Summary of the Swedish Campylobacter Program in Broilers, 2001 through 2005

2008 Journal of Food Protection, Vol. 70, No. 9, 2007, Pages 2008 2014 Copyright, International Association for Food Protection Summary of the Swedish Campylobacter Program in Broilers, 2001 through 2005 I. HANSSON, 1 * L. PLYM FORSHELL, 2 P. GUSTAFSSON, 3,4 S. BOQVIST, 5,6 J. LINDBLAD, 4 E. OLSSON ENGVALL, 5 Y. ANDERSSON, 6 AND I. VÅGSHOLM 5 1 Department of Bacteriology and 5 Swedish Zoonosis Center, National Veterinary Institute, SE-751 89 Uppsala, Sweden; 2 National Food Administration, P.O. Box 622, SE-751 26 Uppsala, Sweden; 3 Swedish Board of Agriculture, Vallgatan 5, SE-551 82 Jönköping, Sweden; 4 Swedish Poultry Meat Association, Box 55633, SE-102 14 Stockholm, Sweden; and 6 Swedish Institute for Infectious Disease Control, SE-171 82 Solna, Sweden MS 07-083: Received 14 February 2007/Accepted 27 March 2007 ABSTRACT A Campylobacter monitoring program in broiler chickens was carried out in Sweden from 2001 through 2005. The objective was to reduce the occurrence of Campylobacter in the food chain through preventive measures, starting with primary production. The program involved collecting samples from all broiler flocks at slaughter and occasional additional times. The annual incidence of Campylobacter-positive slaughter batches progressively decreased from 20% in 2002 to 13% in 2005. Most of the positive batches had a high within-flock prevalence of Campylobacter. However, about 18% of the positive batches had a low-within-flock prevalence; Campylobacter spp. were isolated from at most 50% of the cloacal samples. The incidence of batches contaminated at slaughter ranged between 6 and 9% during the study period. During all 5 years, a seasonal peak of incidence was observed in the summertime. In an additional study, quantitative analyses were performed on neck skin samples and carcass rinse samples. Those results were compared with the positive and negative findings of the cloacal, cecum, and neck skin samples at slaughter. When Campylobacter was found in the cecum, there was a higher level of Campylobacter in the quantitative analyses. Those batches where Campylobacter already had been found on the farm had a higher concentration of Campylobacter than those batches in which Campylobacter was found only at slaughter. During the study period, about one-third of producers seldom delivered Campylobacter-positive batches ( 10% positive batches per year). Thus, it is possible to produce Campylobacter-free broilers in Sweden. Campylobacter spp. are the most commonly reported bacteria in cases of human gastroenteritis in industrialized countries (10). Poultry products are generally believed to be a major reservoir of this organism. However, there is limited information on the prevalence of Campylobacterpositive poultry flocks in many countries. Studies in Europe have indicated that the prevalence varies from 3 to 83%, with lower figures in northern European countries (18) and higher figures in southern European countries (28). In a Senegalese study, Campylobacter spp. were isolated from 63% of broiler chicken flocks (8), whereas in one U.S. study 87% of flocks were colonized (31). However, the results from these studies are not necessarily comparable because different sampling strategies and analytical methods were used. Many studies have been performed to identify transmission routes of Campylobacter spp. into broiler flocks; vertical transmission is not considered to be a common route (7). Significant risk factors identified include older age of broilers (5), the presence of other livestock within 1 km (5, 8), wall-mounted ventilation systems (29), a higher number of workers in the poultry house (29), season (summer) (5, 21, 29), and batch depletion (14). The most important predictors of reduced Campylobacter infection in * Author for correspondence. Tel: 00 46 18 674686; Fax: 00 46 18 674093; E-mail: ingrid.hansson@sva.se. broiler flocks are application and improvement of hygiene barriers (12, 13, 33). In 1991, a voluntary surveillance program for broilers was initiated by the Swedish Poultry Meat Association. This program included collecting samples from all flocks at the slaughterhouse. A revised Campylobacter program was initiated in Sweden on 1 July 2001. Based on the knowledge obtained in the previous program, the objectives of the revised program included reducing the annual incidence of Campylobacter-positive slaughter batches to 0 to 2% and reducing the number of domestic human cases of campylobacteriosis. The results of this program and of some of the additional studies are described here and compared with previously published results of studies associated with the Swedish Campylobacter program in broilers. MATERIALS AND METHODS Study population. In Sweden, a broiler chicken farm usually has one or several broiler houses consisting of one or more compartments. In each compartment, up to seven flocks are produced in a year. A flock is defined as all broilers kept within the same enclosure and constituting a single epidemiological unit. A flock can be divided into one to three slaughter batches. In 2001, 127 producers with 499 compartments produced 70.8 million broilers in Sweden, whereas in 2005, 123 producers with 454 compartments produced 71.5 million broilers. All of the broiler producers and slaughterhouses are situated in the southern part of Sweden.

