Exposure to whole chicken carcasses may present a greater risk of campylobacteriosis compared to exposure to chicken drumsticks.

In New Zealand, the major risk factor for campylobacteriosis has been identified as poultry consumption. New Zealanders consume different types of chicken meat which undergo different processing before entering the retail chain. The manipulations and jointing of chicken carcasses into pieces and the subsequent processing and packaging have the potential to cross-contaminate and reshuffle bacterial pathogens among the different products sold. The aim of this study was to analyse: (a) the differences in the viable count and population genetic structure between Campylobacter isolated from chicken drumsticks and whole carcass meat for retail sale over a 1-year period; and (b) the genetic relatedness of human and chicken isolates collected concurrently. Enumeration of Campylobacter was performed using a spiral plater combined with manual spread plating. Campylobacter isolates were identified by polymerase chain reaction and typed by multilocus sequence typing (MLST). C. jejuni was the dominant species among both whole carcasses (63.5%) and drumsticks samples (73.8%), followed by C. coli (27% and 23.1%, respectively). After sample weight adjustment, whole carcasses showed significantly higher Campylobacter counts than drumsticks, with a significant difference in the counts between the commercial suppliers in both types of retail meat. MLST revealed 28 different sequence types among the two types of meat. Using permutational multivariate analysis of variance, statistically significant differences in the population genetic structures were observed between different suppliers but were not observed between the two types of chicken retail meat. In conclusion, we found differences in Campylobacter viable counts, suggesting consumption of whole carcasses may determine an exposure to a higher number of Campylobacter bacteria than consumption of chicken drumsticks. The Campylobacter population genetic structure did not differ between the two types of chicken retail meat. Therefore, source attribution studies based on MLST are unlikely to be biased by the selection of these types of retail meat during sampling.