Opinion of the Scientific Panel on Biological Hazards on the request from the Commission related to Campylobacter in animals and foodstuffs 1

The EFSA Journal (2005) 173 1-10 Campylobacter in animals and foodstuffs Opinion of the Scientific Panel on Biological Hazards on the request from the Commission related to Campylobacter in animals and foodstuffs 1 (Question N EFSA-Q-2003-081) Adopted on 27th of January 2005 SUMMARY Campylobacteriosis represent an important public health problem with considerable socio-economic impact in the EU. The primary reservoir of thermophilic Campylobacter, the etiological agents of campylobacteriosis, is the alimentary tract of wild and domesticated birds and mammals, and the infective dose seems to be small. Foods represent a significant risk in regard to human campylobacteriosis. This opinion assesses foodborne routes of campylobacteriosis, and identifies possible control options as well as data gaps which require attention. Poultry meat products appear to be a major source of campylobacteriosis, through crosscontamination to ready-to-eat (RTE) foods and through direct hand-to-mouth transfer during food preparation, and to a lesser extent from the consumption of undercooked poultry meat. Reducing the proportion of Campylobacter infected poultry flocks and/or reducing the numbers of Campylobacter in live poultry and on poultry carcasses will lower the risk to consumers considerably. Reducing the proportion of Campylobacter infected poultry flocks can be achieved by the application of strict biosecurity measures in primary production. The setting of performance objectives (PO) or targets in poultry production is recommended. Means for verifying that these POs are met should also be implemented. Setting microbiological standards for Campylobacter in poultry meat products at retail level appear not to be cost-effective as this would imply unnecessary testing of end products. 1 For citation purposes : Opinion of the Scientific Panel on Biological Hazards on «Campylobacter in animals and foodstuffs», The EFSA Journal (2005) 173 1-10 http://www.efsa.eu.int 1

Meat from pigs and ruminant are considered to present a relative low risk to the consumers in regard to campylobacteriosis, but undercooked offal from these food animals is likely to present a considerably risk. The consumption of raw milk and contaminated drinking water are important causes of outbreaks of campylobacteriosis. Milkborne campylobacteriosis can be controlled by proper pasteurisation or by applying alternative measures that eliminate Campylobacter. Otherwise consumers of untreated raw milk should be aware of the associated risk. A reliable safety assurance system in water supply systems is important to exclude Campylobacter from drinking water. Otherwise consumers of water from uncontrolled sources should be aware of the associated risk. Fresh produce contaminated with Campylobacter if consumed raw present a considerable risk to the consumers. Consequently, implementing GAP and GHP throughout the food chain and avoiding the use of untreated faecally contaminated water for irrigation and washing is essential. Bivalve molluscs contaminated with Campylobacter, including oysters, can represent a risk to the consumer. The application of GHP and the control of water quality before harvesting and during depuration are important. Cross-contamination of RTE foods by food handlers and consumers, particularly from raw poultry meat, is a major risk factor for campylobacteriosis. Thus, hygienic food handling practices in the home and in the food services should be promoted. TABLE OF CONTENTS SUMMARY...1 TABLE OF CONTENTS...2 BACKGROUND...3 TERMS OF REFERENCE...4 ASSESSMENT...4 CONCLUSIONS...5 RECOMMENDATIONS...9 SCIENTIFIC PANEL MEMBERS...10 ACKNOWLEDGEMENT...10 ANNEX...10 http://www.efsa.eu.int 2

BACKGROUND The Scientific Committee on Veterinary Measures relating to Public Health (SCVMPH) issued on 12 April 2000 an opinion on foodborne zoonoses. In this opinion the Committee identified Campylobacter spp. as one of the public health priorities among the food borne zoonotic pathogens. The opinion includes a short risk profile on thermophilic Campylobacter spp. in foodstuffs as well as a description of available risk management options. The Committee also addressed Campylobacter spp. in its opinion of 26-27 March 2003 on the human health risk caused by the use of fluoroquinolones in animals. According to the Commission s report on zoonoses 2 a total of 156 232 cases of human campylobacteriosis were reported by 13 Member States in 2001. This number of cases has been increasing over the last 6 years. In some Member States the number of campylobacteriosis cases is already exceeding that of salmonellosis. The Community legislation on food hygiene and control of zoonoses is currently under revision. Proposals for the new legislation are being dealt with in the Council, the European Parliament and the Commission. In this context specific rules could be laid down concerning Campylobacter spp. The central idea of the proposed Regulation on the control of salmonella and other specified food-borne zoonoses is the setting of pathogen reduction targets along the food chain, mainly for animal populations, and the establishment of national control programmes in order to meet these targets. In this context targets for Campylobacter could be established. However, there is still a lack of knowledge of the best available means of preventing the Campylobacter infections at the primary production level, which may hamper the setting of realistic targets and drawing up of control programmes. The proposed Directive on the monitoring of zoonoses and zoonotic agents contains a possibility to harmonise the monitoring of certain zoonotic agents, when it is necessary to make the data collected easier to compile and compare. This option may be used to gather more detailed information on Campylobacter spp. in order to fill data gaps. The Commission has also started a revision of the microbiological criteria in Community legislation. The revised criteria would cover all foodstuffs as well as the whole production and distribution chain in line with the proposed new hygiene legislation. Criteria would be set for food products at different stages of the manufacturing process as well as for products on the market. During the discussions with the Member States and the consultation of the stakeholders, wishes to set criteria for Campylobacter spp. in raw milk, poultry carcasses as well as live bivalve molluscs have been expressed. 2 European Commission : Trends and sources of zoonotic infections in animals, feedingstuffs, food and man in the European Union and Norway in 2001 http://www.efsa.eu.int 3

