Chilling control of beef and pork carcasses in a slaughterhouse based on causality analysis by graphical modelling. (December 2020)

• Record Type: Journal Article
• Authors: Kuzuoka, Kumiko; Kawai, Kohji; Yamauchi, Syunpei; Okada, Ayaka; Inoshima, Yasuo
• Is Part Of: Food control. Volume 118(2020)
• Publication Date: 2020-12
• Subjects: Carcass chilling -- Causality analysis -- Graphical modelling -- Hazard analysis and critical control point -- Multivariate data analysis -- Risk management; FBO food business operator -- HACCP Hazard Analysis and Critical Control Point -- MRA multiple regression analysis -- MDA multivariate data analysis -- GM graphical modelling -- SEM structural equation modelling -- rij.rest partial correlation coefficient -- rij correlation coefficient matrix R -- rij invertible matrix R−1 -- NFI normed fit index -- dev deviance -- RM reduced model -- i number setting partial correlation coefficient to 0 -- NM null model -- n number of samples -- ∏(i)ˆ estimate of population correlation coefficient of RM -- |⋅| determinant -- R sample correlation coefficient -- FM full model; Food -- Quality -- Periodicals; Food -- Analysis -- Periodicals; Food handling -- Periodicals; Food industry and trade -- Quality control -- Periodicals; Aliments -- Industrie et commerce -- Qualité -- Contrôle -- Périodiques; Aliments -- Qualité -- Périodiques; Aliments -- Analyse -- Périodiques; Hygiène alimentaire -- Périodiques; Food -- Analysis; Food handling; Food -- Quality; Periodicals; Electronic journals; 664.07
• DOI: 10.1016/j.foodcont.2020.107353
• Languages: English
• ISSNs: 0956-7135

Abstract

Abstract: Failure to maintain carcass chilling in slaughterhouses results in the growth of food poisoning bacteria, a serious food safety hazard. Therefore, skilled and experienced food business operators (FBOs) are needed to operate the chilling system at slaughterhouses. However, to prevent serious incidents of carcass chilling failure, the chilling process must be managed by understanding all the influencing factors based on scientific evidence, which is demanded by the Hazard Analysis and Critical Control Point (HACCP) system. Appropriate control of carcass temperature requires an accurate understanding of extrinsic and intrinsic factors present after slaughter and dressing. This study aimed to clarify the influencing factors (affectors) of the chilling processes for beef and pork carcasses in a slaughterhouse using graphical modelling (GM). GM is a multivariate data analysis method that has been widely used for statistical causality analysis by visual and flexible modelling. GM was carried out using the following parameters: outside temperature and humidity, number of carcasses in a chilling room on each operating day and during every afternoon operation, clock time of sealing a chilling room, the preset temperature in a chilling room, chilling room temperature at 16:30 (JST) of the slaughter and dressing day and 8:00 (JST) of the next day, and surface and core temperatures of carcasses. The results showed that the major affector of carcass temperature was not the outside temperature, as stated by FBOs, but rather the number of carcasses in the chilling room. The outside temperature was well-compensated by the performance of the chilling equipment. Furthermore, seasonal fluctuations in the number of livestock brought into the slaughterhouse, daily fluctuations in slaughter and dressing efficiency, and performance of chilling equipment were considered as lurking factors. Even skilled FBOs believe that incidents of carcass chilling failure cannot occur during low-temperature seasons. However, incidents of carcass chilling failure occur in this season because of the overloading of chilling rooms with carcasses. The charts determined by GM in this study can help FBOs visually understand the affectors of the chilling process. Education and understanding of stakeholders, including FBOs, are essential for risk management. In addition, our findings provide a foundation for generating layers and nodes in a multilayer perceptron, which is an artificial neural network that can control the chilling process automatically. Highlights: This study aimed to identify the affectors of carcass chilling. Carcass number in the chilling room most significantly affected chilling. The influence of outside temperature on carcass temperature was negligible. Quantitative risk assessment is useful to prevent chilling failure. Graphical modelling charts facilitate better control of carcass chilling.