A model using an inter-sectorial data integration process indicates that reducing Campylobacter cross-contamination at slaughter mitigates the risk of human campylobacteriosis effectively

The risk of human campylobacteriosis due to Danish broiler flocks cross-contaminated (CC) at slaughter with Campylobacter spp. was assessed. National surveillance data (2018) on flock Campylobacter status (called animal level component (AL)) and on contamination of chilled carcasses ready for consumption (food safety component (FS)), were compared. The AL component consisted of PCR testing results on pools of cloacal swabs collected from 3,012 conventional flocks, while the FS component presented culture testing of leg skins from 999 (of the 3,012) randomly sampled flocks. Datasets were integrated on flocks tested in both components, by combinations of farm-ID, house-unit and sampling date. The CC flocks were those entering the slaughterhouse as AL-negative, but resulting FS-positive. All remaining carcass positive flocks were instead classified as Non-CC flocks. The apparent prevalence (AP) of carcass positive flocks and the colony forming units per gram (cfu/g), measured by the FS component, were fed into a published simulation model, to assess under three simulation scenarios: the mean monthly risk per serving during 2018, relative (RR) to that of 2013 (reference year in the current Danish Action Plan). In the baseline scenario, the original AL status and the FS cfu/g were maintained. In the alternative scenarios I and II, the FS cfu/g were set = 0 (i.e. negative) for the Non-CC and for the CC flocks, respectively. Thus, scenario I and the differences between the other two scenarios, provided the contribution of the CC flocks to the AP and to the RR. The (overall) annual median log10 cfu/g was ≈ 2.8 (min. = 1.0; max. = 4.0) for the Non-CC flocks and 1.4 (1.0; 3.9) for the CC flocks. The median monthly difference in AP, between the baseline scenario and scenario II was 7% (min = 2% in January; max = 19% in August), while the difference in risk was 0.04% (0.001%; 0.11%), which was similar to the