Search

SUMMARYThis study assesses the contribution of different sources of human campylobacteriosis in Denmark using two different source-attribution approaches. In total, 794 non-human isolates and 406 isolates from human cases (domestic, travel related, and cases with unknown travel history) were collected. Isolates were characterized by multilocus sequence typing,flaAtyping and susceptibility to antibiotics. Both models used indicate that the major burden of human campylobacteriosis in Denmark originates from the domestic broiler chicken reservoir. The second most important reservoir was found to be cattle. The Asymmetric Island model attributed 52% [95% credibility interval (CrI) 37–67] to Danish chicken, 17% (95% CrI 3–33) to imported chicken, and 17% (95% CrI 7–28) to cattle. Similarly, theCampylobactersource-attribution model apportioned 38% (95% CrI 28–47) to Danish chicken, 14% (95% CrI 10–18) to imported chicken, and 16% (95% CrI 7–25) to cattle. The addition offlaAtype as an extra discriminatory typing parameter did not change the attribution of cases markedly.

Danish organic broiler meat, represented by carcasses sampled at the end of processing after chilling, was more frequently contaminated with thermotolerant Campylobacter spp. than conventional broiler carcasses; the yearly mean prevalence being 54.2% (CI: 40.9-67.5) for organic and 19.7% (CI: 14.8-24.7) for conventional carcasses. Campylobacter jejuni was the most frequently isolated species. The difference in prevalence was obvious in all quarters of the year. Contamination of organic and conventional broiler carcasses was more likely to occur in the warmer summer months, in this case in the third quarter, as also documented for conventional broiler flocks. When contaminated, the mean concentration of Campylobacter on neck skin samples of organic and conventional carcasses was not significantly different (P=0.428); 2.0±0.65 log 10 cfu/g and 2.1±0.93 log 10 cfu/g, respectively. Assessing the relative risk of becoming ill following exposure to Campylobacter on conventional or organic broiler meat indicated that the risk per serving from organic carcasses was 1.7 times higher than that of conventional carcasses. The higher risk of illness from organic broiler carcasses compared with conventional broiler carcasses emphasizes the importance of implementing control measures in organic broiler production.

In Denmark, the incidence of human campylobacteriosis cases, as well as the Campylobacter prevalence in broiler flocks, is strongly influenced by season with a summer peak in July-August. Therefore, it was considered that the prevalence of Campylobacter in broiler meat sold at retail in Denmark might also be influenced by season. A retrospective survey analysis was performed on 2001-2007 national surveillance data of the prevalence of thermotolerant Campylobacter in all conventional broiler flocks at slaughter, and in randomly sampled broiler meat at retail. There was a significant effect of season on the occurrence of Campylobacter in meat at retail; the largest effect was found for domestic chilled meat. Thus, the Campylobacter prevalence in Danish broiler flocks, which fluctuated with season, was found to be a strong predictor for the occurrence of Campylobacter in fresh, chilled, Danish broiler meat. However, besides flock prevalence, there was also a direct effect of season on the occurrence of Campylobacter in Danish broiler meat at retail.

Campylobacter jejuni in fresh chilled chicken meat is known to be a major risk factor for human gastrointestinal disease. In the present study, the survival under chilled conditions of different C. jejuni strains exposed to different gas mixtures usually used for gas packaging of food was examined. Bolton broth and fresh, skinless chicken fillets were inoculated with six and four strains, respectively, and exposed to the gas mixtures 70/30% O(2)/CO(2), 70/30% N(2)/CO(2), and 100% N(2) (the latter only investigated in broth) at refrigeration temperature (4-5 degrees C). In broth culture, the strains survived significantly longer when exposed to 100% N(2) and 70/30% N(2)/CO(2) than in the oxygen-containing gas mixture, 70/30% O(2)/CO(2) (P0.0001). For the two anaerobic gas mixtures, the reductions only reached 0.3-0.8 log(10) CFU mL(-1) within the same period. In the presence of oxygen, the numbers of C. jejuni were reduced by a minimum of 4.6 log(10) CFU mL(-1) over 21 days. When inoculated onto chicken fillets, the C. jejuni strains also died significantly faster in the oxygen-containing gas mixture, 70/30% O(2)/CO(2) (P0.0001), reaching reductions of 2.0-2.6 log(10) CFU g(-1) after 8 days. In the gas mixture without oxygen (70/30% N(2)/CO(2)), no reductions were observed.

