Pathogen exposure to multiple hurdles could result in variation in the number of survivors, which needs to be
carefully considered using appropriate regression models for dealing with survivor dispersion. The aim of this
study was to evaluate the impact of the hurdles on the random component of the measured variation and on its
unexplained part (over or under-dispersion) representing the departure from randomness, i.e. non-randomness, in survivors of a multi-strain mixture of L. monocytogenes. The pathogen inactivation curves were fitted to the Weibull model within the Conway-Maxwell-Poisson process. In all the 20 hurdle combinations, the surviving cells, whether they showed an upward curvature or linear kinetics, displayed the randomness revealed by the degree of dispersion of the inactivation parameters (-b and p). In 15 combinations, a significant dispersion coefficient (c0), which reflected the non–random component of variation was evident, denoting either over-dispersion (c0 >0 in 13 combinations) or under-dispersion (c0 < 0 in 2 combinations). The observed dependence of the under- and over-dispersion conditions on the inactivation rate was confirmed by a Monte Carlo simulation based on the inactivation parameter -b. Including both randomness and non-randomness provides a more accurate estimation of survivors, which certainly impacts on intervention practices.