TERMS OF REFERENCE In view of the above, the Commission asks the European Food Safety Authority to deliver a scientific opinion on Campylobacter spp. in animals and foodstuffs. In doing so, it is asked in particular to: identify categories of foodstuffs where Campylobacter spp. represent a significant risk to public health; identify possible control options to reduce this risk along the food chain, and evaluate their effectiveness, with special reference to the measures taken at the primary production and the setting of microbiological criteria; identify the gaps in available data together with the best means of collecting this information. ASSESSMENT In this opinion, foodborne routes of campylobacteriosis have been assessed. It is, however, acknowledged that other routes of infection exist, including person-to person transmission, contact with dogs and cats, occupational contact with food animals, and environmental and recreational water. These non-food routes contribute to an unknown, but most probably considerable, extent to the incidence of human cases, but are not addressed in this opinion. The contribution of the various food and non-food sources to the incidence of campylobacteriosis within EU is country- and season-dependent, due to various factors such as climate, consumption patterns, drinking water distribution, food production systems, as well as the degree of implementation of control measures. The epidemiology of thermophilic Campylobacter, the etiological agents of campylobacteriosis, is complex. The principal reservoirs are the alimentary tract of wild and domesticated birds and mammals. Due to its wide occurrence in nature, Campylobacter are frequently isolated from a variety of foodstuffs. Campylobacter may be spread to humans by direct contact with infected animals or contaminated animal carcasses, or indirectly through the ingestion of contaminated food or water. Direct spread from one person to another can occur, but is uncommon in industrialised countries. Most cases are sporadic, while common-source outbreaks usually account for a rather small proportion of cases. The dose-response relationship appears to be complex. However, it is considered that only a few hundred bacteria may cause illness. The incubation period for campylobacteriosis is one day to one week and infections usually result in mild to moderate symptoms including diarrhoea (frequently with blood in the faeces), abdominal pain, fever, headache, nausea and/or vomiting. Symptoms, which are usually self-limiting, may last one day to one week, and in up to 20% of cases, illness lasts for more than a week. More invasive manifestations may occur in up to 1% of patients, mostly involving the elderly or very young individuals. A fatal outcome is rare and is usually confined to very young or elderly patients or the immunocompromised suffering from an invasive infection. Chronic sequelae, including reactive arthritis and Guillain-Barré syndrome, sometimes occur. Normally, about 90% of human isolates are C. jejuni, while C. coli accounts for most of the remaining cases. Although antimicrobial treatment usually is not indicated for enteritis of moderate severity, such http://www.efsa.eu.int 4

treatment is sometimes justified. Antimicrobial resistance in Campylobacter, especially to fluoroquinolones, is an increasing medical problem. The true incidence of campylobacteriosis is thought to be considerably higher than the number of reported cases (from 8 to 100 times as high). The high incidence of Campylobacter spp. diarrhoea as well as its duration and possible sequelae, makes campylobacteriosis very important from a public health perspective with significant socio-economic impact. The following summarizes the conclusions when assessing the various food categories in relation to their importance as a risk to public health in regard to campylobacteriosis, and identifying possible control options and data gaps. The conclusions are presented according to the terms of reference. Finally, recommendations drawn from the scientific assessment are presented. CONCLUSIONS TOR#1 (Identify categories of foodstuffs where Campylobacter spp. represents a significant risk to public health) Foods represent a significant risk in regard to human campylobacteriosis. It is, however, acknowledged that other routes of infection exist, including environmental and recreational water, occupational contact with food animals, contact with cats and dogs and person-to-person-transmission. These non-food routes contribute to an unknown, but most probably considerable, extent to the incidence of human cases. Raw drinking milk and contaminated drinking water are important causes of outbreaks of campylobacteriosis. Most waterborne outbreaks have been the result of failures in the water supply system. Poultry meat appears to be a major source of sporadic campylobacteriosis. Sporadic campylobacteriosis accounts for the majority of cases. Other food categories identified as a risk in relation to sporadic campylobacteriosis include barbecued foods, untreated drinking water (e.g., from private wells, rivers and streams), undercooked pork, poorly-pasteurized and raw milk. Poultry meat can be contaminated with high numbers of Campylobacter and can infect people through cross-contamination to ready-to-eat foods and direct hand-tomouth transfer during food preparation, and to a lesser extent from the consumption of undercooked poultry meat. Most Campylobacter contaminated poultry carcasses derive from birds colonised with Campylobacter. Campylobacter negative birds can be infected during transportation to the abattoir from contaminated crates, or during processing from other infected flocks via the processing environment. Modelling studies and risk assessments suggest that carcasses cross-contaminated in this way present a minor risk to the consumers. Free-range and organic poultry flocks appear to have higher flock prevalence of Campylobacter spp. than industrilalized poultry due to greater exposure to the http://www.efsa.eu.int 5

