Researchers have developed a new, farm-to-fork quantitative microbial risk assessment (QMRA) model to evaluate the risk of Listeria monocytogenes contamination in fresh-cut cantaloupe, and ran simulations using the model in a new study. Published in the Journal of Food Protection, the study was led by researchers at Cornell University and funded by the U.S. Department of Agriculture National Institute of Food and Agriculture (USDA-NIFA).

Overall, the QMRA’s estimates underscore the importance of a multi-pronged approach to L. monocytogenes risk reduction in fresh-cut cantaloupe. Temperature control post-packing, effective washing practices, and improved data collection, especially at the field level, are crucial.

Model Development and Simulations

The model considers conditions across the cantaloupe supply chain, including field-level harvest, intermediate handling (i.e., in cooling facilities or packinghouses), fresh-cut processing, distribution, retail, and home storage. The researchers ran simulations to provide estimates of L. monocytogenes concentration in a single serving of cantaloupe (134 grams) and to estimate annual illnesses and deaths in the U.S. attributed to L. monocytogenes-contaminated fresh-cut cantaloupe. 

Importantly, the absence of data on L. monocytogenes prevalence and concentration at the field level restricted the model’s scope and prevented assessment of pre-harvest interventions. Moreover, due to data constraints, cross-contamination between facility surfaces and produce was modeled simplistically. While this had minimal impact on illness estimates, it influenced contamination prevalence.

Although the model is not ready for immediate use by industry as it does not allow for operation-specific data inputs, it provides a framework for identifying and evaluating interventions. The model, as well as data and code from the study, can be accessed via GitHub.

Findings and Implications for Industry

The QMRA estimated a median probability of illness per serving of 1.4 × 10⁻¹² for the general population and 6.4 × 10⁻¹¹ for susceptible populations. While the median predicted annual illnesses and deaths were zero, the 95^th percentile estimates reached up to 1,070 illnesses and 264 deaths, indicating the potential for significantly adverse public health outcomes in the worst-case scenarios.

The initial concentration of L. monocytogenes on cantaloupe at harvest, as well as post-packaging time–temperature conditions, were identified as the most influential factors affecting the presence of L. monocytogenes per serving and the risk of illness. The researchers suggest that targeting time and/or temperature conditions post-packaging may be an effective strategy for reducing the risk of listeriosis due to L. monocytogenes on cantaloupe.

Additionally, the initial L. monocytogenes levels at harvest and cross-contamination at the fresh-cut facility had the greatest impact on the prevalence of contaminated servings. 

QMRA Analysis, Validation, and Limitations

Scenario analysis revealed that increased clustering of L. monocytogenes contamination—where contamination is concentrated in fewer units—generally led to better public health outcomes than homogeneous contamination.

The QMRA model was validated against data from a major 2011 U.S. listeriosis outbreak, which resulted in 33 deaths and 147 hospitalizations, and retail-level L. monocytogenes concentration data. Observed contamination levels fell within the predicted range, supporting the model’s reliability. However, high-end percentile estimates (95^th and 99^th) showed wider uncertainty bounds, a common issue in rare-event modeling. Therefore, these estimates should be interpreted with caution.

Future research should focus on expanding field-level data, refining cross-contamination modeling, and improving simulation granularity to better capture serving-level risk variability.