J. Food Prot., Vol. 70, No. 9 SWEDISH CAMPYLOBACTER PROGRAM IN BROILERS, 2001 THROUGH 2005 2009 TABLE 1. Collection and pooling of broiler chicken cloacal swab samples and cecum samples in the Campylobacter program in Sweden, 2001 through 2005 Type of sample No. of samples 2001 2002 2003 2004 2005 Cloacal swab samples per slaughter batch 40 40 40 40 20 Cloacal swab samples per pool 10 10 10 or 20 a 20 20 Pooled cloacal samples per slaughter batch 4 4 4 or 2 a 2 1 Cecum samples per slaughter batch 0 0 0 0 10 Pooled cecum samples per slaughter batch 1 a The pooling method was changed on 1 July 2003; thereafter, twice as many cloacal samples were placed in one pool at the laboratory. In 2001, about 78% of chicken consumed in Sweden originated from Swedish producers, but by 2005 this figure had decreased to 60%, and consumption of imported chicken increased from 26,200 to 44,299 tons (69%) within the same period. When Campylobacter spp. are found in a flock, the farm of origin is advised to implement more stringent biosecurity measures to prevent subsequent flocks from being infected with Campylobacter. There are no statutory sanctions or penalties, but when the program started in 2001, eight of nine slaughterhouses were paying a premium for Campylobacter-free flocks. Monitoring and collection of samples. All broiler flocks were sampled at slaughter. From each batch of broilers, 40 cloacal swabs were taken after stunning and bleeding but before scalding. These 40 cotton swab samples were divided into four pools, with 10 samples in each pool. For cost-efficiency reasons, this approach was changed to 20 samples per pool in 2003 (Table 1). In addition, 10 neck skin samples of about 2 cm 2 were taken from each batch of broilers before chilling, and these 10 samples were pooled to 1 sample. These samples were placed in screw-cap plastic tubes with about 10 ml of Cary-Blair transport medium (9) with 0.16% agar (22). From April to December 2005, a special study was performed in which intact ceca were collected during slaughter from 10 individual broilers in each batch. These 10 samples were pooled into 1 sample in a sterile plastic jar. The samples were sent to the laboratory daily without cooling except for samples taken on Fridays, which were refrigerated at 4 to 8 C and sent to the laboratory on the following Sunday or Monday. Sampling on the farm was voluntary. Cloacal swabs were taken from live broiler chickens; 10 swabs were pooled to form one sample. The sampling procedures were approved by the Swedish Ethical Committee for Scientific Experiments with Laboratory Animals. In an additional study performed in 2005, sock samples were collected on the farm just before the broilers were loaded for transport to slaughter. The farmers took samples from inside the broiler houses by walking in the litter bed at least four times on the longest distance from wall to wall. This sampling was carried out in conjunction with ordinary work in the broiler house. One sock sample consisted of one pair (two socks) of tubular retention bandage. Before sampling, the bandage was moistened with 20 ml of Cary-Blair transport medium (Difco, Becton Dickinson, Sparks, Md.; and Merck, Whitehouse Station, N.J.) and pulled over ordinary shoes. After sampling, the socks were placed in a plastic bag with 20 ml of Cary-Blair transport medium. In the same study, an additional three pooled samples of fecal droppings were obtained at three different locations in the broiler houses or compartments, with each pool consisting of 10 cotton swab samples. The samples were placed in screw-cap plastic tubes with about 10 ml of Cary-Blair transport medium. From May through October 2005, a quantitative study was conducted on neck skin and whole carcass rinse samples collected at eight slaughterhouses. The objective was to study the correlation between Campylobacter findings in samples taken on farms and at slaughter (cloacal and neck samples) and the quantitative load of Campylobacter spp. on the carcasses. The pooled 10 neck skin samples taken