outdoor environment, and therefore could represent a higher risk in regard to campylobacteriosis. Although pigs and ruminants are frequent carriers of Campylobacter, meat from these animals is less commonly contaminated with Campylobacter and with lower numbers than poultry and therefore probably represents a much lower risk than poultry meat to consumers. However, offal (e.g., liver, kidney, heart) is much more commonly contaminated with Campylobacter, and is more likely to present a considerable risk to public health in regard to campylobacteriosis as a consequence of undercooking or cross-contamination to RTE food products. Dairy products (e.g., cheeses, yoghurt, butter) other than raw drinking milk are not considered a significant risk to public health in regard to campylobacteriosis. The importance of fresh produce as a source of campylobacteriosis is unclear. Fresh produce can present a risk to public health in regard to campylobacteriosis as a consequence of using contaminated irrigation or washing water. Seafood is not a major cause of campylobacteriosis. However, Campylobacter can be present in seawater and thereby contaminate seafood, particularly bivalve molluscs, and consequently bivalve molluscs, for example oysters, can present a risk to public health in regard to campylobacteriosis. Ready-to-eat (RTE) foods, including fresh produce, represent an important risk to public health if contaminated with Campylobacter as a result of cross-contamination, for example from raw poultry meat during food handling. TOR#2 (Identify possible control options to reduce this risk along the food chain, and evaluate their effectiveness, with special reference to the measures taken at the primary production and the setting of microbiological criteria) Reducing the proportion of Campylobacter infected poultry flocks and/or reducing numbers of Campylobacter in live poultry and on poultry carcasses will lower the risk to consumers considerablyreducing the proportion of Campylobacter infected poultry flocks can be achieved by the application of strict biosecurity measures in primary production, including the use of non-contaminated drinking water. Reducing numbers of Campylobacter on poultry carcasses can be achieved by reducing faecal spread during slaughter and further processing and by using appropriate physical (e.g., freezing) or chemical decontamination techniques where permitted. For poultry, the setting of a performance objective (PO)/target for Campylobacter is a relevant option. In primary production, the PO/target would be the proportion of infected flocks (flock prevalence). In the slaughter plant, the PO/target could be the proportion of contaminated final raw carcasses and/or the numbers of Campylobacter on them. Verifying that these POs are achieved will require various approaches, including implementing monitoring programmes in primary production and at slaughterhouse level. Means for verifying that these POs are met will be needed, such as monitoring of flock prevalence. http://www.efsa.eu.int 6

Setting microbiological standards for Campylobacter in poultry meat products at retail level appear not to be cost-effective as this would imply unnecessary testing of end products. An important measure for minimising Campylobacter contamination of raw milk is the application of GHP during milking. Milkborne campylobacteriosis can be adequately controlled by proper pasteurization of milk for direct consumption or by applying alternative measures that eliminate Campylobacter, and by avoiding recontamination of treated milk.untreated raw milk. A reliable safety assurance system in water supply systems is important to exclude Campylobacter from the final product. To control Campylobacter in RTE foods, the application of GHP and HACCP principles, as well as the implementation of proper food handling practices, particular avoiding cross-contamination, is essential. Because sewage and animal farm waste can contain high numbers of Campylobacter, an important control option is to prevent spread from these sources to drinking water and primary food production. Preventing contamination of fresh produce with Campylobacter relies on the application of GAP and GHP throughout the production chain. Important control options include not using untreated faecally contaminated water for irrigation and washing. Preventing contamination of bivalve molluscs, including oysters, with Campylobacter relies on the application of GHP and the control of water quality before harvesting and during depuration. A mild thermal process for sea foods is another control option. Consumers can considerably reduce the risk of campylobacteriosis by applying hygienic food handling practices, particularly avoiding cross-contamination of RTE foods from raw poultry; by not drinking raw milk; by thoroughly cooking poultry, red meat, offal and bivalve molluscs; and by not drinking water from uncontrolled sources. TOR#3 (Identify the gaps in available data together with the best means of collecting this information) Hazard characterization: The reporting of human campylobacteriosis in the EU is not harmonized and therefore data are not comparable between countries. There is a need for improved application of new typing methods and interpretation of typing and surveillance data in order to recognise hidden outbreaks. The relationship between number of Campylobacter ingested and risk of disease, i.e., the dose-response relationship, should be clarified. http://www.efsa.eu.int 7

Better information on the incidence of foodborne disease caused by Campylobacter alone or in conjunction with other pathogens in the EU would be helpful when setting priorities in risk management. More information is needed on role of Campylobacter in the epidemiology of e.g., Guillain-Barre syndrome and reactive arthritis. There is a lack of knowledge regarding pathogenicity and virulence of various Campylobacter types in humans, especially those linked to post-infection sequaelae. There is a need for clarification of the relationship between fluoroquinolone use in animals and the incidence of human infections with quinolone-resistant Campylobacter. Exposure assessment: There is a need for a better understanding of the relative contribution of various food categories (poultry, red meat, raw milk, drinking water, shellfish and fresh produce) to the incidence of campylobacteriosis (pathogen accounts, comparison of food and animal Campylobacter strains with human strains in order to better understand the epidemiology of human campylobacteriosis). To contribute to a quantitative microbiological risk assessment, there is a need for more qualitative and quantitative data on the occurrence of Campylobacter spp. in various steps in the different food production chains, particularly poultry, and the effect of various production and processing systems,. There is a need for a better understanding of the contribution of various non-food routes to the incidence of campylobacteriosis There is a need for a better understanding of the dynamics of Campylobacter in primary poultry production and the factors that influence them, including interaction with gut flora, the effect of competititive exclusion treatment, and genetic resistance, as well as the possibility and potential importance of vertical transmission of Campylobacter. There is a need for more knowledge regarding the survival of Campylobacter in poultry slaughterhouses: Are particularly robust strains selected in the slaughterhouse, which predominate on carcasses that reach the consumers? Are these strains more or less pathogenic for humans? There is a lack of information concerning the importance of biofilms in the occurrence and persistence of Campylobacter in the environment and whether they survive inside protozoa. There is a need for more information on the viable but non-culturable (VBNC) forms of Campylobacter, and whether they are capable of initiating infections in animals and/or humans. There is a need for a better understanding of the effects of new packaging technologies and of different chilling and freezing treatments on the occurrence of http://www.efsa.eu.int 8