When assessing effects of decontamination techniques on counts of Campylobacter spp. on broiler meat, it is essential that the results reflect the variations that may exist. Decontamination studies often use high inoculation levels (10(7) to 10(8) cfu) and one or few strains of Campylobacter jejuni, thereby restricting the results to reflect only a limited part of the true situation. This study presents results from physical and chemical decontamination of broiler meat medallions using different strains and different initial concentrations of C. jejuni. For 3 strains of C. jejuni, mean log reductions obtained by freezing at -20°C for 7 d was significantly higher for an initial concentration of 10(7) cfu/sample on the meat compared with an initial concentration of 10(3) cfu/sample. For freezing at -20°C for 24 h or application of 6% tartaric acid and subsequent storage for 24 h, no statistically significant difference in reductions was found for initial concentrations ranging from 10(3) to 10(7) cfu per sample. The mean log reductions obtained by all techniques were strongly dependent on the strain tested. The results reveal that reductions obtained with high inoculation levels of C. jejuni (10(7) cfu/sample) or single or few strains of the species (or both) should not be interpreted as a generic result for the species. If inoculation studies cannot be replaced by investigations of naturally contaminated meat, we advise using a mixture of strains found in the production environment at levels as close as possible to the natural contamination level.

This paper describes a new approach by which changes over time in the relative risk of human campylobacteriosis from broiler meat are evaluated through quantitative microbiological risk assessment modelling. Danish surveillance data collected at retail from 2001 to 2010 on numbers of thermotolerant Campylobacter spp. on Danish produced and imported chilled and frozen broiler meat were the basis for the investigation. The aim was to explore if the risk from the different meat categories had changed over time as a consequence of implemented intervention strategies. The results showed a slight decrease from 2005 to 2008 in the human risk from Danish produced broiler meat, and a decrease from 2005 to 2010 in the risk from imported chilled meat. This risk reduction coincides with control measures implemented to reduce Campylobacter in Danish and imported chilled broiler meat. The human risk of campylobacteriosis from Danish frozen meat increased but remained lower compared to chilled meat. In total, the relative risk from broiler meat available for sale in Denmark increased from 2001 to 2005 after which the risk decreased to a level similar to the period 2001–2002. The use of QMRA in the evaluation of intervention strategies based on monitoring data provided an added value, compared to the traditional approach of only using changes in prevalence. The estimated human health risk is a function of prevalence and the distribution of concentrations, and therefore takes best usage of the available data, while providing the most relevant outcome for food safety risk managers.

To reduce the incidences of human Campylobacter infections, a number of countries are investigating methods for reducing human exposure to Campylobacter from broiler meat. In addition to implementing biosecurity measures at the farm, Campylobacter may be controlled by reducing Campylobacter counts through physical decontamination of the meat. The current study was conducted to compare the Campylobacter-reducing ability of three physical decontamination techniques, forced air chilling, crust freezing, and steam-ultrasound, performed in the plant with naturally contaminated broiler chickens. The effects of all three techniques were evaluated and compared with the effect of freezing. Mean reductions obtained were 0.44 log CFU per carcass, 0.42 log CFU per sample, and > or = 2.51 log CFU per carcass, respectively. All techniques resulted in significant reductions of the Campylobacter concentration on the carcasses (P < 0.05). However, none of the techniques were as effective as freezing based on reductions in Campylobacter counts and on adverse effects. The increase in Campylobacter counts on carcasses following visceral rupture during the evisceration operation also was examined. Visceral rupture resulted in an increase of 0.9 log CFU per carcass, suggesting that Campylobacter counts also may be reduced by optimizing the hygienic design of equipment or by physical removal of fecal contamination.