before chilling were used for both qualitative and quantitative studies. Two carcasses from each batch, one before and one directly after chilling, were used to obtain carcass rinse samples. Each carcass was rinsed in plastic bags with 400 ml of buffered peptone water (Oxoid, Basingstoke, UK), and the rinse fluid was placed in a sterile plastic jar. Each sample consisted of a randomly selected carcass, and rinse samples were collected 4 days a week (not Fridays because of logistics problems). The samples were refrigerated at 4 to 8 C and sent to the laboratory daily, and bacterial analysis started on the day they arrived at the Department of Bacteriology at the Swedish National Veterinary Institute (Uppsala). Qualitative analysis. The sock, cloacal, and neck skin samples were cultured according to the method of the Nordic Committee on Food Analysis (NMKL no. 119) in Preston Campylobacter selective enrichment broth (PEB; Oxoid; Difco, Becton Dickinson; and Merck) and Preston Campylobacter selective agar (PA; Oxoid; Difco, Becton Dickinson; and Merck); each medium was supplemented with 5% (vol/vol) freeze-lysed horse blood. The PEB also was supplemented with Campylobacter selective supplement and Campylobacter supplement, and the PA was supplemented with Campylobacter selective supplement. For culturing of cloacal samples, two sterile cotton swabs were placed in a tube with Cary-Blair transport medium (Difco and Merck) and then placed in 5 ml of PEB. For culturing of neck skin and sock samples, about 10 g of sample was transferred into PEB at a ratio of 1 to 9. The samples in PEB were incubated for 24 2 h, and then 20 l of the culture was streaked onto a PA plate, which was incubated for 48 2 h. A pool of cecum material was prepared by aseptically collecting contents from each of the 10 broiler ceca. The cecum samples were analyzed by direct plating on modified charcoal cefazolin sodium deoxycholate agar (mccda; Oxoid). All cultures were incubated at 41.5 1 C under a microaerophilic atmosphere (Campygen, Oxoid). From each agar plate that produced typical Campylobacter colonies, one colony was subcultured and tested for motility, production of oxidase and catalase, and hippurate hydrolysis (23). All Campylobacter isolates were kept in 1 to 2 ml of serum broth with 15% glycerol and frozen at 70 C. A slaughter batch was deemed positive for Campylobacter when one or more of the pools of cloacal samples tested positive for Campylobacter. Batches in which Campylobacter was found in half or fewer of the sample pools were regarded as having low within-flock prevalence. Slaughter batches with negative cloacal samples but positive neck skin samples were regarded as negative slaughter batches that had become contaminated during slaughter. Slaughter batches with all four pools positive were designated as having high within-flock prevalence. Quantitative analyses. The quantification of Campylobacter in neck skin and whole carcass rinse samples was carried out according to a modified ISO method (ISO/TS 10272-2). The neck

2010 HANSSON ET AL. J. Food Prot., Vol. 70, No. 9 TABLE 2. Prevalence of Campylobacter spp. contamination in batches of broiler chickens in the Swedish Campylobacter program, 2001 through 2005 a Year No. of batches No. (%) of positive batches based on samples from: Cloaca Neck skin No. (%) of positive carcasses b 2001 (6 mo) 2,140 489 (23) 579 (27) 151 (9) 2002 3,842 760 (20) 917 (24) 222 (7) 2003 3,224 566 (18) 659 (21) 157 (6) 2004 3,019 429 (14) 565 (19) 186 (7) 2005 2,975 394 (13) 529 (18) 169 (6) a Samples were collected from broilers at slaughterhouses that were members of the Swedish Poultry Meat Association. b Slaughter batches with Campylobacter-positive neck skin samples and Campylobacter-negative cloacal samples divided by the total number of neck skin samples minus the number of slaughter batches with both positive cloacal and neck skin results. skin samples were analyzed by taking 1 ml of the 10 ml of Cary- Blair medium from the plastic jar with the 10 skin samples, and the carcass