Roles of Vegetable Surface Properties and Sanitizer Type on Annual Disease Burden of Rotavirus Illness by Consumption of Rotavirus-Contaminated Fresh Vegetables: A Quantitative Microbial Risk Assessment.
Miyu FuzawaRebecca Lee SmithKang-Mo KuJoanna L ShislerHao FengJohn A JuvikThanh Huong NguyenPublished in: Risk analysis : an official publication of the Society for Risk Analysis (2019)
Enteric viruses are often detected in water used for crop irrigation. One concern is foodborne viral disease via the consumption of fresh produce irrigated with virus-contaminated water. Although the food industry routinely uses chemical sanitizers to disinfect post-harvest fresh produce, it remains unknown how sanitizer and fresh produce properties affect the risk of viral illness through fresh produce consumption. A quantitative microbial risk assessment model was conducted to estimate (i) the health risks associated with consumption of rotavirus (RV)-contaminated fresh produce with different surface properties (endive and kale) and (ii) how risks changed when using peracetic acid (PAA) or a surfactant-based sanitizer. The modeling results showed that the annual disease burden depended on the combination of sanitizer and vegetable type when vegetables were irrigated with RV-contaminated water. Global sensitivity analyses revealed that the most influential factors in the disease burden were RV concentration in irrigation water and postharvest disinfection efficacy. A postharvest disinfection efficacy of higher than 99% (2-log10 ) was needed to decrease the disease burden below the World Health Organization (WHO) threshold, even in scenarios with low RV concentrations in irrigation water (i.e., river water). All scenarios tested here with at least 99.9% (3-log10 ) disinfection efficacy had a disease burden lower than the WHO threshold, except for the endive treated with PAA. The disinfection efficacy for the endive treated with PAA was only about 80%, leading to a disease burden 100 times higher than the WHO threshold. These findings should be considered and incorporated into future models for estimating foodborne viral illness risks.