Campylobacter (incl. VBNC forms), for example the effect of modified atmospheres on the survival of Campylobacter during retail shelf-life and determination of the optimum rate of freezing, freezing temperature and frozen storage time for reduction of Campylobacter in poultry). The development of a vaccine against Campylobacter in poultry needs to be further pursued. RECOMMENDATIONS The reporting of incidences of campylobacteriosis in humans, prevalences of Campylobacter in the food chain and methods applied should be harmonized in the EU in order to make data comparable both between and within countries and over time. Sampling schemes, qualitative and quantitative methods for detection, as well as typing methods for Campylobacter should be harmonized. For monitoring purposes, the application of standard sampling schemes and standard methods should be promoted. In order to establish the relative contribution of various food and non-food categories to the incidence of campylobacteriosis, epidemiological investigations supported by laboratory data and specific epidemiological monitoring of campylobacteriosis should be strengthened. New approaches to determine the dose-response relationship should be developed. More information on the incidence of post-infection sequelae should be sought. More research should be undertaken regarding pathogenicity and virulence of various Campylobacter types in humans, especially those linked to post-infection sequaelae. More research should be undertaken regarding the dynamics of Campylobacter in poultry production and the factors that influence them. The setting and use of performance objectives (POs)/targets in poultry production should be encouraged. POs/targets could be the proportion of infected poultry flocks (flock prevelance). Means for verifying that the POs set are met should be encouraged, such as monitoring of flock prevalence. The application of strict biosecuritiy measures in primary poultry production, including the use of non-contaminated drinking water, should be encouraged and promoted. Reducing the proportion of Campylobacter infected poultry flocks can be achieved by the application of strict biosecurity measures The effectiveness of implemented risk management tools should be validated through monitoring programmes. These should be established at relevant stages in the food production chain where a http://www.efsa.eu.int 9

particular measure for the control of Campylobacter has been implemented. The concentration and prevalence of the pathogen, as well as its impact on the number of human cases should be addressed. Relevant points for monitoring could be the proportion of infected flocks, poultry carcasses and raw poultry meat products at retail. Milk for direct consumption should be pasteurized or treated by alternative means that ensure control of Campylobacter. The application of GAP and GHP in the production of fresh produce that is to be consumed raw is strongly recommended to prevent contamination of products with Campylobacter. Untreated faecally contaminated water should not be used for irrigation and only potable water should be used for washing. For pre-cut/pre-sliced fresh produce, a HACCP principles should be implemented. To prevent contamination of bivalve molluscs, including oysters, with Campylobacter, the application of GHP and the control of water quality before harvesting and during depuration are strongly recommended. To control Campylobacter in RTE foods, the application of GHP and HACCP principles, as well as the implementation of proper food handling practices, particular the avoidance of cross-contamination should be promoted. Food handlers and consumers should be informed about how they can reduce the risk of campylobacteriosis, including the application of hygienic food handling practices, particularly avoiding cross-contamination of RTE foods from raw poultry; not drinking raw milk; thoroughly cooking poultry, red meat, offal and bivalve molluscs; and not drinking water from uncontrolled sources. SCIENTIFIC PANEL MEMBERS Herbert Budka, Sava Buncic, Pierre Colin, John D Collins, Christian Ducrot, James Hope, Mac Johnston, Günter Klein, Hilde Kruse, Ernst Lücker, Simone Magnino, Riitta Liisa Maijala, Antonio Martínez López, Christophe Nguyen-The, Birgit Noerrung, Servé Notermans, George-John E Nychas, Maurice Pensaert, Terence Roberts, Ivar Vågsholm, Emmanuel Vanopdenbosch. ACKNOWLEDGEMENT The Scientific Panel on Biological Hazards wishes to acknowledge the contribution of the working group that prepared the draft opinion: H. Kruse, C. Nguyen-Thè, A.M. López, J. Corry, G. Klein, B. Noerrung, G. Salvat and I. Vagsholm. ANNEX http://www.efsa.eu.int/science/biohaz/biohaz_opinions/opinion_annexes/867_en.html http://www.efsa.eu.int 10

Annex to The EFSA Journal (2005), 173, 1-105, Scientific Report on Campylobacter in animals and foodstuffs Scientific Report of the Scientific Panel on Biological Hazards on the request from the Commission related to Campylobacter in animals and foodstuffs 1 (Question N EFSA-Q-2003-081) Adopted on 27 th of January 2005 SUMMARY Campylobacteriosis represent an important public health problem with considerable socio-economic impact in the EU. The primary reservoir of thermophilic Campylobacter, the etiological agents of campylobacteriosis, is the alimentary tract of wild and domesticated birds and mammals, and the infective dose seems to be small. Foods represent a significant risk in regard to human campylobacteriosis. This opinion assesses foodborne routes of campylobacteriosis, and identifies possible control options as well as data gaps which require attention. Poultry meat products appear to be a major source of campylobacteriosis, through crosscontamination to ready-to-eat (RTE) foods and through direct hand-to-mouth transfer during food preparation, and to a lesser extent from the consumption of undercooked poultry meat. Reducing the proportion of Campylobacter infected poultry flocks and/or reducing the numbers of Campylobacter in live poultry and on poultry carcasses will lower the risk to consumers considerably. Reducing the proportion of Campylobacter infected poultry flocks can be achieved by the application of strict biosecurity measures in primary production. The setting of performance objectives (PO) or targets in poultry production is 1 For citation purposes: Scientific Report of the Scientific Panel on Biological Hazards on the request from the Commission related to Campylobacter in animals and foodstuffs. Annex to The EFSA Journal (2005) 173, 1-105 http://www.efsa.eu.int 1 of 105