rinse samples were analyzed by taking 1 ml of the 400 ml of rinse fluid from the plastic jar. Both carcass and neck skin samples were analyzed by direct plating 1 ml of serially diluted rinse fluid on one mccda plate for each dilution. The mccda plates (14 cm diameter) were preincubated at 41.5 1 C for 2 to 6 h before use. Plated samples were incubated microaerobically for 48 2 h at 41.5 1 C. Statistical analysis. The data from the study were compiled and analyzed using Microsoft Access (Microsoft, Redmond, Wash.) and StatView (SAS Institute, Cary, N.C.) software. Differences in prevalence of Campylobacter-positive slaughter batches between different sampling sites were examined with the chisquare test. The influence of the flock status on the farm and at slaughter and the load of Campylobacter in the neck skin and carcasses were tested with an analysis of variance. Differences between batches were deemed significant at P 0.05. RESULTS Annual program. During the Campylobacter program of 2001 through 2005, 44,951 cloacal samples, 2,090 cecal samples, and 15,061 neck skin samples collected at slaughter and 3,758 sock samples, 2,972 cloacal samples, and 1,328 fecal samples collected on the farm were analyzed, for a total of 70,160 samples analyzed. The proportion of slaughter batches that tested positive for Campylobacter based on cloacal samples decreased from 20 to 13% (P 0.0001) during the study period (Table 2). When the results from the different samples analyzed were compared, a significantly higher prevalence (P 0.001) was found based on neck skin samples than based on cloacal swabs, a finding that was taken to represent carcass contamination during processing. The overall prevalence of carcass contamination ranged from 6 to 9% (P 0.04) (Table 2), but there were large variations between slaughterhouses (range, 0 to 26%). About 40% of all producers seldom delivered Campylobacter-positive batches, but 60% of producers delivered 90% of the Campylobacter-positive slaughter batches. Most (73%) of the Campylobacter-positive batches had a high within-flock prevalence, where Campylobacter was found in all pools of cloacal samples. From 2001 through 2004, the proportion of Campylobacter-positive batches with a low within-flock prevalence was 27%, with a range of 6 to 38% among the different slaughterhouses. The within-flock prevalence during 2005 was not calculated because only one sample was taken from each slaughter batch (Table 1). During all 5 years, a seasonal peak in prevalence was observed in the summer. The highest monthly prevalence of Campylobacter-positive batches (52%) was recorded in August 2003 (Fig. 1). Additional study: on the farm. From April through December 2005, both sock and fecal samples were obtained on farms. In 95% of the flocks with Campylobacter-positive sock samples, Campylobacter spp. also were found in the fecal samples. However, in 1% of the flocks, Campylobacter spp. were isolated only from the sock samples. The results from the sock samples were then compared with results from the cloacal, cecal, and neck skin samples taken at slaughter. In 1,490 batches with all four types of samples, Campylobacter was found in 181 (12%) of the sock samples taken 8 h before slaughter, in 202 (14%) of the cecum samples, in 234 (16%) of the cloacal samples, and in 307 (21%) of the neck skin samples. The differences between the samples were all significant (P 0.001). Additional study: quantification of neck skin and carcass rinse samples. In 2005, 3,209 quantitative analyses were performed on neck skin and carcass rinse samples. Campylobacter was detected in 225 (15%) of the neck skin samples and in 139 (19%) and 152 (15%) of the carcass rinse samples collected before and after chilling, respectively (Table 3). There was a certain discrepancy between results obtained from batches analyzed before and after chilling. FIGURE 1. Seasonal variations in Campylobacter spp. isolated from broiler chicken cloacal samples taken at slaughter in the Swedish Campylobacter program, 2001 through 2005.

