M21 - Poster Platform: Food/Water Safety Risk Analysis I: Microbial and Chemical RisksSalon E |
| Chair(s): Mark Powell |
|
M21.1 Relative Absence of Known Streptogramin Resistance Determinants among Enterococcus faecium from the U.S. Poultry Production Environment. Hayes, J.R.*, English, L.L., Wagner, D.D., Carr, L.E., and Joseph, S.W.; University of Maryland and U.S. FDA, Center for Veterinary Medicine jhayes@cvm.fda.gov Abstract: The impact of agricultural use of antimicrobials on the future efficacy of therapeutic drugs in human medicine is a growing public concern. Quinupristin-dalfopristin (Q-D) has been approved to treat human disease due to vancomycin-resistant Enterococcus faecium and is related to virginiamycin, a streptogramin that has long been used in U.S. agriculture poultry production. The magnitude of transferable resistance elements selected for by the agricultural use of streptogramin antimicrobials is largely unquantified within the poultry production environment of the U.S. In order to characterize the population of streptogramin-resistant E. faecium that may be selected for by the use of antimicrobials in food animal production, 127 isolates of E. faecium were recovered from litter and transport containers from 55 poultry farms along the Eastern seaboard. These isolates were recovered without selection for streptogramin resistance and subjected to ribotyping to eliminate clonal bias. Colony PCR screening for the streptogramin resistance determinants erm(A), erm(B), msr(C), vgb, vat(D), and vat(E) was performed. Susceptibility testing revealed that 63% of isolated E. faecium were resistant to quinupristin-dalfopristin. However, resistance to streptogramin antimicrobials was largely unaccounted for by PCR screening for specific resistance determinants. Resistance determinants erm(A) and erm(B) were observed among 7.5% and 5% of resistant isolates while the efflux gene msr(C), the streptogramin B hydolase vgb, or the streptogramin A acetyltransferases genes vat(D) and vat(E) were not detected among resistant E. faecium. These results indicate that streptogramin resistance is widespread among E. faecium from the poultry production environment but the mechanisms of resistance within this population remain largely uncharacterized. |
|
M21.2 Past and Future Human Health Risks and Benefits from Animal Antimicrobials. Cox, Tony, Popken, Douglas; Cox Associates tony@cox-associates.com Abstract: Recent qualitative risk assessments and mathematical models of antimicrobial resistance (APUA, 2002; Smith et al., 2003) have warned of potential future human health dangers from using similar or identical antimicrobial drugs in both food animals and human medicine. While historical data suggests that human health risks from key animal antimicrobials, such as virginiamycin (VM), have been low, there is widespread concern that "resistance epidemics" could arise in future that will not be easy to contain or reverse once started. How reassuring is the past about the future? Are current efforts to restrict or eliminate animal antimicrobials justified? This paper applies quantitative risk assessment (QRA) methods to help find out, using human health risks from VM and the nearly identical human antimicrobial quinupristin-dalfopristin (QD) as a case study. Bayesian Monte Carlo uncertainty analysis of a dynamic simulation model of the emergence of resistance in human populations allows past data to inform mathematical analysis of possible future resistance rates. A full QRA with realistic parameter estimates shows that the probability that continued use of VM will reduce the time until an epidemic of human resistance develops is less than 0.001. The time until an “outbreak” (at least two patients in a hospital being colonized at the same time) is not appreciably affected by concurrent use of VM in animals. Human health risks due to animal use of VM are sensitive to the prescription rate of QD. For realistic prescription rates, the quantitative risks are less than 1 x 10-6 even for members of the most-threatened (ICU patient) population, while societal risks are less than 1 excess death expected per decade for the whole US population. Such quantitative estimates provide a potentially valuable complement to more qualitative assessments that indicate the theoretical possibility of future resistance epidemics without quantifying their probabilities. |
|
M21.3 Quantitative and Qualitative Prediction of Human Health Risks from Emerging Resistance to Animal Antimicrobials. Cox, Tony, Popken, Douglas *; Cox Associates tony@cox-associates.com Abstract: A deterrent to wider use of quantitative risk assessment and “farm-to-fork” simulation models in antimicrobial risk assessment is the expense and effort of assembling the required data elements into useful predictive models. This paper presents a computer-aided risk assessment software tool that combines knowledge-based methods (e.g., rules for predicting when bacteria are likely to develop cross-resistance to drugs in the same family) with quantitative dynamic simulation modeling to quickly approximate the human health risks from different antimicrobial drugs used in food animals. The simulation model sums estimated human health impacts over multiple bacterial species, food animal species, meat products, and human sub-populations. It considers not only historical data, but also simulated emergence of resistance in the future under the selection pressures created by antimicrobials used as growth promoters and/or therapeutic agents in food animals. The model has been applied to several classes of drugs, including macrolides, ionophores, and flouroquinolones, to quickly detect and estimate the relative human health risks (measured in expected quality-adjusted life-years lost per year and expected excess treatment failures and resulting mortalities and morbidities per year) for different “drug-bug” populations. The quantitative simulation model has also been used to generate input-output risk relations that, in turn, have been approximated as a set of relatively simple if-then rules using a machine-learning algorithms for inducing rules from data. This rule-based approach allows for relatively rapid, inexpensive ranking of antimicrobial risks based on available data about drug use, animal exposures, human consumption patterns, human treatment data, and cross-resistance rates within drug families. The predictions of the qualitative and quantitative risk assessment approaches are contrasted for several well-known animal antimicrobials. |
