A procedure for surveillance data-driven risk assessment to inform Campylobacter risk-based control

Abstract

In this study is presented a procedure for surveillance data-driven risk assessment, which can be used to inform inter-sectorial Campylobacter risk-based control, e.g. within National Action Plans and One Health (OH) systems. Campylobacter surveillance data (2019 to 2022) and a published quantitative microbial risk assessment (QMRA) model were used, to show the procedure. Moreover, an interface tool was developed in Excel for showing descriptive statistics on measured apparent flock prevalence (AP) and concentrations (colony forming units per gram, cfu/g) on the meat, together with their related QMRA outputs. Currently (mid-2024), Danish fresh broiler meat is produced by four slaughterhouse companies (A, B, C and D), where approximately 30 % of the annually slaughtered broiler flocks are randomly culture tested, on one leg skin (LS) sample per flock sampled from chilled carcasses. Data variables were: date of sampling, farm-ID, within farm house-ID, flock-ID, slaughterhouse name, sample-ID, and Campylobacter concentrations. Flocks were classified as carcass positive with a concentration ≥ 10 cfu/g. The data was fed into the QMRA model to assess: a) the average risk of human campylobacteriosis per serving (during a month or year), and b) the monthly/annual risk of 2022 relative (RR) to the baseline (average) risk from the previous three years. The descriptive statistics and the risk assessment (RA) were carried out at national level and for each slaughterhouse. In 2022, the national RR was 1.03, implying that the average annual risk increased by approximately 3 % compared to the baseline. Nevertheless, for slaughterhouses A, B and D, the annual risk decreased by ≈ 22 %, 21 % and 43 %, respectively; whereas for slaughterhouse C it increased by 48 %. Monthly risk estimates showed seasonal variations, according to the visualized changes of AP and meat contaminations. The national monthly RR was >1 in July and from September to December. During those months: slaughterhouse C had always RR > 1, slaughterhouse A had a relative increase of risk in July, slaughterhouse B in July and November, and slaughterhouse D in October and December. The procedure and the tools used in this study, allow identifying the impact of seasonality and food-chain stages (i.e. slaughterhouses and their broilers sourcing farms) on the risk per serving, so that Campylobacter risk-based control could be implemented accordingly, from farm to fork, across consecutive surveillance periods. The same principles could be applied in other countries, food chains, and/or for other foodborne pathogens, when similar data and QMRA models are available.