinary and Food Administration (DVFA) do not with this recommendation as we do not find that it - according he Regulation - is a requirement that training or examination to in certificate of competence must include practical training nds-on component"). ecital 28 a reference is made to "well trained and skilled onal". Practical training is not explicitly mentioned, and nor is it nnex IV. In Annex IV is as an example mentioned "Practical cts of stunning techniques", but it is the view of the DVFA that e practical aspects are covered by the training in the existing ning courses. t the time of slaughter or killing Page: 2 t ref. DG(SANCO)/2014-7061 -MR of a/an audit carried out from 05 e at slaughter and during related operations. The course participants are during the courses in Denmark receiving a thorough training and thus become skilled to work in a slaughterhouse performing slaughter operations. The skills are assessed during an examination at the end of each course. The skills, including knowledge of practical aspects, learned at the course, will enable staff to carry out their work a

Medlemsstatens svar

wording of the Danish certificates of competence will be ) changed, thus it will be more clearly indicated for which type of t the certificate is valid. In the present Danish certificates o the specie(s) and the type of certificate - indicated in general hich the operator is competent. ioned "equipment used for the t ref. DG(SANCO)/2014-7061 -MR of a/an audit carried out from 05 e at slaughter and during related operations. ndling of animals before they are restrained, restraining uipment, stunning equipment and bleeding equipment". e change will be in effect from 1 February 2015.

Medlemsstatens svar

he Deputy Head of Meat Inspection verifies that audit and spections aimed at standard operating procedures (SOP) and dustry guides of good practice for the humane handling and killing animals does not only focus on whether the slaughterhouse has but also on the actual compliance of the procedures with the ulation requirements and if they have adapted the guides of od practice to the specific slaughterhouse, see also point 5.3. e Meat Inspection Department has instructed all Deputy Heads of at Inspection of this by email of 19 November 2014 (Attached to e e-mail as Annex 2).

Medlemsstatens svar

e Meat Inspection Department have on 19 November 2014 structed all Deputy Heads of Meat Inspection by email that in the ture the current checklists prepared by the DVFA must be used rforming control with regard to the humane handling and ng of animals. (Attached to the e-mail as Annex 2) The DVFA will make sure, that the current checklists and e guideline for Regulation 1099/2009 will be revised fore end of May 2015. Before end of May 2015 a course on animal welfare to improve awareness of the requirements of the regulation will be held for veterinary officers dealing with the protection of animals at the time of killing. Page: 4 t ref. DG(SANCO)/2014-7061 -MR of a/an audit carried out from 05 e at slaughter and during related operations. 3. In 2015 the Deputy Heads of Meat Inspection will participate in a regular audit on animal welfare carried out by the veterinary officer. The Deputy Heads of Meat Inspection will verify that the veterinary officers perform the audit correctly. In case of noncompliance the Deputy Heads of Meat Inspection will ensure that the veterinary officer acts correctly and ensure - if necessary - the use of sanctions due to infringements of the Regulation.

Medlemsstatens svar

1. The Veterinary Officers are required to perform a minimum of 4 annual inspections and 1 annual audit aimed on the animal welfare requirements. The control results have to be documented in the veterinary control report. In case of non-compliance, the veterinary officer verifies that the business operator and animal welfare officer takes corrective actions and if necessary uses the appropriate sanction. In cases, where it is allowed according to guidelines on inspection frequencies, the veterinary officer guides the establishment. 2. FVO recommendations have been sent to all the Deputy Heads of Meat Inspection, i.e. not only to slaughterhouses visited by FVO inspectors by e-mail of October 21, 2014 (attached). The Meat Inspection Department has by e-mail of 18 June 2014 (attached) sent instructions to all the Deputy Heads of Meat Inspection with an update on inspection frequencies. The Meat Inspection Department has by e-mail on 19 November 2014 given all Deputy Heads of Meat Inspection further instructions regarding findings and Page: 5 t ref. DG(SANCO)/2014-7061 -MR of a/an audit carried out from 05 e at slaughter and during related operations. recommendations. (Attached as Anne

Medlemsstatens svar

instruction concerning approval of slaughterhouses is being sed. It will include a reference to the requirements of Article 4(2) of Regulation (EC) No 1099/2009. The revised instruction is cted to be in place in the beginning of 2015. her, the instruction concerning audits in slaughterhouses will ude a specific reference to the requirements of Article 14(2) of lation (EC) No 1099/2009. This change will be incorporated in Control Frequency Guide for 2015 and will be in effect from 1 ary 2015.

