Developing sampling simulation and quantitative microbial risk assessment to improve food safety management systems

Challenges in a modern food system can be defined with large-scale production and hazards that are rare and not evenly distributed. With these characteristics, traditional experimental research design with smaller sample size can be limited to answer current issues in food safety dealt with in this study such as (1) How powerful is given sampling plans in powdered food product production? (2) Can targeting certain levels or/and serotypes of Salmonella in raw poultry products help reducing the public health risk? Tools like simulation, modeling, Quantitative Microbial Risk Assessment (QMRA) can be helpful in answering these questions combined with numerous data collected from government and industry. These tools can provide the ability to develop science-based food safety management strategies that are specific to different processing environments using collected data. This research focused on answering rising questions in the current food safety issues using combined data analytics. The objectives of this PhD research were (1) Adapt previously developed 2-dimensional and 3-dimensional sampling tool for 1-dimensional sampling tool specific for powdered food products. (2) Examine the power of representative sampling plans with realistic contamination profiles in recalled and non-recalled PIF batches from a published paper using developed 1D sampling simulation. (3) Develop a QMRA for assessing risk in certain levels and serotypes of raw poultry products. (4) Assess public health risks in raw chicken parts contaminated with different levels of all Salmonella serotypes and high-and low-virulent serotypes using QMRA. To achieve the first two objectives, we adapted a previously published sampling simulation to PIF sampling and benchmarked industry-relevant sampling plans across different numbers of grabs, total sample mass, and sampling patterns. Overall, (1) systematic or stratified random sampling patterns are equal to or more powerful than random sampling of the same sample size and total sampled mass, and, (2) taking more samples, even if smaller, can increase the power to detect contamination. The rest of the research objectives were addressed by developing QMRA. Our risk assessment suggests public health risk in chicken parts is concentrated in the small proportion of finished products contaminated with high-levels and specifically high-levels of high-virulent serotypes. Low-virulent serotypes, such as Kentucky, are predicted to contribute to extremely few human cases. In conclusion, data analytics tools used in this research such as simulation, modeling and QMRA combined with numerous industry and government testing data, showed that they can help develop better food safety management strategies to deal with current issues in the food industry. We are now living in the era of big data, and the ability to explore massive data will guide us to develop proper strategies to deal with evolving foodborne pathogens.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Food Science & Human Nutrition

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2024

Author dc:creator

• Kim, Minho

Contributors dc:contributor

• Stasiewicz, Matthew J
• Miller, Michael J
• Banerjee, Pratik
• Ozturk, Oguz

• Salmonella
• Qmra
• Serotype

Statement dc:rights

• Copyright 2024 Minho Kim

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/124257

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F124257