Determining of acceptable limits of uncertainties in radiotoxicology by using a method based only on external quality assessment results

PROCORAD organizes for more than 20 years proficiency tests in radiotoxicology to estimate the performance of the laboratories in term of accuracy. The results of these intercomparisons also show a wide range of uncertainties values provided by these laboratories. The regulatory obligations related to accreditation, which require that the laboratory shall define the performance requirements for measurement uncertainty and regularly review estimates of measurement uncertainty, and the lack of published or required acceptable limits in radiotoxicology led PROCORAD to propose uncertainties acceptable limits from the "state of art". For this purpose, PROCORAD used a new method to estimate the uncertainties based only on external quality assessment results: the Long-Term Uncertainties Method (LTUM). This method was applied for tritium in urines, gamma/X emitters in urines and actinides in fecal ashes. This study allowed validation of LTUM by a simple and reliable method for radiotoxicology uncertainties measurements. Three targets (optimal, desirable and minimal) for acceptable limits were defined by using the first quartiles, medians and third quartiles calculation of uncertainties provided by LTUM: 10%, 15% et 20% for H-3 respectively, 18%, 21% et 27% for g/X emitters respectively, 15%, 25% et 34% for Pu-238 respectively, 11%, 17% et 26% for Pu-239 respectively, 19%, 26% et 38% for Am-241 respectively and 19%, 30% et 52% for Cm-244 respectively. Applying the LTUM method to the results provided by laboratories shows an underestimation of the uncertainties calculated by laboratories, indicating that all influencing factors are not always taken into account in the uncertainties measurements.