Extension of probability models of the risk of infections by human enteric viruses

This study presents a novel approach for obtaining reliable models and coefficients to estimate the probability of infection caused by common human enteric viruses. The aim is to provide guidance for public health policies in disease prevention and control, by reducing uncertainty and management costs in health risk assessments. Conventional dose-response (DR) models, based on the theory elaborated by Furumoto and Mickey [1], exhibit limitations stemming from the heterogeneity of individual host susceptibilities to infection resulting from ingesting aggregate viruses. Moreover, the scarcity of well-designed viral challenge experiments contributes to significant uncertainty in these DR models. To address these issues, we conducted a review of infection models used in health risk analysis, focusing on Norovirus (NoV) GI.1, pooled Enterovirus group (EV), Poliovirus 1/SM, and Echo-12 virus via contaminated water or food. Using a mechanistic approach, we reevaluated the known DR models and coefficients for the probability of individual host infection in the mentioned viruses based on dose-infection challenge experiments. Specifically, we sought to establish a relationship between the minimum infectious dose (ID) and the ID having a 50% probability of initiating host infection in the same challenge experiment. Furthermore, we developed a new formula to estimate the degree of aggregation of GI.1 NoV at the mean infectious dose. The proposed models, based on "exact" beta-Poisson DR models, effectively predicted infection probabilities from ingestion of both disaggregated and aggregate NoV GI.1. Through a numerical evaluation, we compared the results with the maximum likelihood estimation (MLE) probability obtained from a controlled challenge trial with the NoV GI.1 virus described in the literature, demonstrating the accuracy of our approach. By addressing the indetermination of the unmeasured degree of NoV aggregation in each single infectious dose, our models reduce overestimations and uncertainties in microbial risk assessments. This improvement enhances the management of health risks associated with enteric virus infections.