Medlemsstatens svar

DVFA has by letter of 19 November 2014 (attached to the e- il) made contact to the Danish manufacture of stunning nt and repeated a previous request for receiving a link to the bsite of the company, where instructions are made available. The has also asked to receive files with the instructions, which mpany the stunning and restraining equipment when sold. If the does not receive the requested information the DVFA will rt the company to the police for failing to comply with Article 8 he Regulation. DVFA will before 15 December 2014 make contact to a Danish ch of a foreign owned company, which produces restraining and ning equipment, to inform the company of its obligations rding to Article 8 in the Regulation.

Medlemsstatens svar

eputy Heads of Meat Inspection have been instructed, that Page: 6 t ref. DG(SANCO)/2014-7061 -MR of a/an audit carried out from 05 e at slaughter and during related operations. they must verify that all slaughterhouses (including slaughterhouses which slaughter less than 1000 livestock unit per year) have procedures for animal welfare assessment of each consignment of animals upon the arrival at the slaughterhouse. If the veterinary officer take responsibility for assessing animal welfare upon arrival at the slaughterhouse, this distribution of responsibility between the veterinary officer and the slaughterhouse has to be laid down in the own check program. It must be clear, who is responsible for assessment of animal welfare upon arrival of the animals, when a veterinary officer is not present when the animals arrive. The Meat Inspection Department has instructed all Deputy Heads of Meat Inspection of this by email on 19 November 2014. (Attached as Annex 2 to the email)

Medlemsstatens svar

The DVFA has prepared a report based on data extractions, which - show whether or not the required audit and the four inspections annually have been performed. In this way the Deputy Head of by Meat Inspection and the Heads of Meat Inspection can verify that control on animal welfare is being conducted in accordance with the Regulation and the guidelines for inspection frequencies. A guideline for the report will be drawn up, and the report and guideline is expected to be put into production per January 1, 2015. Before end of May 2015 a course on animal welfare to improve awareness of the requirements of the regulation will be held for veterinary officers dealing with animal welfare. In 2015 the Deputy Heads of Meat Inspection will participate in a regular audit on animal welfare carried out by the veterinary officer. The Deputy Heads of Meat Inspection will verify that the veterinary officers perform the audit correctly. In case of noncompliance the Deputy Heads of Meat Inspection will ensure that the veterinary Page: 7 t ref. DG(SANCO)/2014-7061 -MR of a/an audit carried out from 05 e at slaughter and during related operations. officer acts correctly and ensure - if necessary - t

Medlemsstatens svar

cts are DEPA is committed to fulfil the deadlines for re-authorisations ­ in 2014 a fee system has been implemented in order to ensure that the necessary resources to perform the evaluations are available. However in some cases we are challenged by delayed evaluations from other MS or problems that arise during the evaluations and need to be addressed by the applicant ­ under these circumstances it is necessary to extend the existing authorisation.

plant DEPA is committed to fulfil the deadlines for by re-authorisations ­ in 2014 a fee system has been implemented in order to ensure the necessary resources to perform the evaluations. However in some cases we challenged by delayed evaluations from other MS or problems that arise during the evaluations and need to be addressed by the applicant ­ under these circumstances it is necessary to extend the existing authorisation.

Medlemsstatens svar

and DEPA does not agree that emergency uses are granted in cases where there are other means. Emergency uses are only granted when applications for regular authorizations cannot or have not been submitted or when a regular authorization is not possible due to the need for extensive risk mitigation measures and when there are no other reasonable means of controlling the pest. DEPA encourages submission of applications for regular authorizations and minor uses in cases where this is possible. In 2014 we have granted a minor use for an insecticide with indoxacarb in order to avoid the emergency authorization of dimethoate and we have granted extended use of Opera in maize.

ed DEPA does not agree that emergency uses are red by granted in cases where there are other means. Emergency uses are only granted when applications for regular authorisations cannot or have not been submitted or when a regular authorisation is not possible due to the need for extensive risk mitigation measures and when there are no other reasonable means of controlling the pest. DEPA encourages submission of applications for regular authorisations and minor uses in cases where this is possible. In 2014 we have ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to evaluate controls of pesticides. granted a minor use for an insecticide with Indoxacarb in order to avoid the emergency authorisation of Dimethoate and we have granted extended use of Opera in maize. Response April 2015: DEPA maintains that emergency use is only granted in situations where there are no other reasonable means of controlling a pest in Denmark. We strongly encourage applicants to submit applications for regular authorisation of the concerned products or alternative products, or growers to apply for

