BIKE: Dietary Exposure Model for Foodborne Microbiological and Chemical Hazards.

Jukka Ranta,Pirkko Tuominen,Johanna Suomi,Antti Mikkelä

doi:10.3390/foods10112520

Abstract

BIKE is a Bayesian dietary exposure assessment model for microbiological and chemical hazards. A graphical user interface was developed for running the model and inspecting the results. It is based on connected Bayesian hierarchical models, utilizing OpenBUGS and R in tandem. According to occurrence and consumption data given as inputs, a specific BUGS code is automatically written for running the Bayesian model in the background. The user interface is based on shiny app. Chronic and acute exposures are estimated for chemical and microbiological hazards, respectively. Uncertainty and variability in exposures are visualized, and a few optional model structures can be used. Simulated synthetic data are provided with BIKE for an example, resembling real occurrence and consumption data. BIKE is open source and available from github.

Highlights

Exposure assessment is one of the four parts in risk assessments, the other parts being hazard identification, hazard characterization and risk characterization
The workflow of BIKE is simple: once the input data are defined in a correct format, the modeling and computations in BIKE are automatic
Data for hazard concentrations and the consumption of specific foods needs to be stored in Excel formats, which are to be converted to text-files beforehand, to be read by BIKE

Summary

Introduction

Exposure assessment is one of the four parts in risk assessments, the other parts being hazard identification, hazard characterization and risk characterization. Foodborne exposure assessment relies on both occurrence data and food consumption data While the former provides information on the prevalence and level of contamination in foods or food ingredients, the latter provides information on how often and in what amounts the foods are consumed. For non-Bayesian approaches it is common that parameter estimation is broken into separate steps with unrelated estimation methods and possibly combined with assumptions for ‘nuisance parameters’ that are part of the model but not the primary target of inference. All this may lead to incoherent quantification of the truly multidimensional uncertainty. The full potential of multidimensional Bayesian hierarchical modeling in open-source format remains underexplored and the Bayesian methods are still often only mentioned as a possible refined approach [7,9], and left to be explored in further studies

Methods

Results

Conclusion