Quantitative Risk Assessment Of Salmonella In The Broiler Supply Chain

Salmonellosis is one of the most important zoonotic diseases in public health.Human infection with Salmonella can lead to diarrhea,fever and intestinal inflammation and poultry infection with Salmonella can cause chicken pullorum.In China,70%-80%of food-borne illnesses were caused by Salmonella and more than 90%are meat products.The process of broiler supply chain is complex.Prevalence and contamination level of Salmonella would change through vertical and horizontal cross-contamination in whole supply chain,and the related risks are worth to pay attention.The quantitative microbiological risk assessment(QMRA)can identify the critical control points and evaluate the effects of intervention measures for reducing risk.At present,QMRA of Salmonella in broiler around the world focuses on the stage from retail to consumption.Few studies have been done on the QMRA model from farm to consumption.Due to differences in processing parameters or consumption,the developed QMRA model from farm to consumption was not applicable to the prediction of the risk illnesses for broilers in China.Therefore,it is urgent to construct a QMRA model from farm to consumption to predict the risk,determine the factors which could be contributing to the food safety risk and evaluate the effects of intervention measures for reducing risk.The main contents and results of this study are summarized as follows.(1)Analysis of Salmonella exposure pathways in broilers and determination of exposure assessment model parameters.Salmonella can contaminate chickens through farm,bleeding,scalding,evsceration,washing,potioning,storage,distribution,and retail.The key input parameters in the exposure assessment model were identified by analysis of exposure pathways,including processing parameters(disinfection concentration,time),environmental parameters(temperature),bacterial cross-contamination rate and contamination level.Based on the input parameters of the exposure assessment model required for the risk prediction,the key input parameters of the exposure assessment model were constructed by onsite investigation,microbiological testing,laboratory experiments,literature research and expert advice.The input parameters that coud not be obtained through onsite investigation,microbiological testing,literature research and expert advice,the laboratory experiments were carried out to obtain bacterial cross-contamination rate or contamination level in the exposure assessment model.(2)Bacterial inactivation model was developed in scalding process.Based on practical conditions in poultry scalding in China,the skin-on chicken wingette and skinless chicken breast were immersed in water at 50-70?.There were no significant differences of bacterial reductions between scalding temperature at 50 and 60?,and the treatment time was not a significant variable to inactivate the bacteria(p>0.05).A Logistic(-1.12,0.06)distribution was chosen based the result of Kolmogorov–Smirnov test to describe the bacterial reductions on chicken breasts at 50-60? within 100 s treatment.At the scalding temperature of 70?,the bacterial reductions declined with the treatment time(p?0.05).Weibull,Exponential and Log-linear models were used to fit the bacterial reduction and the Weibull model showed the best fit.For chicken wingette,no significant differences in bacterial reductions were found among three temperatures,and the treatment time was not a significant variable as well(p>0.05).A Logistic(-0.95,0.07)distribution could describe the bacterial reductions at 50-70?.The results of this study could provide parts of the input data for microbial risk assessment of the poultry supply chain in China.(3)Bacterial inactivation model was developed in washing process.Based on practical conditions in poultry washing in China,Na Cl O concentration(20 mg/L,50 mg/L and 100 mg/L)on the bacterial reduction in washing process was investigated.For chicken breasts and wingettes,there were no significant differences of bacterial reductions among20 mg/L,50 mg/L and 100 mg/L Na Cl O,and the treatment time was not a significant variable to inactivate the bacteria(p>0.05).A Normal(-0.75,0.1)distribution could describe the bacterial reductions on chicken breasts at 20-100 mg/L and a Normal(-0.6,0.1)distribution could describe the bacterial reductions on chicken wingettes at 20-100mg/L.The results of this study could provide parts of the input data for microbial risk assessment of the poultry supply chain in China.(4)Bacterial cross-contamination model was developed in washing process.To evaluate the combined effects of initial contamination level(1 log CFU/g,2 log CFU/g,3log CFU/g,4 log CFU/g and 5 log CFU/g),pre-chill incidence(3%,10%,22%,33%and40%)and Na Cl O concentration(0 mg/L,20 mg/L,50 mg/L,80 mg/L and 100 mg/L)on post-chill incidence of chicken breasts,a response surface model based on the central composite design was developed to predict the post-chill incidence.Statistical significance with contamination level,pre-chill incidence and Na Cl O concentration for bacterial post-chill incidence were found(p<0.05).The goodness-of-fit of a predictive model has a satisfactory performance as evidenced by statistical indices(pseudo-R²=0.9;p<0.0001;RMSE=0.21).The calculated B_f was 1.02,which lies in the acceptable range of 0.9-1.05.The A_f value was 1.11,revealing a merely 11%difference between the observations and predictions.These results indicated that the developed model could give a reliable prediction for post-chill incidence within the range of variables employed.The results of this study could provide parts of the input data for microbial risk assessment of the poultry supply chain in China.(5)This study developed a QMRA model for Salmonella in yellow broiler and white broiler supply chain.The growth/survival kinetics of Salmonella for model inputs at different stages was determined through onsite investigation,laboratory experiments and systematic studies of published literature.The mean predicted probability of illnesses per serving was 1.5×10^-8 in yellow broiler and 1.0×10^-8in white broilers,respectively.The results of sensitive analysis showed that full cooking(spearman correlation coefficient:0.6),contamination prevention in farm(spearman correlation coefficient:0.45)and bacterial inactivation at disinfection(spearman correlation coefficient:0.33)washing process in slaughter house are the critical control points for minimizing the health risk in yellow broiler.For white broiler,full cooking(spearman correlation coefficient:0.2),contamination prevention in farm(spearman correlation coefficient:0.45)and bacterial inactivation at scalding(spearman correlation coefficient:0.27)in slaughter house are the critical control points for minimizing the health risk.The results of the scenario analysis showed that the use of electrolyzed water,citric acid,lactic acid and lysozyme in chilling process could reduce the risk illnesses to 1.2×10^-9?2.2×10^-9?6.3×10^-9 and 1.2×10^-8 per serving,respectively.An effectively intervention method can reduce risk of foodborne illness.The risk assessment model needs to be further studied.Food Safety Objectives(FSO)often used in risk management of foodborne pathogens.According to the risk assessment model,setting the FSO standard and realizing the risk early warning is of great significance to risk management.