recommended. Means for verifying that these POs are met should also be implemented. Setting microbiological standards for Campylobacter in poultry meat products at retail level appear not to be cost-effective as this would imply unnecessary testing of end products. Meat from pigs and ruminant are considered to present a relative low risk to the consumers in regard to campylobacteriosis, but undercooked offal from these food animals is likely to present a considerable risk. The consumption of raw milk and contaminated drinking water are important causes of outbreaks of campylobacteriosis. Milkborne campylobacteriosis can be controlled by proper pasteurisation or by applying alternative measures that eliminate Campylobacter. Otherwise consumers of untreated raw milk should be aware of the associated risk. A reliable safety assurance system in water supply systems is important to exclude Campylobacter from drinking water. Otherwise consumers of water from uncontrolled sources should be aware of the associated risk. Fresh produce contaminated with Campylobacter if consumed raw present a considerable risk to the consumers. Consequently, implementing GAP and GHP throughout the food chain and avoiding the use of untreated faecally contaminated water for irrigation and washing is essential. Bivalve molluscs contaminated with Campylobacter, including oysters, can represent a risk to the consumer. The application of GHP and the control of water quality before harvesting and during depuration are important. Cross-contamination of RTE foods by food handlers and consumers, particularly from raw poultry meat, is a major risk factor for campylobacteriosis. Thus, hygienic food handling practices in the home and in the food services should be promoted. http://www.efsa.eu.int 2 of 105

TABLE OF CONTENTS SUMMARY...1 TABLE OF CONTENTS...3 BACKGROUND...6 TERMS OF REFERENCE...7 ASSESSMENT...7 1. INTRODUCTION...7 2. CAMPYLOBACTER- BACTERIOLOGICAL ASPECTS...8 2.1. The organism...8 2.2. Biodiversity of Campylobacter in food animals...11 2.3. Methods for isolation and identification of thermophilic Campylobacter...12 2.3.1. Characteristics important for detection and isolation...12 2.3.2. Selective media...12 2.3.3. Modeling survival...14 3. CAMPYLOBACTERIOSIS AS A HUMAN DISEASE...15 3.1. Hazard characterization...15 3.2. Epidemiology...18 3.2.1. Risk factors...19 3.2.1.1. Risk factors for outbreaks of campylobacteriosis...19 3.2.1.2. Risk factors for sporadic cases of campylobacteriosis...21 3.2.2. Antimicrobial resistance...24 4. OCCURRENCE AND CONTROL IN POULTRY ALONG THE FOOD CHAIN...25 4.1. Gallus gallus...25 4.1.1. Primary production...25 4.1.1.1. Prevalences in flocks...25 4.1.1.2. Within-flock prevalences...25 4.1.1.3. Age of colonization...25 4.1.1.4. Risk factors for infection at farm level...25 Deleted: 18 Deleted: 20 Deleted: 23 Deleted: 24 Deleted: 24 Deleted: 24 Deleted: 24 Deleted: 24 http://www.efsa.eu.int 3 of 105

4.1.1.5. Risk factors for contamination during catching and transportation...27 4.1.1.6. Control options...28 4.1.2. Processing...29 4.1.2.1. Campylobacter contamination of carcasses during processing...29 4.1.2.2. Cross-contamination between Campylobacter positive and negative flocks...31 4.1.2.3. Control options...32 4.1.3. Retail/consumer phase...36 4.1.3.1. Retail phase...36 4.1.3.2. Consumer phase...36 4.1.3.3. Control options at retail/consumer level...38 4.1.4. Risk assessment focusing on comparison of different control options...39 4.1.5. Possible setting of microbiological criteria or objectives...39 4.2. Other poultry species...40 4.2.1. Epidemiology of Campylobacter infections in turkey...40 4.2.2. Epidemiology of Campylobacter infections in ducks, geese and other poultry species...42 5. OCCURRENCE AND CONTROL IN PORK AND OTHER RED MEAT...43 Deleted: 27 Deleted: 28 Deleted: 31 5.1. Occurrence of Campylobacter in pork and other red meat...43 5.2. Control options...44 6. OCCURRENCE AND CONTROL IN MILK AND DAIRY PRODUCTS...44 6.1. Occurrence of Campylobacter in milk...44 6.2. Occurrence of Campylobacter in dairy products...45 6.3. Occurrence of Campylobacter in pasteurized milk...46 6.4. Control options...46 7. OCCURRENCE AND CONTROL IN SEAFOOD...46 7.1. Occurrence of Campylobacter in bivalve molluscs...46 7.2. Other seafoods...48 8. OCCURRENCE AND CONTROL IN READY-TO-EAT-FOODS...48 Deleted: 45 Deleted: 47 8.1. Occurrence of Campylobacter in RTE foods...48 8.2. Control options...50 9. OCCURRENCE AND CONTROL IN FRESH PRODUCE...51 9.1. Background...51 9.2. Occurrence of Campylobacter spp. in fresh produce...52 http://www.efsa.eu.int 4 of 105 Deleted: 51

9.3. Control options...54 10. OCCURRENCE AND CONTROL IN DRINKING WATER...55 10.1. Occurrence of Campylobacter in drinking water...55 10.2. Control options...56 11. CONCLUSIONS...57 12. RECOMMENDATIONS...61 13. REFERENCES...63 SCIENTIFIC PANEL MEMBERS...99 ACKNOWLEDGEMENT...99 ANNEX 1 GLOSSARY...100 ANNEX 2 TABLES...102 Deleted: 62 Deleted: 98 Deleted: 98 Deleted: 99 Deleted: 101 http://www.efsa.eu.int 5 of 105

