With the emergence of quantitative risk assessment models in the field of food safety, there has also been a growing recognition of the need for sensitivity analysis of such models. Key questions that must be addressed in performing sensitivity analysis with food safety risk models include the following: What are the key criteria for sensitivity analysis methods applied to food safety risk assessment?; What sensitivity analysis methods are most promising for application to food safety and risk assessment?; and What are the key needs for implementation and demonstration of such methods?
Microbial food safety process risk (MFSPR) models are examples in which the process of sensitivity analysis is challenged by typical characteristics of those models such as non-linearity, thresholds, interactions, use of both continuous and categorical inputs, and two-dimensional simulation of variability and uncertainty.
The three main purposes of this dissertation are to: (1) evaluate sensitivity analysis methods; (2) present examples of how sensitivity analysis can be applied to MFSPR models and how the results can be presented and interpreted; and (3) provide guidance to risk analysis practitioners regarding the application of sensitivity analysis, particularly with regard to MFSPR models as well as other risk models with similar characteristics. This study included a review of existing methods, and a detailed series of quantitative case studies of multiple sensitivity analysis methods applied to an E. coli O157:H7 food safety process risk model in ground beef. Methods evaluated include Pearson and Spearman correlation analyses, sample and rank regression analyses, sample and rank stepwise regression analyses, analysis of variance, and classification and regression trees.
The guidance is intended to assist exposure and risk modeling practitioners to make effective use of sensitivity analysis methods and the insights they provide. The guidance provides insight regarding: (1) why and when to perform sensitivity analysis; (2) preparation of existing or new models to facilitate sensitivity analysis; (3) defining objectives and case study scenarios; (4) selection of methods; (5) general principles for application of methods; and (6) presentation and interpretation of results.
