Use of risk assessment and predictive microbiology in regulatory science related to the scientific opinions of the EFSA BIOHAZ Panel

Abstract

EFSA's Panel on Biological Hazards (BIOHAZ Panel) deals with questions on biological hazards relating to food safety and food-borne diseases. This covers food-borne zoonoses, transmissible spongiform encephalopathies, antimicrobial resistance, food microbiology, food hygiene, animal-by products, and associated waste management issues. The scientific assessments are diverse and frequently the development of new methodological approaches is required to deal with a mandate. Among the many risk factors, product characteristics (pH, water activity etc.), time and temperature of processing and storage along the food supply chain are highly relevant for assessing the biological risks. Therefore, predictive microbiology becomes an essential element of the assessments. Uncertainty analysis is incorporated in all BIOHAZ scientific assessments, to meet the general requirement for transparency. Assessments should clearly and unambiguously state what sources of uncertainty have been identified and their impact on the conclusions of the assessment. Four recent BIOHAZ Scientific Opinions are presented to illustrate the use of predictive modelling and quantitative microbial risk assessment principles in regulatory science. The Scientific Opinion on the guidance on date marking and related food information, gives a general overview on the use of predictive microbiology for shelf-life assessment. The Scientific Opinion on the efficacy and safety of high-pressure processing of food provides an example of inactivation modelling and compliance with performance criteria. The Scientific Opinion on the use of the so-called ‘superchilling’ technique for the transport of fresh fishery products illustrates the combination of heat transfer and microbial growth modelling. Finally, the Scientific Opinion on the delayed post-mortem inspection in ungulates, shows how variability and uncertainty, were quantitatively embedded in assessing the probability of Salmonella detection on carcasses, via stochastic modelling and expert knowledge elicitation.