Researchers Develop Flexible Model to Assess Supply Chain Contamination Risks in Fresh Produce

Researchers have developed a flexible computer model they believe will help the industry compare potential supply chain contamination risks as well as control strategies to reduce them. 

It also will allow industry groups to run “what-if” scenarios to gauge the impact of newly identified risks or control strategies on the overall supply chain compared to current practices.

Addressing Supply Chain Contamination Risks

A team led by Matthew Stasiewicz, Ph.D., with the University of Illinois at Urbana-Champaign, first modeled supply chain risks for leafy greens contaminated by Shiga toxin–producing Escherichia coli (STEC). They are expanding their efforts both to compare small-scale deviations from best practices and model Salmonella and melons.

Joining him as co-principal investigator is Martin Weidmann, Ph.D., with Cornell University. Also involved is Gabby Pinto, a Master of Public Health and Ph.D. candidate at UIUC.

Stasiewicz said:

"Weidmann brings to the project a really good sense of what information the produce industry needs to make good decisions, and he really knows how to communicate that to the right people, so they can use that information to make the right decisions.” 

This helps them do modeling work that can have impact.

The Importance of Data Quality in the Risk Model

The researchers focused initially on STEC in leafy greens because of both public health significance and data available in published models and risk assessments in literature. 

Stasiewicz said:

“The quality of a statistical model is only as high as the quality of the data you put into it”.

In simple terms, developing a supply chain risk model involves identifying the stages of product production. Each stage is then described mathematically.

In this case, it’s fresh produce, which includes growing, harvesting, processing, retail and consumer handling. Of the five stages, Stasiewicz said they decided to focus on where the produce industry has control and not model specific risks or management at consumer handling. He said:

“Can we represent those steps with math in relation to the thing we’re trying to track?”.

Collaborating with Industry to Address Emerging Risks

The goal of their CPS-funded project was to develop a flexible risk model for STEC and Listeria monocytogenes in leafy greens to evaluate different contamination scenarios and management strategies. During the first year, they looked at how common industry practices, like improved process wash and additional product testing, affected the risk of a positive test at retail and of removing lots with the highest potential contamination levels. As part of their analysis, they examined risk model sensitivity. He said:

“When you change things in the model, what else changes? What things are the outcomes most sensitive to? If you change various practices, how much does that change the outcome?”

Once the researchers solidified what Stasiewicz described as the backbone model, they met with CPS Industry Advisory Council members and industry representatives to understand other risks or concerns. For example, could the model also address the risk of small-scale deviations from best practices, such as discovering an irrigation water treatment system failed for a short time, how bad is this and how aggressively should one respond?  

As part of engaging with the industry, Pinto said they’ve had informal discussions with CPS Symposium attendees the past few years. Not only did it allow them to learn from the industry but it also helped get the word out about their research.

During the project’s third year, the researchers are developing a risk model for a set of small-scale deviations like irrigation water treatment failure, incomplete harvester sanitation, small animal intrusion and poor wash water control. They also hope to engage in industry discussions around managing Salmonella in melons. The two-step process involves figuring out the math based on current science and the likelihood of contamination. Stasiewicz said:

“Once you have that math, how do you take it and run it through a process model?”.

As part of the overall project, they’re creating a relatively simple interactive webpage that requires only a handful of entries before running different scenarios. Stasiewicz said the model is intended for produce groups or associations involved in developing policies, rules or best management practices. He said:

“There are people in those discussions that understand these risk models. We hope that they would pick up these tools to help drive the discussions forward.”

About CPS: Promoting Produce Safety

The Center for Produce Safety (CPS) is a U.S.-based non-profit organization that focuses on providing the produce industry with actionable information to enhance food safety. CPS is committed to promoting safe produce through research, risk management tools, and collaboration with industry stakeholders.