1. 5/21/2015

2. The reduction of the level of human illness from food-borne pathogens is a public health goal in the European Community and many countries worldwide. As epidemiological studies show that poultry meat and eggs are important sources for consumers' exposure to zoonotic pathogens such as Salmonella and Campylobacter the reduction of the prevalence of contaminated poultry meat or eggs is amajor area of focus. In addition, risk management strategies aim at communicating the risks associated with the preparation and consumption of poultry meat and eggs to consumers.

3. The aim of this risk communication is to raise awareness and to change consumers' behaviour to avoid risky procedures while preparing these foods. Examples for such messages are ‘cook poultry meat and eggs thoroughly’ or ‘wash your hands after handling raw meat.’ Generally, two pathways for human exposure to pathogens in poultry meat and eggs exist, undercooking and cross contamination (FAO/WHO, 2002, 2003). 5/21/2015

4. During frying or other cooking procedures for meat, fairly high temperatures occur on the outside of the meat that most likely will kill the bacteria located on the surface. If the poultry meat is undercooked, some of the internal bacterial pathogens may survive and viable bacteria may be ingested while consuming the (under)cooked meat. Whereas undercooking is mainly a risk associated with bacterial pathogens located in eggs and inside of chicken or turkey meat, cross-contamination events transferring bacteria from poultry meat or egg shells directly or indirectly to other foods in the majority of cases will involve bacteria located on the surface of poultry meat and eggs.

5. It can thus be expected that the efficacy of risk communication messages will depend on the distribution of pathogens in or on the poultry food commodities. For example, labelling a package of chicken meat to inform the consumer of the need for thorough cooking prior to consumption may reduce the risk of undercooking but it will have no effect on the cross-contamination risk.

6.  However, it does not raise awareness that bacteria from the surface of the chicken meat can stick to the hands of the cook or could be spread in the kitchen environment and subsequently may contaminate ready-to-eat foods like salads or already cooked foods accompanying the meal. 5/21/2015

7. The effectiveness of consumer communication strategies aiming at a reduction of undercooking poultry meat or boosting the consumption of hard-boiled instead of soft eggs will depend on the level of the cross-contamination risk. In other words, the prevalence and level of external contamination of these products with bacterial pathogens in relation to the prevalence and numbers of bacteria inside of poultry meat and eggs will determine the possible success of the message aiming at reducing the undercooking rate.

8. This study is aimed at elucidating whether cross-contamination events or undercooking are a greater risk for human illness from zoonotic pathogens associated with poultry in order to prioritise what message should be given to the consumer. An analysis of the relative contribution of different exposure pathways finally leading to ingestion of the pathogens and resulting in illness will be performed. this analysis will focus on the two major bacterial pathogens in poultry, Salmonella spp. and Campylobacter jejuni and their occurrence in poultry meat (chicken, duck, turkey) and chicken table eggs.

9. A literature research has been performed in the time period from May to June 2008 using the ISI Web of Knowledge databases by The Thomson Corporation. Mining for publications with data considering the location of Campylobacter spp. or Salmonella spp. on the surface of or inside of poultry meat and eggs was performed using the key words “Salmonella” and “Campylobacter”, each in combination with the terms “poultry”, “chicken”, “turkey”, and “egg”. Moreover, these search terms were combined with “quantitative”, “cross- contamination”, and “undercooking”.

10. 3.1. Location of the bacterial pathogens 3.1.1. Poultry meat During the literature research 16 studies have been identified which either address the location of Salmonella and Campylobacter on the surface of poultry meat products, or that determine the prevalence and or concentration.. An overview of these studies addressing the type of poultry meat and the country of origin of the respective study is given in Table 1. Six of the studies (thereof one with two datasets) simultaneously looked at the prevalence of Campylobacter and Salmonella. Eight studies (again one with two datasets) concentrated on Campylobacter spp. and two studies looked at Salmonella, only. Quantitative data were available from 11 studies. 5/21/2015

12. The nine datasets giving prevalence for Salmonella spp. on the surface of poultry meat products encompassed 2936 samples. The average prevalence of Salmonella on the surface of poultry meat is 22.6%. Only one study, the percentage of duck breasts that were internally contaminated with Salmonella. 3.8% of 53 samples were internally contaminated. Thus, the prevalence of Salmonella on the surface of poultry meat is about 6 times higher than the prevalence of this pathogen inside of the meat. The average prevalence for Campylobacter on the surface of poultry meat was 62.3% (3235 samples, 14 datasets).

13. A total of 613 samples from five studies were analysed for internal contamination with the pathogen. Average Campylobacter prevalence inside of poultry meat was 10.3%. In five studies meat samples were simultaneously analysed for Campylobacter on the surface and inside of the meat. On average, the prevalence of external contamination of poultry meat with Campylobacter spp. is 6 times higher than a contamination deeper in the meat. Generally, the Salmonella prevalence on the poultry meat surface is lower than the observed Campylobacter prevalence. Campylobacter prevalence was 1.5 to 17 times higher on samples which have been analysed for both pathogens.