Medlemsstatens svar

wed The Chemical Inspection Service (CIS) has performed control nd is with the marketing of PPPs in Denmark for over 50 years and therefore has very good knowledge of the official market. The control with the unofficial part, which comprises possible illegal and/or counterfeit PPPs, is performed partly via the cooperation with the Danish Tax and Customs Authority and the Danish AgriFish Agency and partly via reports from the companies, which sell the authorized PPPs in Denmark. The market has, however, changed in the later years and the CIS Page: 1 ctober 2015 of Report ref. DG(SANCO)/2014-7184 -MR of a/an audit carried out from 17 PPPs must be ctive control. in order to e- packers, f how the analysis will be evised control proved and ow the case of digms for

The CIS agrees that a strategy for the control ccount of PPPs must be formulated, resulting in a ow, more up to date and effective control. ls is Therefore, the CIS will implement a market analysis in order to create an overview over importers, manufacturers, re-packers, retailers and wholesalers of PPPs and an overview of how the distribution of PPPs take place today. This will be followed by a risk based inspection plan and a revised control guideline. The present control guideline will be improved and concretised further in order to describe what and how the inspectors shall control and how they shall act in case of violations. The control guideline will include paradigms for communication with the different operators.

Medlemsstatens svar

icial The CIS control with PPPs is performed by 3 inspectors and 1 them lawyer. One of the inspectors has controlled PPPs in more than 20 82/2004. years and has therefore a lot of experience. As of November 1, will go through he area. duction into cement and an The training vant employees rticipating rea. PPPs will be levant e of PPPs like Page: 2 ctober 2015 of Report ref. DG(SANCO)/2014-7184 -MR of a/an audit carried out from 17 for example washing sites, is performed by following the AgriFish Agency inspectors when they control the use of PPPs on farms.

The CIS control with PPPs is performed by 3 PPPs inspectors and 1 lawyer. One of the keep inspectors has controlled PPPs in more than e 6 of 20 years and has therefore a lot of experience. ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to evaluate controls of pesticides. r 1, 2014, two new inspectors ice and they will go through a ram in order to be able to control g for new employees comprises on into the rule area, an to control and enforcement and on to the CIS control guideline training takes place partly by erformed by relevant employees sh EPA and by lawyers and ticipating in inspections with nspectors on the area. up to date of new and amended will be implemented by ad hoc seminars with relevant lawyers in the Danish EPA. CIS inspectors in inspection of Ps like for example washing formed by following the AgriFish tors when they control the use rms. ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to evaluate controls of pesticides.

Medlemsstatens svar

operators In Denmark, registration of all distributors of PPPs is not a riate requirement. Only the about 100 distributors which have . permission to sell poisonous and very poisonous PPPs are registered (Poison permission). We have estimated that about 150 distributors sell to farmers and other professional users and about 1600 distributors sell to private consumers. The CIS performs a risk based control: a regular control with those distributors which have permission to sell poisonous products; moreover control is performed with the distributors to professional users and last with the distributors which earlier have violated the rules. A control with all distributors for private consumers will not be proportional and will be very resource demanding, whereas the CIS agrees that a more frequent control with the 150 distributors for farmers and other professional users would be relevant. The CIS will have a market analysis performed in order to gain an overview over importers, manufacturers, re- packers, retailers and wholesalers of PPPs and an overview of how the distribution of PPPs take place today. On the basis of the market analysis an inspection guideline for the control in th

of The CIS will have a market analysis performed in order to gain an overview over ate importers, manufacturers, re-packers, retailers n and wholesalers of PPPs and an overview of how the distribution of PPPs take place today. On the basis of the market analysis an inspection guideline for the control in the sales chain will be developed.

Medlemsstatens svar

ed to Today if suspicious PPPs are found during an inspection at a Article distributor, the CIS will request documents for purchase and sale of the PPP in question. If CIS estimates that it concerns a violation which will be reported to the police, the case will be handed over to the police for investigation. Page: 3 ctober 2015 of Report ref. DG(SANCO)/2014-7184 -MR of a/an audit carried out from 17 CIS plans to expand the control at the distributors to include document control during the visit regarding purchase and sale of relevant PPPs.

CIS plans to expand the control at distributors to include document control during the visit ticle regarding purchase and sale of relevant PPPs.