|
M21.5 Development of Guidance Regarding Application of Sensitivity Analysis Methods to Food Safety Risk Models. H.C. Frey*, Mokhtari, A., Danish, T.; North Carolina State University frey@eos.ncsu.edu Abstract: The objective of this work is to provide guidance to practitioners regarding sensitivity analysis applied to food safety risk process models. Such models are challenging because they typically include: (1) nonlinearities; (2) thresholds; (3) continuous, discrete, and categorical inputs; and (4) two-dimensional simulation of variability and uncertainty. This project included a review of existing methods, an expert and practitioner workshop regarding sensitivity analysis methods, a detailed series of quantitative case studies of multiple sensitivity analysis methods applied to two models, a practitioner workshop regarding review of the case studies and of draft guidance, and development of a guidance document. The two models used as testbeds include Listeria monocytogenes among selected categories of ready-to-eat foods and E. coli O157:H7 in beef in the U.S. Methods evaluated include local sensitivity analysis (e.g., nominal range sensitivity, differential sensitivity), regression and related methods (e.g., sample and rank correlations, linear regression, analysis of variance, and regression trees), and graphical methods. The methods were compared with regard to multiple criteria, including applicability, robustness of results, ability to rank and discriminate among inputs, and others. The guidance emerging from this research includes: (1) why and when to perform sensitivity analysis; (2) preparation of existing or new models to facilitate sensitivity analysis; (3) defining objectives and case study scenarios; (4) selection of methods; (5) general principles for application of methods; and (6) presentation and interpretation of results. This presentation will focus on the key lessons learned as well as recommendations pertaining to sensitivity analysis applied to food safety risk process models. |
|
M21.6 Quantitative exposure assessment of verocytotoxigenic Escherichia coli O157 in British meat products . Jones, RD*, Kelly, L, Gettinby, G; Gale, P.; Wooldridge; M.; Veterinary Laboratories Agency, University of Strathclyde, Unversity of Strathclyde, WRc-NSF Ltd and Veterinary Laboratories Agency r.d.jones@vla.defra.gov.uk Abstract: Since the emergence of verocytotogenic Escherichia coli (VTEC) O157 in the 1970’s, the United Kingdom, particularly Scotland, has witnessed large scale outbreaks of infection in which meat products have been considered the reservoir or vehicle of infection. To quantify the amount of VTEC O157 at the point of consumption within specific meat products derived from British livestock, namely, cattle, sheep, pigs, and chickens, an exposure assessment has been developed. Within the assessment, the farm module, for both cattle and sheep, is described using a stochastic transmission model in which intervention measures can be investigated. As part of the model, both the prevalence of externally contaminated and gut colonized animals are considered. The outputs from these epidemiological models serve as inputs to the transport to consumption modules in which the changes in prevalence and concentration of VTEC O157 along the food pathway are modeled based on the Modular Process Risk Model (MPRM) approach. Using this approach, the dynamics of the organism along the food pathway are described using a combination of six microbiological and manufacturing processes: growth, inactivation, partitioning, mixing, removal and cross contamination. Description of these processes accounts for the variation resulting from, for example, the distribution of VTEC O157 in animal populations, environmental factors such as temperature, and microbiological characteristics. It is aimed that incorporation of this variation will enable investigation of, for example, the differences in the incident rates between Scotland and England, thereby providing insight into a factor of current concern. Results from both the transmission model and the exposure assessment will be presented. |
|
M21.7 Data for Campylobacteriosis Risk Assessment Associated with US Poultry. Coleman, M.E.* , Golden, N.J., Kadry, A.M.; USDA/FSIS/OPHS peg.coleman@fsis.usda.gov Abstract: Published manuscripts and reports on modeling campylobacteriosis risks associated with broiler chickens were extended as part of a feasibility study for risks to US consumers. Additional sources of data were available for exposure assessment, including the Continuing Survey of Food Intake by Individuals (CSFII) on U.S. consumption and other sources for prevalence and levels of Campylobacter in poultry. The CSFII includes data on ~21,000 US survey participants in 2- or 3-day food diary surveys that can be used to estimate serving sizes and frequencies of consumption of chicken and turkey products. Predictive microbiology data also exist indicating no growth of Campylobacter at temperatures below approximately 30ºC, and modeling of decline rather than growth will be considered for typical processing and storage conditions. Data are also available on pathogenicity and virulence from in vitro and in vivo studies relating to dose-response assessment, in addition to human clinical data. Other scientists have noted the unexpected trends of decreasing campylobacteriosis with increasing dose and questioned the validity of pooling the data from the two strains administered to human volunteers. The weight of evidence suggests explicit modeling of strain effects is needed. Overall, the additional data sources considered in the feasibility study will assist the agency in developing a more formal weight of evidence approach for future campylobacteriosis risk assessment. |
