Factors affecting the quantitative uncertainty of the estimated short-term intake. Part II—Practical examples

The effects of the spread of residue concentrations in the samples derived from the selected supervised trials and the number of trials were studied on the magnitude and uncertainty of the short-term dietary intakes calculated with the proposed new procedure (IESTIₚ) and that one used currently by the FAO (Food and Agriculture Organization) and WHO (World Health Organization) Joint meeting on Pesticide Residues (JMPR) (IESTIc). The residue data of 10 pesticides were obtained from supervised trials conducted on apples and pears. The methods described in Part I were used for the calculations of the uncertainty. The results indicate that the ratio of IESTIP to IESTIc (φIESTI) is directly proportional to the ratio of the estimated maximum residue level (MRL), recommended by the JMPR; to the highest residue (HR) observed in supervised trials, and it may have a wide range depending on the particular conditions. The φIESTI becomes greater with the increase of the difference between the mrl or maximum residue limit (MRL, established by the Codex Alimentarius Commission, CAC) and HR, and becomes smaller if the difference between the large portion (LP) and unit mass (U) decreases. The φIESTI ranged between 2 and 5.1 in the 16 cases examined indicating that the IESTIₚ calculation method leads to higher intake estimates. The ratio of CVIESTIₚ and CVIESTIc ranged typically between 0.62 and 1.71. It rapidly increased up to 12 trials. For a larger number of trials, the ratio remained practically constant (1.69–1.71). The processing factor (PF) equally affects the MRL and HR values, therefore, it will not practically influence the φIESTI. The uncertainty of the estimated median residues depends on the spread and number of values in the residue datasets, which affects the uncertainty of the conversion factor (CF) and subsequently the uncertainty of the estimated IESTIₚ. Residue values obtained from minimum nine independent trials are required for the correct calculation of the 95% confidence intervals of the calculated median residues. The uncertainty of the analytical results directly affects the median, HR values and indirectly the calculated mrl and the MRL derived from it. Therefore, it should also be considered for the calculation of the combined uncertainty of the conversion factors. For the correct interpretation of the results of dietary exposure calculations, the upper 95% confidence limit of the short-term intake should also be considered. However, it is not the current practice of regulatory agencies or JMPR.