Medlemsstatens svar

equired The CIS controls that the PPPs are stored safely at the distributors, both at the warehouses and in the stores, and that poisonous PPPs are stored in locked rooms. "Safe storage" is based on CISs estimation: does the storage looks safe, proper with enough room and not disorderly, overloaded and chaotic. The city councils in Denmark give building permissions for the lay-out of warehouses, where PPPs are stored, in order to prevent discharge into the environment in case of spillage The city councils also are the environmental authority and have the responsibility of most of the environmental regulation control. The city councils subsequently handle the control with the lay-out of the buildings to prevent discharge of emission into the environment.

onal The CIS controls that the PPPs are stored safely at the distributors, both at the warehouses and in the stores, and that poisonous PPPs are stored in locked rooms. "Safe storage" is based on CISs estimation: does the storage looks safe, proper with enough room and not disorderly, overloaded and chaotic. The city councils in Denmark give building permissions for the lay-out of warehouses, where PPPs are stored, in order to prevent ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to evaluate controls of pesticides. discharge into the environment in case of spillage The city councils also are the environmental authority and have the responsibility of most of the environmental regulation control. The city councils subsequently handle the control with the lay-out of the buildings to prevent discharge of emission into the environment. Response April 2015: The present guideline for inspections will be improved and further guidance will be added in order to describe what and how the inspectors shall control "Safe storage" for all PPPs and the storage location, size and const

Medlemsstatens svar

mely From 2014 more physic-chemical analysis will be performed on the PPPs selected for control. In connection with the control on (EC) analysis of parallel imported PPPs a GC-screening will be performed and a comparison of HPLC-chromatograms with the original products, to see if there are differences. Differences could be an indication of a product being counterfeit. The CIS selects PPPs for analysis in the period from March to June. The laboratory performs the analysis from July to December same year. It will cost many more resources to select the PPPs with one specific active substance at a time in order to be able to Page: 4 ctober 2015 of Report ref. DG(SANCO)/2014-7184 -MR of a/an audit carried out from 17 sis immediately after in order to shorten the ing to the result of the analysis.

From 2014 more physic-chemical analysis s will be performed on the PPPs selected for control. In connection with the control ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to evaluate controls of pesticides. parallel imported PPPs a GC- ll be performed and a f HPLC-chromatograms with products, to see if there are Differences could be an f a product being counterfeit. cts PPPs for analysis in the March to June. The laboratory analysis from July to December t will cost many more resources e PPPs with one specific active a time in order to be able to analysis immediately after in rten the period from sampling to f the analysis. il 2015 e CIS will for each sample whether the PPP is likely to be a t product. This assessment can he label looks different from the ed by the DEPA; if the packaging orized; if the PPP leaks through g or if the DEPA registration t a legally approved PPP number. If the CIS suspects that ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to e

Medlemsstatens svar

ting of In case of violation of the rules on marketing of PPPs, the required distributor will receive a letter from the CIS after the inspection example it is enjoined that non-authorised PPPs sold. If the products may not be possessed or CIS will undertake a hearing of the distributor issue an enforcement order on the destruction of h no longer must be possessed or used. The get a deadline of 14 days to respond to the case. If as any comments the CIS will estimate the case comments. Subsequently, CIS can issue an to the distributor that the illegal products must re a fixed deadline. Moreover the distributor entation to the CIS that the illegal products are estimates that there is a risk of misuse of the an confiscate the products. anish AgriFish Agency cooperate on a uniform for violation of the rules on PPPs. ports repeated or severe violations of the rules to his depends on the specific evaluation of the tion- and control case. By marketing of foreign Ps in Denmark the distributor will be reported 013 determined rules on increased penalties of marketing, use and possession of illegal PPPs. eased penalties were amongst other things rease the sanctions in cases where illegal PPPs

In case of violation of the rules on marketing of PPPs, the distributor will receive a letter ired from the CIS after the inspection visit, where for example it is enjoined that non-authorised PPPs may no longer be sold. If the products may not be possessed or used anymore, the CIS will undertake a hearing of the distributor that the CIS will issue an enforcement order on the destruction of the products which no longer must be possessed or used. The distributor will get a deadline of 14 days to respond to the case. If the distributor has any comments the CIS will estimate the case and the received comments. Subsequently, CIS can issue an enforcement order to the distributor that the illegal products must be destroyed before a fixed deadline. Moreover the distributor must submit documentation to the CIS that the illegal products are destroyed. If CIS estimates that there is a risk ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to evaluate controls of pesticides. of misuse of the illegal PPP CIS can confiscate the products. The CIS and the Danish AgriFish Agency cooperate