|
M21.8 Methylmercury Toxicity: on Fish, Children, Pharmacokinetics and Toxicodynamics. Byczkowski, J.Z.; Consultant, Fairborn, Ohio januszb@aol.com Abstract: The current information on pharmacokinetic and toxicodynamic aspects of methylmercury toxicity, which may be helpful in understanding the risk of exposure and potential adverse health outcomes of its ingestion with food, are reviewed. It is estimated that for about 5 percent of the U.S. general population of women there may be some risk of exposing their unborn and/or nursing infants to methylmercury levels that potentially may cause some adverse effects in the central nervous system. This estimate is based on the most conservative reference dose of 0.1 mg MeHg/kg/day which bears a substantial uncertainty and is skewed mostly in the health-protective direction. It is concluded that any potential adverse effects of exposure to methylmercury in seafood must be carefully balanced against the obvious nutritional and health benefits of seafood and breast-feeding. |
|
M21.9 EXPOSURE ASSESSMENT OF LISTERIA MONOCYTOGENES IN COOKED HAM USING THE MPRM METHOD.. Rodríguez, F.P.*, Gimeno-Garcia, R.M.G., Cosano, G.Z., Díaz, A.V., Jimenez, E.C. ; University of Cordoba, Spain b42perof@uco.es Abstract: This study focuses on using MPRM method to assess the exposure of Listeria monocytogenes in cooked ham. Listeriosis outbreaks have increased over the last two years in Spain (10-44 cases in 1999-2002, respectively). Because of the ability of L. monocytogenes to survive at refrigeration temperatures, the well known cold chain during food production doesn´t ensure food safety. We considered cooked ham as a susceptible product to be contaminated by L. monocytogenes during handling in industries and retail points. So only production steps occuring after cooking of ham are taken into account, such as handling in industry, package, handling in retail points, storage, consumption patterns and domestic habits. The reason why we chose the steps after cooking is that L. monocytogenes is inactivated at cooking temperatures.A general framework, Modular Process Risk Model (MPRMs) (Nauta 2001) is introduced as a systematical procedure to develop exposure assessment step within Quantitative Microbilogical Risk Assessment (QMRA). MPRMs method identifies all processing steps as one of six basic processes (modules) in any food pathway: growth and inactivation (microbial processes); mixing, partitioning, removal and cross contamination (product handling processes). We found that the modules included in this study are microbial growth, partiotioning and cross contamination. Several predictive models were used to explain microbial growth during storage; these models include factors such as pH, temperature, nitrate, sodium chloride and spoilage flora. Partitioning module was analized through some probability distributions (Nauta, 2001). Finally, cross contamination was mainly described by transfer ratios between surfaces (steel, hands, spigot,...). The yeilded results demonstrated that exposure levels are greatly influenced by survival and transfer rates during pathogen transmission. |
|
M21.10 Microbial Risk Assessment: Lessons learned from the National Academy of Sciences Committee on Review of the Use of Scientific Criteria and Performance Standards for Safe Food. Donahue, D. W., Riviere, J. E., Schaffner, D. W.; 1) University of Maine, 2)N. C. State University, 3)Rutgers University ddonahue@umche.maine.edu Abstract: Microbial risk assessment has become an increasingly important method used to improve public health and address emerging foodborne disease issues because it offers a systematic approach to assessing the impact of microorganisms in the food chain. The use of quantitative and qualitative risk assessments (QRAs) for microbiological food safety issues has developed from a rich history of use in chemical and environmental toxicology. Chemical risk assessment is a relatively mature field compared to that of microbial risk assessment, and is due in part to the requirement for drugs and chemicals to be approved or registered by either the FDA or EPA prior to potential human exposure. Despite these differences in maturity, the overall chemical risk assessment paradigm has a number of similarities to microbial risk assessment as it is currently practiced. The lessons learned from chemical QRA can be applied to the new challenges of developing microbial QRAs. This presentation will summarize lessons learned as a result of the authors’ participation in the recent National Academy of Sciences committee on scientific criteria and performance standards for safe food. Data gaps in the four major components of risk assessment will be identified with respect to microbial QRA in foods. Several methods of handling the data gaps will be explored. The argument for linking chemical and microbial risk assessments for foods to the concept of food safety objectives will be presented. The linkage between risk assessments through food safety objectives to public health outcomes will be made. |