Probabilistic risk model of norovirus transmission during handling and preparation of fresh produce in school foodservice operations

Abstract

Human noroviruses (NoV) are recognized worldwide as important pathogens and the primary cause of foodborne disease outbreaks from contaminated food in the U.S. They are often transmitted by infected food handlers manipulating foods during preparation, such as fresh fruits and vegetables. This paper provides a study to model the transfer of NoV between food handlers and vegetables during salad preparation in school food services based on direct observation data. Three transfer pathways were modeled by considering different initial contamination sources (environment, handlers and contaminated produce). The probability of infection by NoV was also estimated based on the NoV levels at consumption obtained from each simulated transfer pathway. A scenario analysis ranging a wide concentration from 10² to 10⁷ NoV infective particles was performed to represent different levels of NoV in the initial contamination sources. In addition, a sensitivity analysis was applied to identify the most important model inputs and determine the safest handling practices to be implemented in school food service operations. The pathway describing transfer from contaminated surfaces or handlers to foods indicated that initial levels of ≤10⁴ NoV particles/fomite resulted in <0.5% cases per serving of NoV infection. When initial levels were higher, % cases of NoV infection was estimated to be ca. 3%. This rise in % cases of infection was linked to higher doses (5% serving with ≥15 NoV particles/serving) and prevalence levels at consumption (>0.2). In the pathway modeling cross contamination from contaminated vegetables to non-contaminated vegetables, all scenarios could lead to infected individuals, although number of cases of infection were lower (<1.3%), despite concentration levels were higher. On the contrary, for this pathway, prevalence was 2-fold lower than that observed in the pathways describing transfer from contaminated surfaces and hands. Based on the sensitivity analysis, NoV transfers to fresh produce may be minimized by improving hand washing, and therefore effective training programs need to be carried out specifically addressing hand washing. Moreover, the produce's washing step showed to be an effective control measure, depending on the desinfectant efficacy, by reducing % cases of NoV infection from 6 to 1%. The model in this study might be used, in the future, to evaluate the impact on the risk associated with NoV transmission of specific and effective training programs, aimed at food chain operators.

Keywords: Cross-contamination; Enteric viruses; Lettuce; Microbial risk assessment; Stochastic modeling.