Medlemsstatens svar

d by Denmark's implementation of the requirement in Article 14(4) nsure that the general principles of nt as set out in Annex III are ssional users by 1 January 2014" is to a r new differentiated pesticide tax, which 2013, and by which we make sure that a horised pesticides and that the size of the tax products depends on the intrinsic properties ides/active substances (their loads with nd health). Furthermore, a basic tax on nce is levied on all pesticides in addition order to ensure that there is a tax on all esticides have a very low tax in general the tax model). s to ensure, that the professional user will ve to: s as sound crop rotation, preventive cultivation techniques, resistant cultivars, etc. ning systems for a number of pests, l methods, ides with lower load on the environment and human health, the authorized products. x is the most important tool to ensure that principles are met. Page: 6 ctober 2015 of Report ref. DG(SANCO)/2014-7184 -MR of a/an audit carried out from 17 As a supplement to the tax, we have established several other initiatives regarding advisory service, demonstration farms, education etc. In order to follow (or monitor) the development of the IPM imp

al Denmark's implementation of the e requirement in Article 14(4) regarding how ndations (updated) received 30/4/2015 Competent Authorities of Denmark on 22 October 2014 in response to Report ref. DG(SANTE)/2014- 14 to 24 June 2014 in order to evaluate controls of pesticides. we will "ensure that the general principles of integrated pest management as set out in Annex III are implemented by all professional users by 1 January 2014" is to a large extent based on our new differentiated pesticide tax, which came into force by July 2013, and by which we make sure that a tax is levied on all authorised pesticides and that the size of the tax level of the individual products depends on the intrinsic properties of the individual pesticides/active substances (their loads with respect to environment and health). Furthermore, a basic tax on amounts of active substance is levied on all pesticides in addition to the load based tax in order to ensure that there is a tax on all pesticides. (microbial pesticides have a very low tax in general and are not included in the tax model). The purpose of the tax is to ensure, that the professional user will have an economic incentive to: Use preventive m

Medlemsstatens svar

ndation should be deleted or changed. As the draft report there is only described one rt with insufficient documentation and it was eport with insufficient documentation we the specific situation. Besides this DVFA eed for further special action on this point. ocus on describing the control adequately. er actions holds courses on how to ly when needed. Furthermore DVFA checks n reports.

Medlemsstatens svar

lans for establishments handling meat, or registration must be taken under review hin a period of one year. ine to the CAs that deficiencies found on is control area must be handled in accordance specific guides for various control topics.

Medlemsstatens svar

hterhouses and Pork slaughterhouse: walls and equipment) were not maintained in loor, presence of rust and flaking paint on me instances above exposed Page: 1 -6686-MR of an audit carried n and placing on the market of t. Will be ater than the below floor. Repairs of documented by local ember 2013. mation of t frequency in d documented by with raw material for he same day. Page: 2 -6686-MR of an audit carried n and placing on the market of he local ishments ensure that This includes r plate freezing (- stration. perature of ective action. e registration form. for deficiencies in regarding ficial control. nce regarding Page: 3 -6686-MR of an audit carried n and placing on the market of order for this non- ed as well. CA procedures for Regarding trace the product, tion in a case of mplement another y basis. btained from deboning whereas must take place ly report by the itten and en validated by the ecretariat for Meat inary officers (DVOs). uidelines. -6686-MR of an audit carried n and placing on the market of Authority

Medlemsstatens svar

The reorganisation of the laboratory service has had an influence on the turnaround time in 2012 and partly in 2013. The analytical control of feed is now carried out at two laboratories, which guarantee that the capacity is adequate. In order to reduce the costs for the business operators, the analytical control is now organised so that the samples normally will be analysed in full series and not necessary on the day of arrival. This can prolong the turnaround time. The laboratory is also working on a new organisation of the sampling, which will reduce the turnaround time. Concerning screening methods for coccidiostats used in Denmark the situation in 2012 was that monensin, lasalocid, salinomycin and narasin/nicarbazin were used (Informations from Vetstat). The LC/MS screening method is able to detect these substances except nicarbazin. This substance is only used in Denmark when mixed with narasin 1:1. So when we analyze for carry-over of narasin, we also analyse indirectly for nicarbazin.

The reorganisation of the laboratory service has had an influence on the turnaround time in 2012 and partly in 2013. The analytical control of feed is now carried out at two laboratories, which guarantee that the capacity is adequate. In order to reduce the costs for the business operators, the analytical control is now organised so that the samples normally will be analysed in full series and not necessary on the day of arrival. This can prolong the turnaround time. The laboratory is also working on a new organisation of the sampling, which will reduce the turnaround time. Concerning screening methods for coccidiostats used in Denmark the situation in 2012 was that monensin, lasalocid, salinomycin and narasin/nicarbazin were used etent Authorities of Denmark on 28 February 2014 in response to Report ref. DG(SANCO)/2013- 13 to 19 September 2013 in order to evaluate measures in place for the identification of hazards and oils, fats and products derived thereof. (Informations from Vetstat). The LC/MS screening method is able to detect these substances except nicarbazin. This substance is only used in Denmark when mixed with narasin 1:1. So when we analyze for carry- over of narasin,

Medlemsstatens svar

ive DVFA has a comprehensive verification system as mentioned in FVO report 2013-6857. DVFA will ensure that all feed inspectors have been quality supervised by the end of 2014. The quality supervision guideline will be changed by July 2014 so the control of feed is fully incorporated.

l DVFA has a comprehensive verification system as mentioned in FVO report 2013- 6857. DVFA will ensure that all feed inspectors have been quality supervised by the end of 2014. The quality supervision guideline will be changed by July 2014 so the control of feed is fully incorporated.

