Risk Modelling - an Explanation of Friday 13th Syndrome (Failure) in Well-operated Continuous Sterilisation Plant

There is a sense in which we are all familiar with Friday 13th syndrome. We are intuitively familiar with the practical notion that no matter how good the design and operation of plant there will be an occasional bad outcome or rare event. Often there is too small a data set for analysis, especially if these rare events are simply put down to human error. The important insight offered by Risk Modelling to explain rare events is applied to sterilisation. Sterilisation is the most numerous process operation worldwide. It is essential to the preparation and processing of a wide range of bulk media including fermentation media, food, plant and equipment and a wide range of pharmaceuticals such as parenterals (examples include cholera and flu shots). In a comparative example the widely used single value assessment (SVA) of risk for non-sterility of UHT processed milk packs is compared with that from a Risk model. The effect of distributions of values in each of the UHT process parameters for sterilisation is highlighted through a Monte Carlo assessment (MCA) with a distribution of practical values for the concentration of contaminant spores (Bacillus stearothermophilus and B thermodurans), thermal resistance of the spores, heating temperature and the residence time of the milk in the steriliser. Results offer a number of practical surprises - they support our intuitive notion regarding Friday 13th syndrome and underscore that in operation of UHT plant we should get a higher proportion of non-sterile milk packs (failures) than is predicted by the SVA. The exciting prospect of applying Risk modelling lessons and insights to a range of processes and in the assembly of bio-statistics for biological and microbiological processes is briefly discussed.