J. Food Prot., Vol. 70, No. 9 SWEDISH CAMPYLOBACTER PROGRAM IN BROILERS, 2001 THROUGH 2005 2011 TABLE 3. Results of quantitative analyses of broiler chicken neck skin samples and whole-carcass rinse samples obtained before and after chilling as part of the Campylobacter program in Sweden, 2005 Sample type No. of samples analyzed No. (%) of positive samples Campylobacter count (log CFU) Maximum Samples Mean Neck skin 1,467 225 (15) 4.6 2.4 0.06 Carcass rinse Before chilling 750 139 (19) 7.6 4.8 0.10 After chilling 992 152 (15) 6.5 4.2 0.10 Twenty-nine percent (40 of 139) of the batches tested positive before chilling but negative after. In contrast, 7% tested positive after chilling but not before. The mean level of Campylobacter in neck skin samples was 2.4 log CFU/ml, whereas the results for whole carcass rinse samples were 4.8 and 4.2 log CFU per carcass before and after chilling, respectively (Table 3). No obvious correlation could be found between the results from the neck skin and the whole carcass rinse samples (Fig. 2). Comparison of results from the two additional studies. The results of the two additional studies revealed that those slaughter batches in which Campylobacter was also found on the farm had a significantly higher load of Campylobacter than those slaughter batches in which Campylobacter was found only at slaughter (P 0.001) (Fig. 3). In eight of the batches, Campylobacter was found on the farm but not in cloacal, cecal, or neck skin samples taken at slaughter. Five of these batches also tested negative after quantification of the neck skin, but one of the carcass rinse samples was positive, resulting in a large standard deviation (Fig. 3). When Campylobacter was found in the cecum, the SD Campylobacter load in the neck skin and carcass samples was higher than that found when only the cloacal and/or neck skin samples were positive for Campylobacter (P 0.001) (Fig. 4). DISCUSSION The annual incidence of Campylobacter-positive broiler chicken batches at slaughter was reduced during the period of the Swedish Campylobacter program, possibly because of producers increased knowledge of the importance and use of hygiene barriers. However, the objective of 0 to 2% Campylobacter-positive slaughter batches was not achieved. This failure can be explained by various factors, including the fact that the true Campylobacter prevalence was at least twice as high as that predicted when the program started. Nevertheless, the large number of farms that produced broilers without Campylobacter on a regular basis demonstrated that it is possible to consistently produce Campylobacter-free broilers. Only a small number of farms contributed to most of the Campylobacter load by consistently producing infected batches. In some cases there was an obvious cause, such as contaminated water, infected poultry or other livestock in the vicinity, or noncompliance with recommended hygiene regulations. Some contamination could be blamed on ventilation systems that allowed higher numbers of insects to enter the broiler houses (17). In 2004, a comparison was made between farms that often, occasionally, or seldom produced Campylobacter-positive flocks (17). No difference was found between these farms in the recovery of Campylobacter from samples of the environment outside the house, but a significant difference was found between the farms for the sock samples obtained inside the broiler houses. The presence of hygiene barriers is probably the most important factor linked to a low risk of flock colonization (21, 33). The quantitative studies presented here revealed that broilers that were Campylobacter positive on the farm had FIGURE 2. Quantitative analyses of broiler chickens in the Swedish Campylobacter program, 2005. Results for neck skin samples were compared with those for whole carcass rinse samples obtained before (n 730) and after (n 724) chilling. Ten neck skin samples pooled into one sample and one carcass rinse sample taken before and one taken after chilling were collected from each slaughter batch.

2012 HANSSON ET AL. J. Food Prot., Vol. 70, No. 9 FIGURE 3. Quantitative analyses of broiler chickens in the Swedish Campylobacter program, 2005. Results for neck skin samples and two carcass rinse samples (one taken before and one taken after chilling) from each slaughter batch were compared with those of samples taken on the farm and at slaughter. a higher concentration of Campylobacter per carcass than did broilers contaminated during transport and processing. This higher concentration probably represents a higher risk to the consumer. Thus, the focus of Campylobacter control should be on preventing Campylobacter on the farm and eliminating Campylobacter from the slaughter process. The results from the Swedish Campylobacter program revealed low within-flock Campylobacter prevalence. Possible reasons for this low prevalence are that a late introduction of Campylobacter leads to a slower spread in a flock and that some as yet unidentified Campylobacter subtypes spread more slowly than other subtypes. Some batches of broilers produced on low-prevalence farms became contaminated during transport to the slaughterhouse, presumably from inadequately cleaned crates. In previous studies (4, 