Comparison between the Exponential and Weibull-gamma Dose-response Model Approaches to Quantitative Microbial Risk Assessment

Introduction: Quantitative microbial risk assessment (QMRA) can be to provide scientific advice to the risk managers who will use the information to decide upon the risk management option(s) that will be implemented to achieve the desired level of consumer protection from microbial hazards. According to the Codex Alimentarius Commission, a QMRA should include four steps, i.e., hazard identification, exposure assessment, hazard characterization, and risk characterization. In these above successive steps, a crucial aspect of QMRA is the assessment of dose-response relationship since they are critical to the finally risk evaluation and also the selection of this model had a significant impact on the magnitude of risk estimates.

Purpose: The objective of this study was to compare two dose-response models, i.e., exponential and Weibull-gamma, and determine which model was more appropriate for QMRA of Listeria monocytogenesin ham products in Korea.

Methods: The model was constructed for QMRA of L. monocytogenes with consumption on ham products in Korea, according to Codex guidelines. Frame-work model as product-retail-consumption pathway composed with initial contamination level, the time and temperature in distributions, and consumption data sets for ham products and also used the developed predictive growth. To determine more appropriate of two dose-response models QMRA of L. monocytogenes used and compared the exponential and Weibull-gamma model. The simulation model and formulas with Microsoft^®Excel spreadsheet program using these data sets was developed and simulated with @RISK.

Results: The probability of foodborne disease by L. monocytogenes with consumption of the ham products per person per day, in which simulated with the exponential and Weibull-gamma dose-response models, was estimated similarly as 5.80 x 10^-14 , 3.71 x 10^-14 in low risk population, and 2.80 x 10^-10 , 7.07 x 10^-10in high risk population, as a mean values, respectively. However, the lesser estimated risk level, the Weibull-gamma model greater risks than the exponential model with Pearson correlation coefficients value was 0.8009.

Significance: This study can be used as scientific information and showed the practical possibility using and picking up dose-response models for QMRA.