2018 ASHS Annual Conference

A modeling approach to using the knowledge (Knowns) gained from industry bacterial testing programs to bound and limit the unknowns and unknown unknows regarding microbial risk. The overall incidence rates for positive results in these databases are very low (1 in 5000 to 10000 tests depending on the product, season and other factors) limiting the utility of individual tests. However, large testing databases can be mined to yield information regarding the overall system under study and also yield information about what would not be consistent with the data and therefore, is not of concern. This mining approach can be bolstered by targeted intensive testing to yield additional information. Having this “is“ and “is not” information about the system affords an approach to assessing and developing practical mitigation strategies in an environment where pasteurization is probably out of reach. This type of analysis should enable processors and their customers to make thoughtful decisions about their risk mitigation programs.