Monitoring food safety hazards in the pork supply chain

According to the General Food Law (Regulation (EC) 178/2002), food business operators (FBOs) are responsible for the production of safe products. Apart from implementing quality systems, regular monitoring for the potential presence of microbial or chemical hazards is, therefore, necessary to control food safety. Such monitoring programs are preferably risk-based focusing on the most relevant hazards and samples. Within the pork supply chain, monitoring is performed at the slaughterhouse and currently includes monitoring on, among others, Toxoplasma gondii, Mycobacterium avium and antibiotic residues. The aim of the research described in this report was to optimize the monitoring for the above-mentioned food safety hazards and evaluate whether other relevant or emerging food safety hazards should be monitored at the slaughterhouse that are currently not included in the monitoring program. For this purpose, we performed the following three studies: 1. we explored which factors were related to the presence of T. gondii, M. avium and antibiotic residues; 2. we explored whether antibiotic screening can be performed more cost-efficiently; and 3. we identified emerging risks via social media analysis and expert elicitation. The following sections summarize these results. 1. Farm factors related to T. gondii, M. avium and antibiotic residues Literature review, Bayesian Network (BN) modelling and statistical analyses were performed to identify which farm factors influenced the presence of T. gondii, M. avium and antibiotic residues. The literature review revealed that limited information was available for M. avium. For T. gondii and antibiotic residues, multiple factors can influence their presence. The most relevant factors in both cases were the farm characteristics - herd size and farm type (especially outdoor access) - as well as biosecurity measures such as cleaning and disinfection, handling visitors, ventilation, and possible access of other animals (such as cats (i.e. for T. gondii), rodents and birds). Subsequently, BN modelling was performed based on a large data set collected at the slaughterhouse level from 2015 – 2022 on T. gondii and M. avium. However, since only a limited number of samples were positive for the specified pathogens, BN modelling demonstrated to be less useful for exploring relevant farm factors. Statistical analyses showed that the time period, number of transactions (i.e., farm size), and outdoor access are factors influencing the T. gondii and M. avium detection probability. These factors could be included in a risk-based monitoring program. Further research is needed including data on biosecurity measures in order to establish whether and how these factors influence pathogen presence. 2. Cost-effective antibiotic screening Antibiotic screening, in general, shows that the percentage of non-conform antibiotic samples is very low (around 0.05%). Cost-effectiveness analyses indicated that on-site antibiotic screening using swabs in combination with lateral flow tests (LFD) could be an alternative to the current screening procedure based on lab analyses only. This hypothesis was tested at the pilot scale in which a tetracycline lateral flow device (LFD) was used on pigs treated with doxycycline. This showed that swabs collected directly after treatment appeared barely positive, and swabs collected at the end of the withdrawal period were already (visually) negative. A further comparison at the lab scale showed that the sensitivity of the LFD was approximately 8-fold lower for doxycycline compared to oxytetracycline. Additional sampling of oxytetracycline-treated animals indicated that this alternative screening procedure might be feasible for animals treated with either of these tetracyclines, but more testing is needed to determine to what extent >MRL tissue levels correlate with positive LFD read-outs. 3. Emerging risks In order to establish whether the current monitoring program at the slaughterhouse should be expanded for alerting new or emerging risks, social media were evaluated using text mining, grey literature was screened, and experts were interviewed. Although the evaluation of social media did not reveal any unknown food safety issues, the methodology applied can be used for future evaluations to explore relevant topics for the pig supply chain and the sentiment towards these topics. The grey literature and expert interviews showed that upcoming changes in the pig supply chain, such as climate change and changes in farming systems, are expected to impact the presence of pathogens and mycotoxins. It is recommended for FBOs to keep track of the trends and keep monitoring pathogens in pork and mycotoxins in feed to establish trends in the data.