Commutability study on three CRMs evaluating their suitability for calibration, trueness verification and statistical quality control of methods measuring metal concentrations in human blood.

Abstract

The commutability of three certified reference materials for Cd, Cr, Hg, Ni, Pb and Tl in blood was evaluated for the combination of digestion ICP-MS with dilution ICP-MS or dilution GFAAS using the difference in bias approach. Maximum non-commutability bias (MANCB) was derived from the median of the tolerance limits set by six proficiency testing (PT) providers. The evaluation showed that this criteria are too strict and that either unfeasible high number of measurements or a larger MANCB is required to demonstrate commutability. The effect of increasing the MANCB on the maximum tolerable measurement uncertainty was found small to moderate demonstrating the commutability of the materials when used as a calibrator. Also the inclusion of the uncertainty of non-commutability into the measurement uncertainty resulted in only small increases of the measurement uncertainty, demonstrating the suitability of the materials for trueness control and method validation. The bias of the three CRMs was for most elements within the 99% confidence limits of the bias of the patient samples, demonstrating the suitability of the materials for statistical quality control.