Listeria monocytogenes outbreaks and the development of a quantitative risk assessment model to investigate the impact of varying levels of Listeria monocytogenes

Listeriosis outbreaks

Listeriosis is a severe, foodborne infection linked with higher case fatality and hospitalization than other bacterial pathogens. The CDC estimated 1,600 people get listeriosis yearly, and about 260 die. The infection is most likely to sicken susceptible populations such as pregnant women and newborns, older people (aged 65 or older), and people with weakened immune systems.

According to the CDC there were two outbreaks in the US in 2022. An outbreak caused by Listeria monocytogenes was found in Big Olaf Ice Cream. The outbreak caused 28 people to get sick, 27 hospitalizations, and one death. A second outbreak was caused by Brie, baked Brie, and camembert cheese made by Old Europe Cheese, Inc., and sold under multiple brand names. The products caused six illnesses and five hospitalizations. In 2021, four outbreaks due to Listeria caused 44 people to get sick, 41 to be hospitalized, and six deaths. One of the deadliest outbreaks in the US happened in 2011. The outbreak involved 147 cases associated with cantaloupes.

The outbreaks due to listeriosis are seen worldwide. For example, according to a recently published study in Spain, a Listeria outbreak caused by stuffed pork sickened >200 people in 2019 and killed three. The outbreak was the largest listeriosis outbreak in Spain and one of the biggest in Europe. Spain has seen a rising trend in such infections since 1997 based on hospitalization records (Martinez et al., 2022). Listeriosis outbreak caused by contaminated stuffed pork, Andalusia, Spain, July to October 2019, Euro surveillance, Vol 27, 43.

In 2017 and 2018, the world’s largest and deadliest listeriosis outbreak occurred in multiple provinces of South Africa. It was caused by contaminated polony, a ready-to-eat, processed meat product. 1,060 cases and 216 deaths were reported during the outbreak period from January 11, 2017, to July 17, 2018.

In Germany, a large nationwide outbreak that included 134 laboratory-confirmed L. monocytogenes isolates from 112 patients outbreak linked to Blood Sausage in 2018-2019.

In Italy, at least 90 people have been sickened, and three have died in a large Listeria outbreak. The outbreak’s source is probably Listeria in sausages made with poultry meat produced by Agricola Tre Valli.

Listeriosis incidence (cases per 100,000 population) has not decreased during the last five years. According to the CDC, there were 0.24 incidents in 2014 and 0.26 in 2018, and ECDC reported 0.46 in 2014 and 0.47 in 2018. Therefore, Listeria remains a significant reason for foodborne disease.

Risk assessment models with Monte Carlo probabilistic simulation

Quantitative risk assessment is the most practical method to assess the risks associated with microbial contamination of foods. These risk analysis models are developed by combing Monte Carlo probabilistic simulation with probability distributions inserted into the model. Probability distribution models can utilize normal, lognormal, uniform, triangular, PERT, and discrete distributions.

The model then repeatedly calculates the results, each time using a different set of random values from the input probability distributions. A Monte Carlo simulation involves thousands of recalculations before it is complete. The result of a Monte Carlo simulation is a distribution of possible outcome values. The data enables to calculate the probabilities of different outcomes.

To use the Monte Carlo model, one can use a commercial program such as @RISK . The FDA has its risk model FDA-iRISK model.  Users enter data to build scenarios that reflect the hazard and issues of interest, with support from built-in, standard data-entry templates. FDA-iRISK, with Analytica® Decision Engine software, then uses a custom, built-in equations and algorithms to perform Monte Carlo in pre-structured models.

The risk model developed for Listeria

A quantitative risk assessment model was developed (Sampedro et al. 2022, Quantitative risk assessment model to investigate the public health impact of varying Listeria monocytogenes allowable levels in different food commodities: A retrospective analysis. Int. J. Food Microbiol., 383 (16 December 2022).

The main purpose of to estimate the probability of infection with Listeria monocytogenes in various ready to eat foods at various levels of contamination with the pathogen. The effect of the removal of highly contaminates samples was evaluated.

The model assumptions

The susceptible population was assumed to be 10–10,000×, and the total number of estimated cases in the US was 1,044 and 2,089, using the FAO/WHO dose-response models. The model assumed that >90% of cases were due to deli meats, followed by RTE salads (3.9–4.5 %), soft and semisoft cheese, and RTE seafood (0.5–1.0 %) and frozen vegetables (0.2–0.3 %). The at-risk population was responsible for 96.6–98.0 % of the total cases, with the highly susceptible population responsible for 46.9–80.1 %.

The total number of samples evaluated in the study worldwide (WD) to estimate the prevalence of L. monocytogenes in food commodities over the last 30 years (1990–2020) was around a half-million samples (62.5 % collected in the US). RTE deli meats were the most sampled category representing 67% (WD) and 87 % (US). Overall L. monocytogenes prevalence ranged from 1.4–9.9 % worldwide (WD) and 0.5–3.8 % (US). Prevalence rates were significantly lower in the US (1.3–5.2×) than those reported worldwide. For RTE food categories (RTE seafood, RTE deli meats, and soft and semisoft cheeses), whose values correspond to 0.5–10.3 % in 2008–2015 and 0.8–2.7 % in 2016–2018. L. monocytogenes enumeration data were limited and commodity-dependent. Total positive samples ranged from 26.2–56.5 % (WD) and 8.2–86.3 % (US) for the enumeration rates.

Model results

Samples with counts higher than 100 CFU/g varied between 6.4–36.2 % (WD) and 5.5–35.7 % (US).WD counts above 100CFU/g were highest for soft/semisoft cheese, RTE seafood, US frozen vegetables, and RTE deli meats. Mean L. monocytogenes counts ranged from 0.2 to 1.6 log CFU/g (WD) and −0.5 to 1.6 log CFU/g (US).

A proposal investigated was to include a quantitative microbiological criterion (i.e., <100 CFU/g) and a sampling scheme to test every lot and remove the highly contaminated lots from the market by changing the specifications from absence in 25 g to a quantitative plan (either a two-class or three-class). The lots with counts higher than 0.1 log CFU/g ranged from 30.9-87.6%, for counts of than 1.0 log CFU/g and 1.9–37.9%,

Removing lots with counts higher than 1 CFU/g yielded a prevalence reduction of 15.7–88.3 %, whereas removing samples higher than 100 CFU/g reduced the prevalence to 4.5–33.3 %. Also, removing samples with counts higher than 1 CFU/g decreased the number of cases by 61.1–100 %, whereas using a 100 CFU/g limit, the reduction varied between 31.8 and 96.6 %.