Medlemsstatens svar

From January 2014 a new database for registered and approved feed establishments has been put into operation by the DVFA. Registration of the information "direct drying" is included. This data is considered additional information for the registered Page: 1 Authority on 27 February 2014 t ref. DG(SANCO)/2013-6749 -MR of a/an audit carried out from 10 he identification of hazards and management of risks along the feed chain vity "production", not a separate "activity". aims to prepare guidelines for the inspectors of the official rol and to initiate registration of this information within the t half of 2014. consider indirect drying of feed as covered by the stered activities "production" and/or "storage" (e.g. in ark most grain is indirectly dried by air due to the weather ition during harvest). DVFA consider recycling of food lus as covered by the registered activity "production" and/or rading "distribution" of feed materials. It is currently not ned to register this data as additional information for the stered activity of the establishments. , it is the opinion of DVFA that the referred feed blishments are registered in line with the requirements of cle 9 of Regulation (EC) No 18

From January 2014 a new database for registered and approved feed establishments d etent Authorities of Denmark on 28 February 2014 in response to Report ref. DG(SANCO)/2013- 13 to 19 September 2013 in order to evaluate measures in place for the identification of hazards and oils, fats and products derived thereof. peration by the DVFA. information "direct . ered additional information activity "production", not a . e guidelines for the fficial control and to on of this information within 014. ect drying of feed as stered activities "storage" (e.g. in is indirectly dried by air condition during harvest). ling of food surplus as stered activity by trading "distribution" It is currently not planned ta as additional information activity of the etent Authorities of Denmark on 28 February 2014 in response to Report ref. DG(SANCO)/2013- 13 to 19 September 2013 in order to evaluate measures in place for the identification of hazards and oils, fats and products derived thereof. establishments. us, it is the opinion of DVFA that the ferred feed establishments are registered in ne with the requirements of Article 9 of gulation (EC) No 183/2005 for all their tivities.

Medlemsstatens svar

will in the future ensure that companies only use feed that is not labelled as "waste". DVFA will ensure this by a more sed control on the labelling of food co-products. DVFA will s even more on guiding and teaching the inspectors (both feed ectors and food inspectors, who inspect feed at the food co- ucers).

FA will in the future ensure that mpanies only use feed that is not labelled "waste". DVFA will ensure this by a more cused control on the labelling of food co- oducts. DVFA will focus even more on iding and teaching the inspectors (both ed inspectors and food inspectors, who spect feed at the food co-producers). date received from CA 17 June 2014 e proposed guiding and teaching are anned to be started in 2014.

Medlemsstatens svar

he DVFA will ensure that the requirements laid down by Article of Regulation (EC) No 183/2005 concerning dioxin monitoring are known by the inspectors in the Food Control Offices and are included in the control. will additional assess the need for update of guidelines for inspectors concerning dioxin monitoring at establishments also r the scope of Regulation (EC) No 852/2004. Page: 2 hority on 27 February 2014 t ref. DG(SANCO)/2013-6749 -MR of a/an audit carried out from 10 he identification of hazards and management of risks along the feed chain

e DVFA will ensure that the requirements id down by Article 5(2) of Regulation (EC) 183/2005 concerning dioxin monitoring e known by the inspectors in the Food etent Authorities of Denmark on 28 February 2014 in response to Report ref. DG(SANCO)/2013- 13 to 19 September 2013 in order to evaluate measures in place for the identification of hazards and oils, fats and products derived thereof. e Control Offices and are included in the control. DVFA will additional assess the need for update of guidelines for the inspectors concerning dioxin monitoring at establishments also under the scope of Regulation (EC) No 852/2004. Update received from CA 17 June 2014 Both of the proposed tasks are planned completed before end of 2014.

Medlemsstatens svar

VFA will in 2014 adjust the current guidelines for the inspectors egarding control of measures put in place regarding carry over of ccidiostats

DVFA will in 2014 adjust the current guidelines for the inspectors regarding control line of measures put in place regarding carry over of coccidiostats.

