Alternatives in Estimation of Measurement Uncertainty in Analysis of Pesticide Residues in Vegetables and Fruits

Estimation of Measurement Uncertainty in Pesticide Residues in Vegetables and Fruits

Abstract

         It is a requirement under ISO/IEC 17025:2017 that laboratories can assess the measurement uncertainty associated with analytical results. Estimating measurement uncertainty (MU) is necessary in pesticide residue analysis to comply with relevant regulations. In this study, three calculation methods in MU estimation were examined. The conventional bottom-up method for MU estimation using the ISO GUM approach was compared with two alternative top-down methods, including the estimation of MU based on intra-laboratory quality control (QC) data and proficiency testing (PT) data. The aim was to determine the method of choice for the estimation of MU in routine reports. The first typical method of MU by the ISO GUM was the conventional method. On the other hand, two alternative top-down methods were used: one using the QC data of 28 spiked samples collected for two years, and the other using the 3-round PT data collected from 33 data points. The results showed the relative expanded uncertainty of the conventional ISO-GUM method and two alternative methods based on QC data and PT data were 8.0, 31.7, and 32.4%, respectively. By comparing with the 50% default value used by international regulatory authorities, all approaches met the requirements to demonstrate that the laboratory’s MU did not exceed this value. However, the minimum value obtained from the ISO GUM was different from the two alternative top-down methods, which were comparable. The top-down methods were more appropriate and suitable alternatives for estimating measurement uncertainties.