Focus on Food Safety: Listeria recall exposes flaws in CFIA’s ERA model

We should have learned our Listeria lessons, but the recall of plant-based beverages in late summer 2024 shows there is still much to be done to improve regulatory compliance.

Last summer, a major Listeria recall occurred at the Joriki Beverages plant in Toronto. The infection had caused more than 20 illnesses and three deaths. As with any food safety incident, root cause analysis is helping systematically map where errors occurred. Much of the current attention is focused on Canadian Food Inspection Agency’s (CFIA’s) establishment-based risk assessment (ERA) model for food establishments. This model, in its conceptual state, is quite useful. Rate the food safety risk within the company and then base the frequency and intensity of inspection based on the risk profile. This is a way to gain efficiency for tax-funded services.

In the ERA model, which was developed through consultation with food safety specialists across the country, a list of 173 factors were profiled and ranked by importance. These factors were then grouped into three categories: Inherent Risk Factors, Mitigation Factors, and Compliance Factors.

Inherent Risk Factors

Inherent Risk Factors include commodity, type of products, volume, type of activities conducted at the establishment, direct distribution to vulnerable populations, and processing steps. These inherent risk factors were then ranked by the health impact or disability adjusted life year (DALY) attributable to the commodity and product.

One issue that’s related to Inherent Risk Factors is the sorting of commodities apart from traditional ones like meat, seafood, dairy, eggs, honey, maple syrup, and processed fruits and vegetables into one large category of ‘Manufactured Foods’.

Manufactured foods run the gamut for risk profiling and are hard to quantify because of the vast range of products and process innovation that can occur within this category. Interestingly, they were ranked low for risk, despite the huge range of potential risk profiles in this category.

Plant-based beverages were classed as Manufactured Foods, as identified in an investigation by the Globe and Mail, thereby giving a lower risk rating than comparable commodities like dairy. However, the intrinsic properties of the beverage, including a pH close to neutral, a high level of fermentable substrate within the product, no water activity control, and lack of antimicrobial additives suggests this product is prime for pathogen survival and microbial outgrowth. Bacterial growth is highly likely if there is a failure in the pasteurization and aseptic filling process, or packaging contamination. Based on the Listeria Policy categorization, plant-based beverages would typically be rated as Category 1—highest risk product for the lack of intrinsic properties protecting the product.

In my work, I coach many companies on food safety and quality management and will frequently ask about their knowledge of inherent risk in their products. Do they know the water activity, pH, titratable acidity, Brix, and other various factors that define the risk basis of their products? When working with small businesses, which make up the bulk of our Canadian food manufacturing, the answer is quite often no. These companies do not have the sophistication of quality control programs to track, either at the specification level or lot level, the properties of the product that make it high or low risk.

Mitigating Factors

Mitigating Factors include applying additional processes and process controls, application of microbial sampling plans, evidence of food safety certifications and third-party audits, employment of qualified quality control personnel, and evidence of supplier verification schemes. All of these add robustness, but again, the presence of these programs, and the quality of performance of each of these elements within the ERA model is all based on single points in time.

Compliance Factors

Compliance Factors include information derived from CFIA activities and actions. This includes inspection results and impact assessment, history of enforcement actions, history of recalls and food safety confirmed complaints. With many establishments inspected under the Safe Food for Canadians licensing requirements, the frequency of inspection was already very low, making the quantity of these compliance factors consequentially low.

The ERA model and each establishment’s risk profile is predominantly based on self-reported data. Most of the rating criteria are not determined by audit or verification from an external body such as CFIA. Whenever organizations self-select their data and response, there’s always a bias to respond in a way that places the organization in the least risk possible category. It’s tempting to offer what you think is the easiest answer rather than reflect on worst-case scenarios. Secondly, if a company is doing product innovation, is a new risk categorization prompted by the system? Continuous improvement should be part of every company’s food safety management program. It is not clear how frequently establishments re-register within ERA to reflect updates and changes within their risk profile.

Based on an investigation from the Globe and Mail, it appears that 1443 out of approximately 8000 establishments did not answer the survey at all, thereby receiving low risk ranking without any verification from inspection.

The use of mathematical models, artificial intelligence, and algorithm-based decision-making is only as good as the data we feed. When data is missing, not correctly verified, or is not addressing the entire scope of the system, it will not reflect reality. In today’s technology-driven world, we become more and more reliant on it for answers. Meanwhile onsite observation, expert common sense and scientific consensus is still the prevailing way of assessing risk.

A full-scale audit of the ERA model will be performed in winter 2025. The entire sector is looking forward to understanding what went wrong. There is great anticipation for major improvements to bring trust back to this high impact food safety model.

Dr. Amy Proulx is professor and academic program co-ordinator for the Culinary Innovation and Food Technology programs at Niagara College, Ont. She can be reached via email at aproulx@niagaracollege.ca.

This column was originally published in the Feb./Mar. 2025 issue of Food in Canada.