BACKGROUND The Scientific Committee on Veterinary Measures relating to Public Health (SCVMPH) issued on 12 April 2000 an opinion on foodborne zoonoses 1. In this opinion the Committee identified Campylobacter spp. as one of the public health priorities among the food borne zoonotic pathogens. The opinion includes a short risk profile on thermophilic Campylobacter spp. in foodstuffs as well as a description of available risk management options. The Committee also addressed Campylobacter spp. in its opinion of 26-27 March 2003 on the human health risk caused by the use of fluoroquinolones in animals. According to the Commission s report on zoonoses 2 a total of 156 232 cases of human campylobacteriosis were reported by 13 Member States in 2001. This number of cases has been increasing over the last 6 years. In some Member States the number of campylobacteriosis cases is already exceeding that of salmonellosis. The Community legislation on food hygiene and control of zoonoses is currently under revision. Proposals for the new legislation are being dealt with in the Council, the European Parliament and the Commission. In this context specific rules could be laid down concerning Campylobacter spp. The central idea of the proposed Regulation on the control of salmonella and other specified food-borne zoonoses is the setting of pathogen reduction targets along the food chain, mainly for animal populations, and the establishment of national control programmes in order to meet these targets. In this context targets for Campylobacter could be established. However, there is still a lack of knowledge of the best available means of preventing the Campylobacter infections at the primary production level, which may hamper the setting of realistic targets and drawing up of control programmes. The proposed Directive on the monitoring of zoonoses and zoonotic agents contains a possibility to harmonise the monitoring of certain zoonotic agents, when it is necessary to make the data collected easier to compile and compare. This option may be used to gather more detailed information on Campylobacter spp. in order to fill data gaps. 1 Scientific Committee on Veterinary Measures relating to Public Health (SCVMPH) (2000) Opinion on food-borne zoonoses. European Commission. 2 European Commission : Trends and sources of zoonotic infections in animals, feedingstuffs, food and man in the European Union and Norway in 2001 http://www.efsa.eu.int 6 of 105

The Commission has also started a revision of the microbiological criteria in Community legislation. The revised criteria would cover all foodstuffs as well as the whole production and distribution chain in line with the proposed new hygiene legislation. Criteria would be set for food products at different stages of the manufacturing process as well as for products on the market. During the discussions with the Member States and the consultation of the stakeholders, wishes to set criteria for Campylobacter spp. in raw milk, poultry carcasses as well as live bivalve molluscs have been expressed. TERMS OF REFERENCE In view of the above, the Commission asks the European Food Safety Authority to deliver a scientific opinion on Campylobacter spp. in animals and foodstuffs. In doing so, it is asked in particular to: identify categories of foodstuffs where Campylobacter spp. represent a significant risk to public health; identify possible control options to reduce this risk along the food chain, and evaluate their effectiveness, with special reference to the measures taken at the primary production and the setting of microbiological criteria; identify the gaps in available data together with the best means of collecting this information. ASSESSMENT 1. INTRODUCTION The zoonosis campylobacteriosis, which is caused by thermophilic Campylobacter spp., is an important public health problem in most areas of the world. According to the EU reporting of zoonoses, the incidence of campylobacteriosis is significant throughout EU and there has been a general upward in the incidence over the last decade (EC, 2004). In several Member States, the incidence of campylobacteriosis has surpassed that of salmonellosis in recent years and has become the most commonly reported bacterial gastrointestinal disease (EC, 2004). Typically, about 90% of the isolates from human campylobacteriosis cases are identified as C. jejuni (EC, 2002), while C. coli accounts for most of the remaining cases. Thermophilic Campylobacter spp., the etiological agents of campylobacteriosis, are widespread in nature (Jones, 2001). The principal reservoirs are the alimentary tracts of wild and domesticated birds and mammals. Consequently thermophilic Campylobacter spp., especially C. jejuni and C. coli, are commonly isolated from water sources, food animals such as poultry, cattle, pigs, and sheep, as well as from cats and dogs (Jones, 2001; FAO/WHO, 2002). C. jejuni is predominantly associated with poultry, but can also be isolated from cattle, sheep, goats, pigs, dogs and cats, while C. coli is predominantly found in pigs, but can also be isolated from http://www.efsa.eu.int 7 of 105

poultry, cattle and sheep (EC, 2004; Nielsen et al., 1997; Pezzotti et al., 2003). In animals, thermophilic Campylobacter spp. seldom cause clinical disease (WHO, 2000). As a result of the widespread occurrence of thermophilic Campylobacter spp. in nature and in food animals, the bacteria can readily contaminate various foodstuffs, including poultry, beef, pork, other meat products, raw milk and milk products, and, less frequently, fish and fish products, mussels and fresh vegetables (Jacobs- Reitsma, 2000). Because Campylobacter spp. can be transferred from animals to man, human campylobacteriosis is considered to be a zoonosis (WHO, 2000). Campylobacter spp. may be transferred to humans by direct contact with contaminated animals or animal carcasses, or indirectly through the ingestion of contaminated food or drinking water (FAO/WHO, 2002). While transmission routes in low-income societies are complex and multifactorial, campylobacteriosis in industrialised countries is primarily a foodborne disease, with poultry recognized as a principal source (Friedman et al., 2000; WHO, 2001). Although our understanding of the risk factors for human infection has improved, the relative contribution of the different factors to the total incidence of disease remains to be established. The contribution of the various food and non-food sources to the incidence of campylobacteriosis within EU will be country- and time-dependent due to factors such as climate, food consumption patterns, including the drinking of water from uncontrolled sources, food production systems, degree of implementation of control measures, etc. In this opinion, foodborne routes of campylobacteriosis are assessed. It is, however, acknowledged that other routes of infection exist, including environmental and recreational water, occupational contact with food animals, contact with cats and dogs and person-to-person-transmission. These non-food routes contribute to an unknown, but probably significant proportion of human cases. Nevertheless, as the foodborne routes are considered to be responsible for the vast majority of cases, these routes are further addressed in this opinion, while the non-food routes are not further discussed. Following a short review of the etiological agent and associated human disease, as well as methods for detection, the various food categories identified as representing a considerable risk with regard to campylobacteriosis are addressed, focusing on exposure assessment and control options. Various aspects concerning monitoring and the possible establishment of criteria are also discussed followed by a presentation of conclusions according to the terms of reference. Finally, recommendations drawn from the scientific assessment are presented. 2. CAMPYLOBACTER- BACTERIOLOGICAL ASPECTS 2.1. The organism Campylobacter species are Gram negative, oxidase positive, usually curved or spiral rod-shaped bacteria, 0.2-0.5µm wide and 0.5-8µm long. Most http://www.efsa.eu.int 8 of 105

