MODEL ELEMENTS AS COMPONENTS OF A FOOD-SAFETY MANAGEMENT SYSTEM UNDER DYNAMIC GROWTH/DEATH GENERATING CONDITIONS

SESSION 8: Data Bases, software and decision support tools P. 81.- MODEL ELEMENTS AS COMPONENTS OF A FOOD-SAFETY MANAGEMENT SYSTEM UNDER DYNAMIC GROWTH/DEATH GENERATING CONDITIONS Pierluigi Polese1, Manuela Del Torre2, Mara Lucia Stecchini2 1 Polytechnic Department of Engineering and Architecture, University of Udine, Italy 2 Department of Agricultural, Food, Environmental and Animal Sciences, University of Udine, Italy Keywords: time-dependent probability parameter; non-thermal inactivation; dynamic conditions INTRODUCTION AND OBJETIVES Predictive food microbiology information needs, for increasing their manageability, to be implemented in decision-supporting tools, which can be practically used by a range of stakeholders to improve food safety. To be more effective in managing food safety, especially under dynamic growth/inactivation processing conditions, a combination of growth and devitalization models could be advantageous. The aim of this research was to develop a set of elements, including a time-dependent probability parameter (PΔLog) for dynamicallychanging growth/death conditions, to be included in a decision tool. MATERIAL AND METHODS The effect of temperature, pH, aw, liquid smoke and lactic acid on Listeria monocytogenes inactivation was experimentally evaluated. The dependence of the probability of devitalization was ascertained and the inactivation rate (KD) was determined by building a regression on the pathogen viability data (Log CFU/ml) versus time, relying on the independent action of factors, as in the gamma approach. RESULTS Our decision support approach is based on the gamma concept and contains a set of elements, which pivot around the growth/inactivation partitioning parameter (P) in its dynamic version (Polese et al., 2011, 2017). The concept underlying the already validated time-dependent probability parameter (Pt) for growth (Polese et al., 2017) was extended to the devitalization region, yielding a new element, i.e. PΔLog, expressed as the probability to achieve definite log increase/decrease within a specified time of storage. With the purpose of realistically predicting PΔLog, the variability of growth and death observations was taken into account (Aryani et al., 2016). Considering datasets of L. monocytogenes behaviours collected from published data (Lindqvist & Lindblad, 2009), the performance of KD was found to be satisfactory. The PΔLog option was successfully validated on ComBase data from experiments with food products characterized by different combinations of controlling factors. CONCLUSIONS One of the main drawbacks of using the commonly available predictive software tools is the fact that in many products with a long shelf-life and dynamic internal environment, prediction can result in both growth and inactivation (Iannetti et al., 2017). Providing a decision support approach with model elements that allow both the probability and quantification of the pathogen increase/decrease for a given formulation or storage condition, could be a step in the right direction to overcome these problems.