14. 3.1.2. Shell eggs Literature data base research identified nine studies that looked at the occurrence of salmonella or Campylobacter on the surface and inside of table eggs. An overview of these studies including information on the country of origin is given in Table 2. Three of the studies simultaneously analysed eggs on the surface and inside for Salmonella and Campylobacter spp., five studies address only Salmonella spp. and one study only presents data for Campylobacter.

15. 5/21/2015

16. Eight studies with a total of 14,343 pooled samples qualitatively analysed the egg shell surface and the content of eggs for Salmonella spp. Surface prevalence ranged from 0.04% to 9.0% positives. Only half of the eight studies showed Salmonella contamination of the egg contents as well. the contamination rate inside of the eggs was significantly lower than on their surface. 5/21/2015

17. In four studies table eggs were analysed for Campylobacter prevalence. Eggs from a hatchery in one study were all negative for Campylobacter. two internally Campylobacter-positive composite egg samples out of a total of 184 samples which were all negative for the pathogen on their surface. Two other studies a prevalence of 0.5% and 0.6% of Campylobacter-positive egg shells. All egg contents in these two studies were negative. The prevalence of Salmonella on the egg surfaces in these studies was almost twice as high as the prevalence of Campylobacter.

19. All of the published risk assessments for Campylobacter spp. In chicken meat are addressing cross-contamination during preparation of the chicken meat as exposure pathway. Four of the risk assessments look as well at exposure via undercooked chicken meat. The first three of these quantitative risk assessment models consider cross contamination as the dominant route of exposure, as compared to undercooking for Campylobacter spp.

20. Their model simulations revealed that 74% of campylobacteriosis cases were caused by cross contamination events involving Campylobacter spp. from the surface of chicken meat during the preparation of meals in German private homes. In comparison, only 3% of cases could be attributed to consumption of undercooked products and in 23% of cases more than one exposure pathway (e.g. inadequate hand washing, not cleaning the kitchen environment, or undercooking) was involved in increasing the probability of illness.

21. A quantitative risk assessment of Campylobacter in poultry meat revealed that an elimination of the poultry meat products that are highly contaminated will have a great effect in lowering the campylobacteriosis risk. Although this risk assessment model has a module with consumer handling practices that considers undercooking and cross contamination as exposure pathways, a comparison of the contribution of these two pathways to consumers' exposure to Campylobacter bacteria has not been performed

22. For Salmonella, the relative relevance of cross-contamination events and undercooking to the exposure of consumers to this pathogen from poultry meats has not been addressed by risk assessors in the same magnitude as in the case of Campylobacter spp. and this food commodity. It has shown that in the EU, among the food-borne cases of human salmonellosis, eggs and egg products are still the most frequently implicated sources. Meat is also an important source of food-borne salmonellosis, with poultry and pork implicated more often than beef and lamb. More specific conclusions about raw meat brought into kitchens were not possible.

23. However, to assess consumers' exposure by consumption of Salmonella, only an uptake during consumption of undercooked chicken meat is addressed. The model does not consider cross-contamination events while preparing a contaminated carcass in the kitchen. Key findings of the risk assessment were that a reduction of numbers of Salmonella on the surface of chicken carcasses and even a small reduction in the frequency of undercooking.

24. Even though this exposure pathway was not included, the authors of this risk assessment feel that cross-contamination may in fact be the predominant source of risk of illness. Data on the magnitude of cross-contamination in the domestic kitchen, and the pathways of such cross-contamination are identified as areas where future data collection and research efforts need to focus. It is outlined that the lack of addressing consumers' exposure by cross-contamination in the assessment is a limitation and caveat.

25. These authors look at different egg products considering how consumers handle them during preparation and make a salmonellosis risk ranking for the different egg products. The greatest risk was associated with foods containing raw eggs. By Salmonella coming from the surface of eggs was not addressed. A risk assessment addressing Campylobacter on or in eggs was not available in published literature.

26. A more general model for all bacteria, which addresses individual hygiene practices during food preparation and consumption patterns in private homes, has been developed by Christensen et al., 2005. The model was used to estimate the probability of ingesting a meal where precautions have not been taken to avoid the transfer of microorganisms from raw food to final meal, exemplified by the event that the cutting board was not washed during food preparation.

27. The probability of ingesting a risk meal was highest for young males aged 18–29 years and lowest for the elderly above 60 years of age. The simulated results show that the probability of ingesting a chicken risk meal at home does not only depend on the hygiene practices of the persons preparing the food, but also on the consumption patterns of consumers, and the relationship between people preparing and ingesting food.

28. Plenty of publications on the subject of communicating about food safety risks to consumers exist, but only those specifically addressing consumer handling during preparation of poultry or eggs have been analyzed. The studies are very diverse. An analysis looking at the focus of the respective studies, their observations, and the recommendations given for information is presented in Table 3.