species possess a polar flagellum, at one or both ends which gives them a very characteristic "cork-screw" motility. The most important species of Campylobacter are the thermophilic species: C. jejuni ssp. jejuni, C. coli and C. lari (formerly known as "nalidixic acid resistant thermophilic Campylobacter spp. - NARTC ). Other species which are known to cause human illness are C. upsaliensis, C. fetus ssp. fetus and C. jejuni ssp. doylei. Most physiological/biochemical, epidemiological and survival information concerns C. jejuni, as this species is the most frequently recovered in human disease. UPTC (urease positive thermophilic Campylobacter spp) are variants of C. lari often associated with seawater environments and consequently found in shellfish and sea birds (Megraud et al., 1988; Jones, 2001; On, 2001). Campylobacter spp. are relatively inactive biochemically, obtaining their energy from amino acids or tricarboxylic acid cycles intermediates rather than carbohydrates. This makes them difficult to speciate by use of classical biochemical tests (On, 1996), so they are often identified to species level by use of PCR-based methods (Linton et al., 1997; Bolton et al., 2002; On and Jordan, 2003). Compared with other important foodborne pathogens, such as Salmonella species, Campylobacter spp. seem ill-equipped to survive outside an animal host. They require a microaerobic atmosphere (ca. 5% oxygen and 10% carbon dioxide) and cannot multiply below about 30 o C. However, even at 4 o C, low-level metabolic activity can be detected, suggesting that cell integrity is maintained (Park, 2002). Media for their cultivation usually contain supplements such as blood, FBP (see section 2.3) charcoal, haemin or haematin to protect them from toxic oxygen derivatives (e.g., hydrogen peroxide, singlet oxygen, superoxide). Jacobs-Reitsma (2000) reviewed published information on survival of Campylobacter spp. in foods. Campylobacter spp. survive poorly in dry or acid conditions, and in sodium chloride above 2%. Survival in foods is better at chill temperatures than higher (e.g., ambient), and freezing inactivates many, but not all those bacteria present. Campylobacter spp. are relatively sensitive to heat and irradiation, and so can readily be inactivated during cooking (ICMSF, 1996). Their sensitivity to environmental stresses seems to be confirmed by their lack of genes analogous to those in other bacteria, enabling physiological adaptation to adverse environments - e.g., oxidative stress, osmoregulation, starvation/stationary phase, heat and cold shock (Park, 2002). However, there is some indication that they may be capable of other strategies, which explain their survival. Kelly et al. (2001) and Martinez-Rodriguez et al. (2004) showed that stationary phase cells of C. jejuni fluctuate in numbers and in resistance to heat, deoxycholate and oxidative stress. http://www.efsa.eu.int 9 of 105

Under prolonged periods of storage and starvation the organism changes to a coccoid form which has been described as "viable but non-culturable" (VBNC). Although this form cannot be grown using normal media, there are reports that it can infect animals. However, others have suggested that this could be due to regrowth of a small proportion of cells which never lost culturability, and that these cells may merely be a degenerate form (Bovill and Mackey, 1997; Kell et al. 1998). Reduced ability to form coccoid cells was reported by Kelly et al. (2001) not to affect survival of C. coli under non-growth conditions. A new observation is that Campylobacter spp. can reside within amoebae (Dahlgren et al., 2003), providing an explanation of the survival and persistence of Campylobacter in water particular at low temperatures. Buswell et al. (1998) and Trachoo et al, (2002) found that Campylobacter spp. survive longer in biofilms than as planktonic cells in water, and are also more resistant to biocides, although they are still killed within 45 sec by 50 ppm chlorine (Trachoo and Frank, 2002). On the other hand, Dykes et al. (2003) concluded that Campylobacter spp. grown as a biofilm were less stress resistant than planktonic cells. There is as yet no evidence to suggest that human infections occur due to survival of Campylobacter spp. in biofilms or inside protozoa in chlorinated water supplies for human consumption (Percival and Walker, 1999). Similarly, chlorinated water supplies for poultry appear to be effective in preventing infection from this source (Kapperud et al., 1993; Pearson et al., 1993). Whatever the mechanism that enables thermophilic Campylobacter spp. to survive in the environment, they are well adapted to colonisation of animals, especially birds, intestines, and they are a normal inhabitant of many species of birds. They are usually found close to the intestinal wall, where the oxygen level is close to their optimum, and their mode of motility enables them to move through the viscous intestinal mucus layer. Understanding the epidemiology of Campylobacter infections has been hampered by the lack of convenient and comprehensive typing systems, analogous to those used for Salmonella serovars, and also by the relative rarity of outbreaks of campylobacteriosis. Two serotyping schemes were developed in Canada during the 1980s, the Penner scheme based on soluble heat-stable antigens, and the Lior scheme on heat-labile antigens. A modified and extended Penner scheme, in combination with phage typing, is used by the UK Health Protection Agency Laboratory of Enteric Pathogens, to provide a relatively economic and rapid method for use in surveillance of human infection (Newell et al., 2000). However, molecular-based methods have proved more discriminating/reliable for indepth investigations. Methods particularly useful for epidemiological studies are pulsed field gel electrophoresis (PFGE), polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) of the flagella (flaa and flab) genes and http://www.efsa.eu.int 10 of 105

