Human campylobacteriosis related to cross-contamination during handling of raw chicken meat: Application of quantitative risk assessment to guide intervention scenarios analysis in the Australian context.

ABSTRACT

Quantitative Microbiological Risk Assessment (QMRA) is a methodology used to organize and analyze scientific information to both estimate the probability and severity of an adverse event as well as prioritize efforts to reduce the risk of foodborne pathogens. No QMRA efforts have been applied to Campylobacter in the Australian chicken meat sector. Hence, we present a QMRA model of human campylobacteriosis related to the occurrence of cross-contamination while handling raw chicken meat in Western Australia (WA). This work fills a gap in Campylobacter risk characterization in Australia and enables benchmarking against risk assessments undertaken in other countries. The model predicted the average probability of the occurrence of illness per serving of salad that became cross-contaminated from being handled following the handling of fresh chicken meat as 7.0 × 10-4 (90% Confidence Interval [CI] ± 4.7 × 10-5). The risk assessment model was utilized to estimate the likely impact of intervention scenarios on the predicted probability of illness (campylobacteriosis) per serving. Predicted relative risk reductions following changes in the retail prevalence of Campylobacter were proportional to the percentage desired in the reduction scenario; a target that is aiming to reduce the current baseline prevalence of Campylobacter in retail chicken by 30% is predicted to yield approximately 30% relative risk reduction. A simulated one-log reduction in the mean concentration of Campylobacter is anticipated to generate approximately 20% relative risk reductions. Relative risk reduction induced by a one-log decrease in the mean was equally achieved when the tail of the input distribution was affected-that is, by a change (one-log reduction) in the standard deviation of the baseline Campylobacter concentration. A scenario assuming a 5% point decrease in baseline probability of cross-contamination at the consumer phase would yield relative risk reductions of 14%, which is as effective as the impact of a strategic target of 10% reduction in the retail prevalence of Campylobacter. In conclusion, the present model simulates the probability of illness predicted for an average individual who consumes salad that has been cross-contaminated with Campylobacter from retail chicken meat in WA. Despite some uncertainties, this is the first attempt to utilize the QMRA approach as a scientific basis to guide risk managers toward implementing strategies to reduce the risk of human campylobacteriosis in an Australian context.