Medlemsstatens svar

n the new official control concept for feed and feed stablishments starting January 2014 it is anticipated that the risk nalysis and the HACCP plan shall be inspected every year at n most feed producers. However, in establishments with an ordinary nspection frequency every other year, then only every other year. ome establishments will be inspected as and when needed. As a onsequence of the new concept, control of hazard analysis and ACCP-based procedures of the establishments will be a more ntegrated part of the ongoing control. VFA will additional assess the need for update of guidelines for he inspectors concerning relevant risks in the risk analysis. Page: 3

In the new official control concept for feed ed and feed establishments starting January 2014 it is anticipated that the risk analysis and the HACCP plan shall be inspected every year at tely most feed producers. However, in put establishments with an ordinary inspection frequency every other year, then only every 7 etent Authorities of Denmark on 28 February 2014 in response to Report ref. DG(SANCO)/2013- 13 to 19 September 2013 in order to evaluate measures in place for the identification of hazards and oils, fats and products derived thereof. ents will be d. As a ept, control of ased ments will be a ongoing control. the need for e inspectors n the risk analysis.

Medlemsstatens svar

n in The contingency plan available on the TRACES tool kit was on 3 June 2013 put on a special me contingency site, in connection with the ion TRACES microsite that already existed on the internal website of the DVFA. Also a 2600-mail was created on 3 June 2013 on the internal website to inform all relevant CA users, and additionally an email was forwarded on this matter to all relevant units, also on 3 June 2013.

Medlemsstatens svar

Please see deadlines inserted in the text in the of column: "Action Proposed by the Competent thority" The DVFA will provide written instructions to the BIPs on the managing of incoming notification from TRACES regarding Pending has laboratory results and Transhipment by the end of 2013. pt at Furthermore, the DVFA will provide written instruction to the LVUs on the managing of incoming notification from TRACES regarding new/valid intra certificates destined for Denmark as well as confirmation of receipt at destination by the end of 2013. Page: 1

Medlemsstatens svar

f This recommendation should be deleted. The CoE recommendations are designed in such a way that they he contain both legislative provisions, which Contracting Parties to ied in the European Convention for the Protection of Animals kept for Farming Purposes shall implement, and guidelines, which Contracting Parties may implement on a voluntary basis. The text of a provision uses the word "shall ", and the text of a guideline uses the word "should". During discussions of recommendations in the Standing Committee established under the convention much attention were given to the different consequences of using "shall" or "should".

Medlemsstatens svar

performing official controls in the area of dequate sampling equipment as required tion (EC) No 882/2004 so they can perform ly, effectively and in line with the specific EU inants.

We will ensure that staff carrying out official controls in the area of inants have access to adequate sampling equipment as required rticle 4 (2) (d) of Regulation (EC) No 882/2004, so they can make fficial controls efficiently and in accordance with specific EU legislation on contaminants by appropriate sampling equipment will be purchased to the extent that it is necessary and where it is not already acquired, as soon as possible and no later than September 1, 2013

Medlemsstatens svar

uideline on control of chemical the control of hazard analyzes is specified, by the l be placed on the DVFA website. We ad show to all local control units during the f the guideline. It is our experience that ective way of reaching inspectors as most of all get the same information. It will be considered if n sampling is necessary.

Ensure that all staff responsible for official controls including those taking samples under contaminants monitoring programmes, receive appropriate training on contaminants and in particular, in the area of food sampling, as required by Article 6 and Annex II of Regulation (EC) No 882/2004 and are kept up to date so that they can perform their task competently as required by Article 4 of the same Regulation.

Medlemsstatens svar

based control. The laboratory projects are groups consisting of the DVFA aboratories and the NRL. The DVFA has now listing proposed projects, relevant methods of lability thereof and the priority of the projects. The to update these tables at every meeting for clarity for traceability over time. Page: 1 6276-MR of an audit carried Action Proposed by the Competent Authority

Ensure that risk-based planning of laboratory projects for contaminants operated by the DVFA is documented as required by Article 8 (1) of Regulation (EC) No 882/2004.

