Importance of Predictive Microbiology for risk minimization in food production processes - 2. Shortcomings and benefits

Although all Predictive Microbiology (PM) models are simplifications of the biological mechanisms and the actually available models still have their l imitations, comparisons with independent data from literature indicate that predictions of most models are in the worst case fail-safe and their syste matic errors do not exceed those of inoculated pack experiments. Once a mod el has been validated for a type of food, it can be applied at all stages o f food production and distribution. PM models are already used to conduct H ACCP studies and are powerful tools for microbiological risk assessment, in particular. Under both aspects, there are various applications for PM: det ermination of necessary time-temperature combinations at heating processes, estimating the risk of pathogen growth during planned storage time and con ditions, examination of new formulations for potential microbiological haza rds. Traditionally, these tasks were done by using microbiological challenge tes ting (MCT, inoculated pack experiments). But these experiments are expensiv e and time consuming. Moreover, their results are only valid for the produc t in test and the conditions of its processing and storage. If changes are planned or occur occasionally, new MCT's have to be conducted under changed conditions. PM models are most effectively used at the stage of product development. As they allow a last first estimation about microorganisms behaviour, their a pplication enables food microbiologists to recognize without loss of time b y wailing for MCT results, whether the product has to be modified in formul ation or process, to avoid unacceptable levels of risk for consumers.