Speaker
Description
Interlaboratory comparisons and proficiency tests (PT) are an increasingly popular necessity used by laboratories as an external, independent confirmation of the quality of their results. How to find and choose the most appropriate PT, how can the laboratory influence the provider of PTs? And, vice versa: is providing the samples with a known true value and issuing the evaluation report to the laboratory the only task of the organizer? Can the provider of PT do more for the labs and how? These questions will be answered on the case of IARMA, rather new and small provider of PT tests.
IARMA (International Atomic Reference Material Agency) was established in 2012. The first release of PT samples occured in 2013. The first set of samples was devoted to gamma spectometry laboratories, consisted of three water samples and the sample of soil. 18 participants from 16 states participated on the first ERAD (Environmental RADioactivity). In 2014, tritium ETRIT (Environmental TRITium) was organized, with the set of eight water samples and 25 participants from 16 states. The program was enlarged in 2015 by EGROSS (Environmental GROSS alpha and beta emittors), with 6 water samples in the set. From 2016 on, all three PT tests regularly appear on the worldwide scene. In average, around 70 sets of samples are released yearly. ECARB (Environmental CARBon) joined as a last one in 2020.
Up to now, 148 different samples were prepared and sent in 2876 replicates, 45% of those samples were tap water, 9% ground water, 11% sea water and 5% of other matrices such as hay, seaweed, soil and mushrooms. All together, 7701 results were received and evaluated in the individual reports with short response time, followed by Summary report, with information about the sample matrices, sample preparation, homogeneity and stability tests, as well as assessments of bias, trueness, precision, uncertainty estimations, z-score. Beside regular numerical evaluations, the additional goals were analysed and discussed in each Summary report, such as Metod‘s Minimum Detection Limit, Blank (background) verification, Bias and uncertainty of the measurements results, Repeatability on duplicate samples, False positive / negative.
The interaction with customers is very important and valuable. PTs should be designed in a way that match with customer's needs. On the other side, the statistical evaluation of PT provider might help to individual laboratories in the decision making process about the most appropriate method for their needs and also in the process of method validation.
The resume of statistical analyses of results will be presented, as well as lessons learned during 10 years and vision and challenges which will guide IARMA in the next decade.