6, 19, 20, 25), broiler flocks became colonized by Campylobacter at about 3 to 4 weeks of age and reached a within-flock prevalence of close to 100% in days, with chickens remaining colonized until slaughter. The higher prevalence of Campylobacter obtained based on cloacal samples taken at slaughter compared with that based on cloacal samples taken on the farm supports the hypothesis of infection during transport. The difference was most pronounced during the late summer and early fall (high-prevalence season). Two studies have been carried out in Sweden to investigate the occurrence of Campylobacter-contaminated crates. Both studies reveled that overall more than 50% of transport crates tested were contaminated with Campylobacter despite cleaning (15, 17). Contaminated crates may also contribute to the introduction of Campylobacter and other pathogens to the farm. In Sweden, producers are generally paid a premium when they deliver Campylobacter-free flocks to slaughter. Until 2004, this premium was based on the results from the cloacal samples at slaughter. As a consequence of the transport crate studies (15, 17), however, regulators decided in 2005 that the premium should be based on the sock samples taken on the farm. A seasonal variation in Campylobacter prevalence was observed, with peaks in the summer. This seasonal variation is similar to that observed for campylobacteriosis cases in humans in Sweden and in broilers and humans in other countries (2, 5, 18). A fluctuation in Campylobacter prevalence between years also has been identified in Sweden (1). This fluctuation may be due to between-year variation in the climate. In a Danish study (26), a corresponding increase in the incidence of human campylobacteriosis and the percentage of infected broiler flocks at slaughter was found during periods of higher temperatures. Newell and Fearnley (24) suggested that the seasonal variation in campylobacteriosis in humans coincides with or may precede rather than follow that in poultry, as also reported by Berndtson (3). During the high-prevalence season of 2005, a quantitative study on neck skin samples and whole carcass rinse samples obtained before and after chilling revealed no correlation between the results of the quantitative analyses on these two types of samples (Fig. 2). Although it is possible that no correlation exists, the differences in specimen type should be considered, i.e., 1 ml of the 10 ml of Cary-Blair medium from the plastic jar with 10 neck skin samples was analyzed and compared with 1 ml from the 400 ml of rinse sample from one carcass. Because for each slaughter batch, one sample was collected from only one broiler before chilling and was compared with the rinse sample obtained from only one other broiler from the same slaughter batch FIGURE 4. Quantitative analyses of broiler chickens in the Swedish Campylobacter program, 2005. Results for neck skin samples and two carcass rinse samples (one taken before and one taken after chilling) from each slaughter batch were compared when Campylobacter was found in the cecum samples or cloacal and/or neck skin samples at slaughter.

J. Food Prot., Vol. 70, No. 9 SWEDISH CAMPYLOBACTER PROGRAM IN BROILERS, 2001 THROUGH 2005 2013 after chilling, these results may not be representative of whole slaughter batches consisting of 2,000 to 50,000 broilers. Carcasses from flocks negative for Campylobacter on the farm but contaminated during transport and processing were usually those with low levels in the quantitative analyses. When Campylobacter was found in the cecum, it was more likely that a higher concentration of Campylobacter would be found in the quantitative analyses of the neck skin and carcass rinse before and after chilling than when Campylobacter was detected only in the cloacal and/or neck skin samples (Fig. 4). Similar results were found in a Danish study (30), with a correlation between Campylobacter concentrations in the intestinal contents and on carcasses of chickens after defeathering. This finding also supports the suggestion that monitoring of cecum samples for Campylobacter will provide information most useful for controlling the public health problem of Campylobacter in chicken. The numbers of reported domestic cases of campylobacteriosis in humans in Sweden were 2,839 in 2001 and 2,255 in 2005, a decrease in incidence from 31.9 to 24.9 per 100,000 persons (32). However, the long-term objective of the Campylobacter program to reduce human domestic campylobacteriosis cases to less than 1,000 per year has not been achieved. In many countries that export poultry meat to Sweden, the incidence of Campylobacter in broilers is higher than that in Sweden (11). The doubled importation of poultry meat products during the study period could have concealed the effect of the decrease in Campylobacter prevalence in Swedish broilers. The revised Campylobacter program elicited new and useful information about the occurrence of Campylobacter in Swedish broilers. The incidence of Campylobacter-positive slaughter batches was progressively reduced from 20 to 13%. Slaughter batches with low within-batch prevalence were identified, split slaughter was confirmed as a risk (16), and contamination of carcasses was found to