Using an approach that incorporates both laboratory-based studies and mathematical modeling, the purpose of this work is to develop a quantitative risk assessment model to describe the risk of foodborne viral gastroenteritis and hepatitis A associated with poor personal hygiene practices of infected food handlers. The objectives are as follows:
Collect data on surface persistence and disinfection for hepatitis A virus (HAV), Norwalk virus [NV, a representative genogroup I norovirus (NoV)], and Snow Mountain virus (SMV, a representative genogroup II NoV) as evaluated using real-time quantitative RT-PCR and supplemented with infectivity assays using HAV strain HM175 and mouse norovirus (MNV-1); <LI> Collect data on hand persistence and hand-washing/disinfection efficacy for HAV, NV, and SMV using real-time quantitative RT-PCR and supplemented with infectivity assay using HAV strain HM175; <LI> Collect data on virus transfer (between hands, surfaces, and foods) for HAV, NV, and SMV using real-time quantitative RT-PCR and supplemented with infectivity assay using HAV strain HM175; <LI> Evaluate virus titers from infected individuals over the course of infection for NV and SMV using real-time quantitative RT-PCR; <LI> Use the data from the four previous objectives, including the incorporation of dose-response data from human challenge studies, to estimate the risk of foodborne viral gastroenteritis and hepatitis A associated with poor personal hygiene practices of infected food handlers.
NON-TECHNICAL SUMMARY: Recent estimates indicate that the human gastrointestinal viruses are the most common cause of foodborne disease. These viruses are frequently transmitted by infected food handlers. Unfortunately, little is known about the risk of human disease and the efficacy of proposed methods to control virus transmission via contaminated foods. This work will provide significant understanding of persistence and transfer of gastrointestinal viruses in the food handling environment, their risk of infecting humans, and ways that such risk can be reduced.
APPROACH: Working with four viruses [Norwalk virus (NV), Snow Mountain virus (SMV), murine norovirus (MNV-1), and hepatitis A virus (HAV)], the project has both a lab-based component aimed at gathering data for inclusion in the risk model, and a mathematical risk modeling component. Virus inactivation, persistence, and transfer will be measured in laboratory-based studies using methods of mammalian cell culture infectivity assay (which quantifies infectious virus) and real-time quantitative RT-PCR (which quantifies viral genome copies). These data, along with additional data collected from the literature and by expert elicitation, will be incorporated into the exposure model. Virus titers from experimentally challenged individuals will be measured over the course of infection; these data will be combined with previously collected dose-response data to develop hazard characterization models. The modeling will have a probabilistic framework with inputs expressed in stochastic terms (i.e., as probability density/mass functions). Initially, the modeling will be done using MatLab software with further refinements planned to improve accessiblity. The propagation of variability and uncertainty will be done using Monte Carlo simulation. Upon completion of the risk characterization, what-if scenario analysis will be applied to evaluate potential mitigation strategies.
