Quantitative microbial risk assessment model to estimate the risk of Escherichia coli O157:H7 from fresh Australian apples

Fresh apples are among the most highly consumed fruits in Australia with ongoing expansion of domestic and export markets increasing consumption. Apples have emerged as being at potential risk of contamination with Escherichia coli O157:H7, Salmonella spp., and Listeria monocytogenes from the orchard and packhouse environments. Apples have not been associated with foodborne infection in Australia, however industry risk managers identified knowledge gaps of the effectiveness of food safety controls used in the industry and the level of risk that apples present. This recognition indicated the need for systematic and rigorous risk assessment to evaluate the risk, and sources of risk, to consumers from foodborne pathogens in apples. E. coli O157:H7 was chosen as an exemplar in the development of a stochastic, quantitative, risk assessment because Shiga toxin-producing E. coli (STEC), which includes E. coli O157:H7,can cause infections leading to Haemolytic Uraemic Syndrome (HUS) and Thrombotic Thrombocytopaenic Purpura (TPP). STEC and HUS are commonly notified diseases in Australia and there is substantial information on E. coli O157:H7 in horticultural products available in the literature that can be applied to quantitative risk assessment. The objective of the thesis was to assess the risk of E. coli O157:H7 from consumption of fresh Australian apples, determine the principal factors influencing risk, and develop strategies to minimise risk. Thus, a quantitative microbial risk assessment model was developed designed around the presence, concentration and ecology of known microbial hazards in inputs at each stage from orchard to consumer, which identified the critical risk-determining steps in the apple supply chain and, from that, the most effective risk management strategies. An extensive review of the literature revealed the lack of information and data relevant to current practices in Australian apple orchards and packhouses, the performance of food safety management systems in use, and the level of risk awareness among growers and packhouse quality assurance managers. However, relevant to this risk assessment were many appropriate data sources for the prevalence and concentration of generic E. coli and E. coli O157:H7, including irrigation water, birds and other wildlife in orchards, apple washing systems, and environmental monitoring in packhouses, and dose-response models for E. coli O157:H7. To address the data gaps identified, novel data on the range of industry food safety controls and the efficacy of those controls were studied in the Australian industry and, from those observations, incorporated into the modelling. The influence of food safety culture on risk awareness and understanding of microbial contamination among growers and packhouse operators was also investigated to gain insight on the application of current controls and provide additional information to industry risk managers to assist rational decision making. Current Australian orchard and packhouse practices related to food safety control were determined through an observational study that incorporated semi-structured interviews and a hygiene audit, with the aim to provide relevant and accurate input data to the risk model. The microbial hazards in orchards and packhouses were characterised, samples of wash water and apples analysed for microbial loads, and the relationship between microbial loads and fruit quality was assessed. Industry interpretations of food safety guidelines were also assessed. Overall food safety management in orchards and packhouses was systematically assessed based on business context, the application of food safety controls, and food safety assurance activities. The hygiene audit results showed that despite variable application of food safety controls and low assurance activities, foodborne pathogen contamination in apples was low at less than two percent, consistent with other tree fruit surveys. The performance of packhouse food safety management systems was assessed semi-quantitatively, to determine if industry practices are appropriate to minimise the risk from contamination of fruit with foodborne pathogens. Results were compared with the results of the observational study and between packhouses. Overall, a high level of food safety management performance was demonstrated. Comprehensive measurement of organisational and product context factors, and food safety control and assurance activities identified strengths including an intrinsically low level of non-conforming product, supportive management, and systematic use of feedback. System weaknesses were also identified including limited validation of food safety controls and a low requirement for formal food safety qualifications and knowledge among workers. The results confirmed findings from the observational study while providing detailed information of the management systems and the connections between their components to enable continuous improvement. The role of organisational culture as an enabler of food safety management was studied to assess how culture influences food safety outcomes and how food safety culture improvement can enhance continuous improvement in food safety systems. The mechanisms that drive culture improvement were assessed in relation to risk assessment and the implementation of risk management strategies. The role of food safety culture improvement in risk reduction was elucidated and strategies to progress food safety culture articulated. A stochastic risk assessment model was constructed to estimate the changes in E. coli O157:H7 prevalence and concentration, between the orchard and consumption, with the probability of illness per apple estimated. The risk assessment followed the approaches and guidelines recommended by FAO/WHO and Codex for microbiological risk assessment. The model was developed in Analytica® with each stage of the supply chain constructed with the key inputs affecting contamination with, and survival of, E. coli O157:H7. Integration of technological, organisational and sociological factors determined in the observational, performance measurement and food safety culture studies with the stochastic risk assessment model provides a comprehensive food safety risk management framework. The mean risk of illness from consumption of one apple was estimated to be 1.43 x 10-^-10 per serving with a standard deviation of 1.05x10^-8 . This risk level results in one case of E. coli O157:H7 illness every five years. Sensitivity analysis was conducted using Spearman rank correlation which identified use of chlorine sanitiser in wash water, removal of damaged apples, and equipment cleaning by washing and sanitising as having the greatest influence on the minimal risk to consumers. Hypothetical scenarios that reflected the range of industry practices were also explored to determine their predicted effect on the estimated risk, including the use of unsanitised wash water and a low level of verification activities. This thesis has developed information and knowledge on approaches to, and interpretations of, food safety management practices in orchards and packhouses, a food safety management system diagnostic tool to evaluate performance, a quantitative risk assessment model, and “a roadmap” to improve food safety culture, thus providing multiple tools and novel information on factors that influence packhouse microbial food safety risk management. The risk assessments provide valuable information to industry and regulators to inform risk-based decisions, and support risk management strategies for apples in Australia and New Zealand. Despite assumptions made in the modelling, many of the findings and outputs of this thesis can be applied to other horticultural products and in countries other than Australia and New Zealand.