occur both during transport and during the slaughtering process (15, 16). Cecum sample results seem to be the most appropriate predictor of the incidence of Campylobacter-positive broiler flocks. Environmental Campylobacter loads on the farm were mostly constant for both high- and low-incidence farms (17), which indicates that house hygiene barriers appeared to be an important factor linked to low prevalence of flock contamination. Flocks that tested positive for Campylobacter on the farm had a higher concentration of Campylobacter on carcasses at slaughter than did those flocks in which Campylobacter was found only at slaughter. Around 40% of producers seldom delivered Campylobacter-positive flocks, which shows that it is possible to produce Campylobacter-free broilers in Sweden. ACKNOWLEDGMENTS The authors gratefully acknowledge the broiler producers and the staff at the Department of Bacteriology of the Swedish National Veterinary Institute for assistance with the microbiological analyses and the staff at the slaughterhouses participating in the study for making this work possible. The authors also thank the European Commission, the Ivar and Elsa Sandberg Research Foundation, and the Swedish Farmers Foundation for Agricultural Research for financial support. REFERENCES 1. Anonymous. 2004. Zoonoses in Sweden 2003. Trends and sources of zoonotic infections recorded in Sweden during 2003. National Veterinary Institute, National Food Administration, Swedish Board of Agriculture, and the Swedish Institute for Infectious Disease Control, Solna. 2. Bang, D. D., E. M. Nielsen, K. Knudsen, and M. Madsen. 2003. A one-year study of Campylobacter carriage by individual Danish broiler chickens as the basis for selection of Campylobacter spp. strains for a chicken infection model. Epidemiol. Infect. 130:323 333. 3. Berndtson, E. 1996. Campylobacter in broiler chickens. The mode of spread in chicken flocks with special reference to food hygiene. Ph.D. thesis. Swedish University of Agricultural Sciences, Uppsala. 4. Berndtson, E., M.-L. Danielsson-Tham, and A. Engvall. 1996. Campylobacter incidence on a chicken farm and the spread of Campylobacter during the slaughter process. Int. J. Food Microbiol. 32:35 47. 5. Bouwknegt, M., A. W. Van De Giessen, W. D. C. Dam-Deisz, A. H. Havelaar, N. J. D. Nagelkerke, and A. M. Henken. 2004. Risk factors for the presence of Campylobacter spp in Dutch broiler flocks. Prev. Vet. Med. 62:35 49. 6. Bull, S. A., V. M. Allen, G. Domingue, F. Jorgensen, J. A. Frost, R. Ure, R. Whyte, D. Tinker, J. E. Corry, J. Gillard-King, and T. J. Humphrey. 2006. Sources of Campylobacter spp. colonizing housed broiler flocks during rearing. Appl. Environ. Microbiol. 72:645 652. 7. Callicott, K. A., V. Fredriksdottir, J. Reiersen, R. Lowman, J.-R. Bisaillon, E. Gunnarsson, E. Berndtson, K. L. Hiett, D. S. Needleman, and N. J. Stern. 2006. Lack of evidence for vertical transmission of Campylobacter spp. in chickens. Appl. Environ. Microbiol. 72:5794 5798. 8. Cardinale, E., F. Tall, E. F. Gueye, M. Cisse, and G. Salvat. 2004. Risk factors for Campylobacter spp. infection in Senegalese broilerchicken flocks. Prev. Vet. Med. 64:15 25. 9. Cary, S. G., and E. B. Blair. 1964. New transport medium for shipment of clinical specimens. J. Bacteriol. 88:96 98. 10. European Commission. 2004. Trends and sources of zoonotic agents in animals, feedstuffs, food and man in the European Union and Norway in 2002. Report to the European Commission in accordance with Article 5 of the Directive 927117/EEC, prepared by the Community Reference Laboratory on the Epidemiology of Zoonoses, BgVV, Berlin. 11. European Food Safety Authority. 2006. EFSA s Second Community Summary Report on trends and sources of zoonoses, zoonotic agents, antimicrobial resistance and foodborne outbreaks in the European Union in 2005. Available at: http://www.efsa.europa.eu/en.html. Accessed 2 November 2006. 12. Evans, S. J., and A. R. Sayers. 2000. A longitudinal study of Campylobacter infection of broiler flocks in Great Britain. Prev. Vet. Med. 46:209 223. 13. Gibbens, J. C., S. J. S. Pascoe, S. J. Evans, R. H. Davies, and A. R. Sayers. 2001. A trial of biosecurity as a means to control Campylobacter infection of broiler chickens. Prev. Vet. Med. 48:85 99. 14. Hald, B., E. Rattenborg, and M. Madsen. 2001. Role of batch depletion of broiler houses on the occurrence of Campylobacter spp. in chicken flocks. Lett. Appl. Microbiol. 32:253 256. 15. Hansson, I., M. Ederoth, L. Andersson, I. Vågsholm, and E. O. Engvall. 2005. Transmission of Campylobacter spp. to chickens during transport to slaughter. J. Appl. Microbiol. 99:1149 1157. 16. Hansson, I., E. Olsson Engvall, J. Lindblad, A. Gunnarson, and I. Vågsholm. 2004. The Campylobacter surveillance program for broilers in Sweden, July 2001 June 2002. Vet. Rec. 155:193 196. 17. Hansson, I., I. Vågsholm, L. Andersson, and E. Olsson Engvall. Correlations between Campylobacter spp. prevalence in the environment and broiler flocks [published on-line 7 February 2007]. J. Appl. Microbiol. DOI: 10.1111/j.1365-2672.2007.03291.x. 18. Hofshagen, M., and H. Kruse. 2005. Reduction of Campylobacter spp. in broilers in Norway after implementation of an action plan. J. Food Prot. 68:2220 2223.

