Probabilistic Assessment of Cyflumetofen Dietary Exposure Risks By a Bayesian Framework with Markov Chain Monte Carlo

Key Points

• This research aims to improve the assessment of dietary exposure risks for cyflumetofen by addressing uncertainties in model parameters.
• Developed a Bayesian linear regression model for pesticide residue using JASP to obtain posterior distributions.
• Utilized MCMC in R to extract key model parameters (α, β, σ) and predict concentration samples with associated uncertainties.
• Incorporated population strawberry consumption and weight data to calculate exposure levels and risk ratios using Crystal Ball.
• Performed 100,000 Monte Carlo simulations to output risk ratio distributions and compare against traditional methods.
• Traditional methods led to overestimating dietary risks due to neglecting uncertainty in model parameters.
• The proposed Bayesian-MCMC method provided a more accurate quantification of parameter uncertainties.
• The P95 value of risk ratio distributions was significantly different from traditional simulation outputs.