Medlemsstatens svar

receive Single Payment. They are tion from which we by risk analysis and ulation for control for food hygiene in equirements of Regulation (EC) 852/2004 and the to cross-compliance.

primary producers of vegetable food are registered under the food giene regulations. They are therefore subject to official control. g) During the official control for hygiene the focus are at risk analyses, protection of the primary products and prevention of hazard including measures to control contamination arising from fertilizers, plant protection products and biocides. Page: 1 eport ref. DG(SANCO)/2012-6276-MR of an audit carried in food of plant origin on Proposed by the Competent Authority ucers of leafy vegetables receive Single Payment. They are efore also part of the population from which we select the subjects to s-compliance control. While checking for cross-compliance we ys carry out an official control of food hygiene in accordance with requirements of Regulation (EC) 852/2004. requirements laid down in Annex I, Parts A, point II 3 (a), point 5 and 9 (c) are controlled in connection with ordinary hygiene control control of cross compliance, where also the hygiene conditions are rolled. requirement pointed out in Annex I, Parts A, point II 5 (g) and 9 (c) shown at the detailed checklist. This detailed checklist is used by ectors during every hygiene control. Furthermo

Medlemsstatens svar

authority will ensure, that the correct e Commission Regulation is specified the analytical project, including information ty, weight and number of samples. y is responsible for correct sampling according to on. will record information about the specific laboratory is responsible for checking that they amples according to the information registered by t correct, the laboratory will reject the samples. troduced specific procedures to ensure this.

: 2 eport ref. DG(SANCO)/2012-6276-MR of an audit carried in food of plant origin n Proposed by the Competent Authority

Medlemsstatens svar

is of official samples are validated according dited by the Danish accreditation authority ocedure concerning validation and laboratory. ed on the performance criteria laid down in the s. Page: 2 6276-MR of an audit carried Proposed by the Competent Authority

VFA perform risk-based control or risk-based campaigns. As soon as it is ded to run such a campaign, the laboratory will validate the methods and rixes not yet being validated. Approval of the validation will be done by the Accreditation Authority, DANAK. The laboratory's SOPs for validation mplementation of methods ensure that the methods used for detection of minants in official samples are fit for purpose. The laboratory will ensure the test performance meets criteria as defined by current EU legislation. ere is an urgent need of methods not yet validated or not validated for all xes, the samples will be sent to an external accredited laboratory. ethods where the validation data is not fully in line with EU legislation ements to the validation will soon be completed. e future the laboratory will indicate the expanded uncertainty as x +/ - U by x is the analytical result and U is the expanded measurement tainty using a coverage factor of 2. sis results will be reported corrected or uncorrected for recovery. ting method and level of recovery will be given on the certificate.

Medlemsstatens svar

r results according to the recovery and we report lts. We recognize that we did not is reports that our results are corrected for recovery. by making an additional note on the report. he DVFA are implementing a new LIMS system omatically will be able to handle and report the g recovery and expanded measurement be fully operational from the 1st. of January Page: 3

Page: 3 eport ref. DG(SANCO)/2012-6276-MR of an audit carried in food of plant origin y the Competent Authority

Medlemsstatens svar

iance It is DFVA´s opinion that overall, the system for control of dioxin/PCB´s mon in fish from the Baltic Sea operates efficiently in DK. for For further action DVFA will make a specific instruction to the involved inspectors (deadline 1 may). The instruction will include: · To check that the Baltic salmon at the companies comply with the weight limit in the national legislation · To check the traceability both ways. · To check that the Baltic herring for human consumption are caught in areas where it is allowed. · To check that the business operator has identified risk products and has implemented a system for documentation.

Medlemsstatens svar

ss DVFA will finalize the guideline on control of chemical contaminants in d, which the control of hazard analyzes is specified, by the summer 2013. e The guideline will be placed on DVFA´s website. DVFA will also organize a new road show to all food control units during the autumn 2013 with presentation of the guideline. It is our experience that road shows are a very effective way of reaching inspectors as most of them participate and all get the same information. Page: 1

Medlemsstatens svar

elevant Danish legislation (Order no 1219 of 12 ensure that the authorization conditions, in icle 5 of Directive 2006/88/EC, are fulfilled by all ization is granted. The shortcomings described in e audit report will be corrected in connection with ty records, disinfection records and biosecurity ted in connection with this review, and this work islation has been revised, the presently authorized about the changes in the authorization conditions, ce with these conditions will be included in the he CA, the results of the control of these conditions the official control report and CCA will audit that A (see Recommendation No 3).

Medlemsstatens svar

ollaboration review the register of the authorized only those APBs that meet the conditions are

Medlemsstatens svar

DVFA began to use DIKO, which is the digital visits performed by CA on APBs shall use this ent the control and make the control report. The written tion of control visits and administrative control in Page: 1 NCO)/2011-6411-MR of an audit carried to aquaculture animals Competent Authority which is the control object register with among APBs and information on frequency of control staff performs the control, etc. The written on in KOR is attached. ly being revised to include the control of the conditions described in Recommendation No 1 written in the control reports. bed under recommendation 1 ­ 4 have been it CA.

Medlemsstatens svar

aborating a written procedure to ensure that CA he general principles of certification and all requirements cates annexed to Regulation 1251/2008. Page: 2