2014 HANSSON ET AL. J. Food Prot., Vol. 70, No. 9 19. Jacobs-Reitsma, W. F. 1997. Aspects of epidemiology of Campylobacter in poultry. Vet. Q. 19:113 117. 20. Jacobs-Reitsma, W. F., A. W. Van De Giessen, N. M. Bolder, and R. W. A. W. Mulder. 1995. Epidemiology of Campylobacter spp. at two Dutch broiler farms. Epidemiol. Infect. 114:413 421. 21. Kapperud, G., E. Skjerve, L. Vik, K. Hauge, A. Lysaker, I. Aalmen, S. M. Ostroff, and M. Potter. 1993. Epidemiological investigation of risk factors for Campylobacter colonisation in Norwegian broiler flocks. Epidemiol. Infect. 111:245 255. 22. Luechtefeld, N. W., W.-L. L. Wang, M. J. Blaser, and L. B. Reller. 1981. Evaluation of transport and storage techniques for isolation of Campylobacter fetus subsp. jejuni from turkey caeca specimens. J. Clin. Microbiol. 13:438 443. 23. Nachamkin, I. 1995. Campylobacter and Arcobacter, p. 113 117. In P. R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, and R. H. Yolken (ed.), Manual of clinical microbiology. ASM Press, Washington, D.C. 24. Newell, D. G., and C. Fearnley. 2003. Sources of Campylobacter colonization in broiler chickens. Appl. Environ. Microbiol. 69:4343 4351. 25. Newell, D. G., and J. A. Wagenaar. 2000. Poultry infections and their control at farm level, p. 497 510. In I. Nachamkin and M. J. Blaser (ed.), Campylobacter, 2nd ed. ASM Press, Washington, D.C. 26. Patrick, M. E., L. E. Christensen, M. Waino, S. Ethelberg, H. Madsen, and H. C. Wegener. 2004. Effects of climate on incidence of Campylobacter spp. in humans and prevalence in broiler flocks in Denmark. Appl. Environ. Microbiol. 70:7474 7480. 27. Petrie, A., and P. Watson. 1999. Statistics for veterinary and animal science. Blackwell Science, Oxford. 28. Pezzotti, G., A. Serafin, I. Luzzi, R. Mioni, M. Milan, and R. Perin. 2003. Occurrence and resistance to antibiotics of Campylobacter jejuni and Campylobacter coli in animals and meat in north-eastern Italy. Int. J. Food Microbiol. 82:281 287. 29. Refregier-Petton, J., N. Rose, M. Denis, and G. Salvat. 2001. Risk factors for Campylobacter spp. contamination in French broilerchicken flocks at the end of the rearing period. Prev. Vet. Med. 50: 89 100. 30. Rosenqvist, H., H. M. Sommer, N. L. Nielsen, and B. B. Christensen. 2006. The effect of slaughter operations on the contamination of chicken carcasses with thermotolerant Campylobacter. Int. J. Food Microbiol. 12:226 232. 31. Stern, N. J., P. Fedorka-Cray, J. S. Bailey, N. A. Cox, S. E. Craven, K. L. Hiett, M. T. Musgrove, S. Ladely, D. Cosby, and G. C. Mead. 2001. Distribution of Campylobacter spp. in selected U.S. poultry production and processing operations. J. Food Prot. 64:1705 1710. 32. Swedish Institute for Infectious Disease Control. 2006. Statistik över smittsamma sjukdomar. Available at: http://gis.smittskyddsinstitutet. se/mapapp/build/intro.html. Accessed 2 November 2006. 33. Van De Giessen, A. W., M. Bouwknegt, W. D. Dam-Deisz, W. Van Pelt, W. J. Wannet, and G. Visser. 1998. Reduction of Campylobacter infections in broiler flocks by application of hygiene measures. Epidemiol. Infect. 121:57 66.