29. 5/21/2015

30. Within the European Union, labelling requirements aiming at reducing the risk of consumers' exposure to Salmonella in poultry meat are laid down in Article 6 of Regulation 2073/2005 (European Commission, 2005). Manufacturers have to label raw poultry meat intended to be eaten cooked to inform the consumer of the need for thorough cooking prior to consumption. 5/21/2015

31. In the case of eggs, provisions for the marking of packs are laid down in Article 12 of Regulation 589/2008 (European Commission, 2008). As a measure to reduce the risk of multiplication of Salmonella bacteria during storage of table eggs in private homes, an indication advising consumers to keep eggs chilled after purchase shall be on egg packs. Both labeling requirements can be interpreted as measures which may reduce European consumers' exposure to bacterial contaminations in meat or inside of eggs. Cross-contamination risks are not mentioned.

32. Fischer et al. (2007) conclude that cross-contamination and inadequate cooking are the most important factors that determine food safety during the preparation of poultry meat. The question remains whether conclusions can be drawn on the magnitude of exposure coming from cross-contamination or undercooking events of poultry meat and eggs. A first approach to finding an answer to this question was a literature research for studies.

33. In the case of poultry meat more data sets specifically looking at the prevalence on the surface and separately inside of the meat were available for Campylobacter spp., whereas in case of eggs more studies on Salmonella spp. were found. For poultry meat and Campylobacter all studies looking at the surface and inside showed higher prevalence and numbers of this pathogen on the meat surface. For Salmonella, only one study could be found that looked at the prevalence of Salmonella inside of poultry meat. However, it seems that as well for Salmonella bacteria the prevalence of surface contamination is several times higher.

34. Generally, the prevalence and number of Salmonella pathogens is lower than the contamination of poultry meat with Campylobacter. In conclusion, the analysis of studies presenting prevalence data and level of external contamination of poultry meat with Salmonella spp. And Campylobacter spp. in relation to the prevalence and numbers of the pathogens inside of these food products revealed a six times higher contamination on the foods' surface. Assuming that bacteria located on the meat surface will be killed during cooking procedures, these data clearly demonstrate the superior importance of cross contamination events over undercooking of the meat for consumers’ exposure to these two bacterial pathogens.

35. The role of undercooking may be more relevant when whole poultry carcasses (like turkeys prepared for Thanksgiving) are considered. In case of minced poultry meat a contamination will not be restricted to the surface and Salmonella or Campylobacter bacteria could also be present in the center of a portion. More over, bacteria on the outside can survive for a relatively long period of time during cooking under high temperatures.

36. A study by Atanassova et al. (2007)on the prevalence of Campylobacter in turkey breast fillets and cuts demonstrated that in Germany 58.3% of retail samples were positive for this pathogen and recently published results from a European monitoring study show an average prevalence of 30.7% of Salmonella spp. In fattening turkeys (Anonymous, 2008b).

37. A Canadian study compared the prevalence of Salmonella and Campylobacter in broiler chicken and turkey flocks at the time of slaughter. They found higher prevalence rates for both pathogens in turkey flocks, giving evidence for the possible importance of turkey meat for consumers' exposure with pathogens. Research on the number of pathogens on the surface and inside of turkey meat should be encouraged to gather data for risk assessment and management.

38. Epidemiological data tell us that egg consumption is the largest contributor to salmonellosis cases in the population. Studies showed higher prevalence levels for surface contamination (Table 2). Only one study from found Campylobacter inside of eggs and a higher contamination rate of egg contents with Salmonella than on the surface of the respective eggs. prevalence of Salmonella was almost twice as high as the prevalence of Campylobacter.

39. Bacterial transfer could be possibly prevented by identification and elimination of certain tasks and activities involving higher levels of cross-contamination. The preliminary analysis of current risk communication strategies showed that they mainly focus on getting consumers to avoid undercooking of poultry meat and consumption of dishes with raw eggs, and less on information that focuses on avoiding cross-contamination.

40. a survey in The Netherlands where subjects had to imagine situations involving chicken contamination and report how they would react behaviourally if this situation occurred. It was found that 60% of the subjects would allegedly avoid the risks by not consuming chicken for a while. Moreover, warnings that a food can be a health hazard based on consumer behavior could also been seen as a way of removing the legal responsibility from the producer to the consumer.

41. the majority of study participants wanted to see such kind of information on the food packaging. Thus, labeling of poultry meat and eggs with warnings that this food might be a source of cross contamination in kitchens can be considered as a promising risk management option. However, more research is needed to analyse the possible public health effect of this risk communication strategy.

42. In conclusion, in order to reduce consumers' exposure to pathogens such as Salmonella and Campylobacter in poultry meat and eggs during preparation of these foods, management activities should focus on cross-contamination risks. More risk assessments addressing exposure through this exposure pathway and research on the efficacy of risk communication activities aiming at the reduction of cross- contamination events would be helpful to support future risk management decisions.

43. Reference Petra Luber,2009. Cross-contamination versus undercooking of poultry meat or eggs — which risks need to be managed first?. International Journal of Food Microbiology 134, 21–28.