PROGRESS: 2006/09 TO 2009/02 <BR>
OUTPUTS: Combined laboratory-based data collection and mathematical modeling were used to characterize the risk of foodborne viral disease associated with poor personal hygiene of infected food handlers. In the lab portion, we examined the persistence and transfer of human norovirus (HuNoV) strains and a surrogate virus (murine norovirus, MNV-1) on and between representative food contact surfaces (stainless steel, ceramic and Formica), foods (lettuce and deli meat), and human fingerpads. Quantification of virus persistence and transfer was accomplished using quantitative real-time (RT-qPCR) and/or mammalian cell culture infectivity assay. Human NoV were stable on food contact surfaces with an average reduction of 1.5-2.3 log over 45 days of exposure to ambient conditions; there were no statistically significant differences between surfaces. On inoculated lettuce, HuNoV titers decreased from 0.8-1.8 log over a 2 week storage period at refrigeration or room temperature. On human fingerpads, the HuNoV were equally persistent with little drop in viral RNA concentration 120 min after inoculation. Only short-term persistence (<7 days) was observed when purified viral RNA was placed on surfaces or lettuce, suggesting that some of the quantifiable RNA detected after long-term exposure to ambient conditions was associated with infectious virus. The persistence of MNV-1 and purified MNV-1 RNA was substantially different from that observed for the HuNoV strains. Statistically significant differences in transfer efficiency were observed over time (decreases), with pressure (increases), and by food or virus. For example, transfer efficiency of HuNoV from ceramic to lettuce ranged from 0-26% while for MNV-1, it ranged from 1-11%. Transfer of viruses from surfaces to deli meat was much more efficient (up to 95%). A probabilistic exposure model of the transmission dynamics of HuNoV in the retail food preparation environment was developed. Key inputs included degree of fecal shedding, hand hygiene behaviors, efficacy of virus removal and/or inactivation, and transferability of virus between surfaces. The model has a temporal dimension allowing contamination to be estimated as a function of time over the simulation period. Sensitivity and what-if scenario analyses were applied to identify the most important model inputs and evaluate potential mitigation strategies. The key inputs affecting estimates of the number of infectious viruses present in contaminated food servings, given the current model structure and assumptions, were as follows: mass of feces on hands, concentration of virus in feces, number of bathroom visits, degree of gloving compliance, hand-washing efficiency, and hand-washing compliance. The model suggests that gloving and hand-washing compliance are most effective in controlling contamination of food products when practiced simultaneously. Moreover, the bathroom environment was identified as a major reservoir of HuNoV, even in the absence of an infected individual on site. This unique mathematical model will facilitate comparison of potential mitigations aimed at reducing the transmission of foodborne viral infections. <BR> PARTICIPANTS: Dr. Blanca Escudero-Abarca (NCSU) and Dr. Pengbo Liu (Emoery), both post-doctoral research associates, conducted most of lab-based research. A small portion the laboratory component was completed by Dr. Efstathia Papfragkou as part of her Ph.D. dissertation work. Two undergraduate students (Ms. Alexia Taylor and Ms. Katie Ginsel) provided support for laboratory experiments, providing them an opportunity to train in mammalian cell culture and molecular techniques. Dr. Amir Mokhtari, serving as a post-doctoral research associate, completed the exposure modeling. <BR> TARGET AUDIENCES: Although the research community is the true target audience, the end-products provided by this project are relevant to federal and state regulatory agencies interested in controlling the transmission of foodborne viruses and in microbial risk assessment. The risk model is also relevant to the food industry as they, too, seek to control transmission of these agents. <BR> <BR>
IMPACT: 2006/09 TO 2009/02<BR>
Recent estimates indicate that the HuNoV are responsible for a large proportion of the world's foodborne disease burden. Unfortunately, the risk of human disease associated with these viruses is poorly characterized, in large part because of the absence of data upon which model parameters may be based. The purpose of this work has been to develop a quantitative exposure assessment model to describe the risk of foodborne viral gastroenteritis associated with poor personal hygiene practices of infected food handlers. In so doing, it was necessary to collect data on the environmental persistence and transfer of these viruses. Significant changes in knowledge, actions, and/or conditions which have occurred as a consequence of this project include the following: --Confirmation that the HuNoV are highly stable under ambient conditions, and can persist (with limited reduction in titer) on surfaces for up to 45 days, on foods for up to 14 days, and on human skin. --Demonstration of the ease of transfer of the HuNoV between surfaces (including foods and hands) with efficiency ranging from 0-95%. --Production of data which supports the observation that MNV-1 may not always be the most appropriate surrogate for studying the environmental behavior of the HuNoV. --Demonstration that ethanol-based hand sanitizers are relatively ineffective in removal or inactivation of HuNoV from human fingers, emphasizing the need to development of hand (and surface) sanitizing agents with specific efficacy against this group of viruses. --Laboratory and risk assessment results which support the epidemiological evidence that food preparation surfaces and human hands can act as vehicles for human NoV transmission long after an initial contamination event has occurred. --The exposure model is unique, the first of its kind to apply quantitative risk assessment to the foodborne transmission of the HuNoV. Further, it provides a means by which regulatory agencies and the industry can evaluate the efficacy of potential mitigation strategies which might be considered to control the foodborne transmission of the HuNoV in the future.