An evaluation of parsimony for microbial risk assessment models

Abstract

Microbial risk assessment (MRA) is a process that evaluates the likelihood of adverse human health effects following exposure to a medium in which pathogens are present. Several different classes of models are available to quantitatively characterize risks to human health from exposure to pathogens. Herein, we consider the question of parsimony for specific realizations of representative static and dynamic MRA models and identify conditions under which the more complex dynamic model provides sufficient additional insight to justify the added modeling complexity. To address this question, a standard static individual-level risk model is compared to a deterministic dynamic population-level model that explicitly includes secondary transmission and immunity processes. Exposure parameters are based on a scenario defined by human exposure to pathogens in reclaimed water. A sensitivity analysis is implemented to identify conditions under which static and dynamic models yield substantially different results. Under low risk conditions, defined by a combination of exposure levels and infectivity of the pathogen, the simpler static model provides satisfactory risk estimates. The approach presented here provides a basis for model selection for a broad range of MRA applications. Copyright © 2007 John Wiley & Sons, Ltd.