Systematic Approach to Microbial Risk Assessment from Producers through Retailers

Identify hazards and develop models to allocate resources to minimize hazards in poultry, fruit and vegetable commodities. Develop a systematic microbial risk assessment model to address hazards in poultry, fruits and vegetables.

The overall project approach is to develop a systematic approach to integrating food safety, food technology, and food engineering together with production and transportation economics into reliable models for quantifying the risk of Salmonella and Campylobacter in poultry products as well as Salmonella and entro-hemorrhagic E. coli on fruit and vegetable products. Today, more and more food manufacturers are using new software modules to comply with the federally mandated Hazard Analysis and Critical Control Points (HACCP) monitoring plans. These HACCP modules can be merged with newly developing risk assessments models. The initial focus will be to collect data on poultry products. These findings will then be adapted for fruits and vegetables. The structure of a quantitative microbial risk assessment model has been developed. The first step in the development of microbial risk assessment models will be to collect information by searching data bases and publications, by consulting researchers and by contacting the food industry and regulatory agencies. The information will be classified into the following groups: (1) Experimental data and bacterial growth/survival kinetic models for Salmonella and Campylobacter in poultry, Salmonella and entro-hemorrhagic E. coli in fruit and vegetable products; (2) Physical and chemical properties of fruit, vegetable and poultry products; (3) Distribution of initial microbial loads, cross-contamination and recontamination; (4) Operating conditions from each unit operation that will affect the microbiological load; (5) Anti-microbial washes, pasteurization, thermal processing, irradiation or additional microbiological hurdles and (6) Available dose-response relationships and exposure models. After compiling and evaluating this data, gaps in the data will be filled in through gathering experts' opinions and estimating such factors as initial microbial loads, dose response or exposure assessment.

Produce, fruits and vegetables, was responsible for 26 food borne illness outbreaks during the 10 year period from 1990 - 2001. Each year Americans eat about 84 bananas, more than any other fresh raw fruit. The very young and the elderly, two groups with immature or compromised immune systems, eat the majority of the bananas. Although banana peel is not eaten it could harbor pathogens that could be transferred to the pulp during peeling and consuming. Bananas from local grocery stores were analyzed for their bacterial flora. No pathogens were isolated from the banana peels using direct plating methods. However, when bananas were inoculated with Staphylococcus aureus, Citrobacter freundii as a surrogate for Salmonella sp or with a non-pathogenic E. coli as a surrogate for E. coli O157:H7 was found that all of these organisms could survive at high levels on the peel of bananas and provide the possibility for cross-contamination of the edible pulp of the fruit. As a second stage, a simulation model using at Risk software was developed for the quantitative microbial risk assessment that described the probable level of a consumer becoming sick due to Staphylococcus aureus contamination of bananas. Inputs to the model were conditions that would contribute to the initial Staph contamination during harvesting, through growth of this pathogen at processing, shipping, degreening, retailing and consumption. Previous research had shown that as many as 44% of food workers were positive for the pathogenic strains of Staph. However, significant growth of Staph is required before toxin production is commenced. Probability distributions were used to characterize the incidence and banana-bacteria-consumer events that described the initial contamination, microbial survival/growth/death and cross-contamination. Data from values published in the literature, Staph growth experiments and computer simulations were used for model input data.

The newly developed model predicted incidence of staphylococcal food poisoning associated with contaminated bananas was as being low, 0.02% or about 2 in 10,000 bananas under the conditions described in this study. Our results indicate that bananas should be relatively low risk as a hazard for Staphylococcus toxin. Temperature-time abuse during retailing and cross-contamination during food handling and preparations are the important factors contributing to risk of illness. The results indicated that the developed model has its significance in predicting the microbial populations and risk in banana production and processing.

Research Objectives: 1) Develop quantitative risk assessment models for microbial hazards in poultry, fruits and vegetables. 2) Construct a computer simulation program using Analytica software for the risk assessment models. 3) Determine distribution functions of initial microbial loads, cross-contamination, and recontamination.

Education Objectives: 1) Examine how expenditures for HACCP implementation have impacted risk of food born illness, product liability rates and claims. 2) Use risk analysis to estimate which investments in food safety have paid the greatest dividends. 3) Leverage on the investigators' expertise in training students and corporate personnel on a foundation statistical program, JMP that can be also applied to quality control and quality assurance. 4) Develop a series of CD-ROM based tutorials that will provide food processors with a background in newly developed statistics programs, risk assessment, management and communication and technical expertise in delivering web based instruction.

Extension Objectives: 1) Communicate how risk analysis is used to identify and minimize risks at the pre-harvest and agribusiness sectors, focusing on actual liabilities that food processors face regarding food safety. 2) Set up approximately twenty extension programs with food processors in Oklahoma, Georgia and Arkansas to teach about using risk assessment. 3) Develop CD-ROM self-tutorials specially designed for individual company workshops.