amplified fragment length polymorphism (AFLP) (Newell et al., 2000). Sails et al. (2003) found sequencing the flaa short variable region useful for investigation of outbreaks. Multilocus sequence typing (MLST) is particularly appropriate for C. jejuni because it relies on more stable housekeeping genes, and has proved useful in elucidating the population structure of the organism and showing that genetic exchange plays a major role in its genetic diversity (Sails et al., 2003; Urwin and Maiden, 2003). Campylobacter spp. are known to show much greater genomic plasticity than other bacteria such as the Enterobacteriaceae with evidence for changes in gene order, as well as relatively frequent loss and acquisition of DNA (Wassenaar et al., 2000). Therefore, to minimise misinterpretation of typing data used for epidemiological studies, it is often advisable to use more than one method of typing. 2.2. Biodiversity of Campylobacter in food animals Poultry: Several studies suggest a low biodiversity (one or two types) among the Campylobacter isolates from the same flock (Van der Giessen et al., 1992; Pearson et al., 1993, 1996; Ayling et al., 1996; Berndtson, et al., 1996b; Chuma et al., 1997; Shreeve et al., 1999), although Stern et al. (1997) and Thomas et al. (1997) found greater biodiversity. In particular rearing conditions with multiple contamination sources, such as free range flocks, much greater biodiversty can be observed (e.g., 84 genotypes among 1225 isolates from seven free-range broiler farms (Rivoal, 2000). Single birds can be colonized by more than one PFGE type of Campylobacter (Hook et al., 2005). PFGE studies on Campylobacter in broiler flocks in Sweden showed an addition of clones appearing during the rearing period (Rivoal, 2005; Hook et al., 2005). Possible explanations include the exposure to different Campylobacter strains on different occasions and that the bacterial genome is plastic. Genotyping such as PFGE and MLST could complement and contribute to the interpretation of results from epidemiological studies and determine whether particular Campylobacter types persist along the food chain or establish themselves as house flora in slaughterhouses or processing plants. Red meat animals and other foods: There have been few detailed studies of strains of Campylobacter present in individual animals or groups of animals, but in pigs Campylobacter infection is often acquired from the sow, and pigs can be infected with diverse strains (Weijtens et al., 1997, 1999; Hume et al., 2002; Moore et al., 2002a). Evidence from molecular typing for human infection from various sources: Serotyping and Fla- and PFGE-typing indicate that poultry is unlikely to be the only source of human infections (Fricker and Park 1989; Koenrad et al., 1995; Nielsen et al., 1997; On et al., 1998; Hudson et al., http://www.efsa.eu.int 11 of 105

1999; Nielsen and Nielsen, 1999). Comparison of isolates from human cases with those from cattle, sheep, pigs, pets, wild birds and poultry using MLST has confirmed earlier conclusions, and shown that strains of C. jejuni from all these sources have the potential to cause disease in humans (Dingle et al., 2002; Manning et al., 2003). 2.3. Methods for isolation and identification of thermophilic Campylobacter 2.3.1. Characteristics important for detection and isolation Campylobacter spp. have a reputation for being difficult to grow and for needing complex media. In fact, they can grow in or on quite ordinary media, such as nutrient broth or agar, provided the atmosphere and humidity are appropriate with a microaerobic atmosphere containing 5-7% oxygen and about 10% carbon dioxide. A low level of hydrogen sometimes stimulates growth of thermophilic Campylobacter spp. (Skirrow et al., 1991). They are sensitive to toxic products of oxygen, and so most (particularly solid) media are supplemented with substances such as blood, whole or lysed, 'FBP' (a mixture of ferrous sulphate (F), sodium metabisulphite (B) and sodium pyruvate (P), charcoal or haematin plus BP. They grow better on solid media if the surface is not too dry - although on wet plates they tend to swarm. Consequently, the appearance of colonies can vary considerably, and it is advisable to check colonies growing on selective media for positive oxidase reaction as well as characteristic morphology by Gram stain or phase contrast microscopy. 2.3.2. Selective media Many different selective plating media have been devised, but thorough comparisons of methods for detecting campylobacters in food have not been made Corry et al., 2003a). However, mccda (modified charcoal cefoperazone deoxycholate agar) has performed well when compared with other media, is specified in the draft ISO method, and is widely used. Direct plating can be used when the substrate examined is likely to contain high numbers of Campylobacter spp., e.g., faeces, intestinal contents or raw poultry or raw red meat offal. Detection of Campylobacter spp. in foods usually requires the use of enrichment media. Enrichment media most commonly used are Preston, Exeter and Bolton broths (Corry et al., 2003a). Standard methods. The ISO (1995) method for examination of foods recommended enrichment in Preston broth (18 h at 42 C) if the Campylobacter spp. were not likely to be stressed, or Park and Sanders broth (32 C for 4 h, then add antimicrobials, then 37 C 2 h and 42 C for 40-42 h) if stressed (ISO, 1995). Karmali plus one other plating medium was recommended (Skirrow, Preston or mccda), incubating for up to 5 days at 42 C (Karmali et al., 1986; Skirrow, 1997). The current draft ISO method for detection of Campylobacter in foods (ISO, 2004) recommends using Bolton broth (1:10 ratio of food to broth), incubating in microaerobic http://www.efsa.eu.int 12 of 105