Last Updated: 19-10-2020 15:53
Within the framework of the HBM4EU project, a 3rd interlaboratory comparison was organised for the determination of nine pesticide biomarkers in urine. Included were biomarkers for glyphosate and AMPA (glyphosate and AMPA), chlorpyrifos (TCPy), and pyrethroids (3-PBA, 4-F-3-PBA, cis-DBCA, cis-DCCA, trans-DCCA, and ClF3CA).
The study was performed in March/April 2020 (extended into June) and was conducted to assess the comparability and reliability of analytical methods across the participating expert laboratories.
The HBM4EU QAU had selected four expert laboratories for pesticide biomarkers in urine. The expert laboratories were from four different countries in Europe. For glyphosate/AMPA, two additional laboratories analysed the samples (the lab preparing the control material, and an external laboratory).
Each participant received two control materials of human urine to be analysed for glyphosate and AMPA (present at 0.2-6 ng/ml), and two other control materials, which both contained the chlorpyrifos biomarker (1.7-5.3 ng/ml) and pyrethroid biomarkers (0.1-1.0 ng/ml). The laboratories were requested to perform a single analysis and the submit results to the organiser within 3 weeks. Due to the outbreak of COVID-19 and (partial) closure of laboratories, it took much longer to receive all results.
A first assessment of comparability of results was done by calculation of the mean, the RSD, and the relative uncertainty of the mean. Results were compared against the mean through a Z-score when the relative uncertainty of the mean was within 17.5%. In case the relative uncertainty exceeded this value, no objective reliable quantitative comparability assessment could be done.
For glyphosate all six laboratories reported results. At the lower level the results were insufficiently comparable, at the higher level the results were comparable after exclusion of one outlier. For AMPA, five laboratories reported results which were comparable for both sample materials.
For the chlorpyrifos and pyrethroid biomarkers, three laboratories reported results for all seven biomarkers, and one for six biomarkers (no results for ClF3CA). In ten out of 14 cases, comparability of results could be demonstrated, in some cases after exclusion of one outlier. In four cases the relative uncertainty of the mean was too high for a quantitative assessment of comparability