Background: Measurement uncertainty (MU) characterizes the dispersion of the quantity values attributed to a measurand. Although this concept was introduced to medical laboratories some years ago, not all medical researchers are familiar with it. Therefore, the evaluation and expression of MU must be highlighted. In this paper, the evaluation of MU is described by using four different approaches from different quality assessment data.
Methods: In accordance with Guide to the Expression of Uncertainty of Measurement (GUM) principles, human serum γ-glutamyltransferase (GGT) level was defined as the measurand. Main sources of MU were analyzed; individual components of MU were evaluated, followed by calculation of standard uncertainty, the combined standard uncertainty and the expanded uncertainty.
Results: In method 1, the median of expanded uncertainty (k=2) of GGT in lower level (65±1 U/L) was 5 U/L (9%, 95% confidence interval) and in higher level (116±2 U/L) was 8% (95% confidence interval), respectively. The results of method 2 were lower than that of method 1. There were no significant differences between the two other methods compared with the method 1.
Conclusions: Three out of the four different approaches based on different quality assessment data yielded similar results. Proficiency testing or external quality assessment data used for MU evaluation can be regarded as a supplementary method in clinical laboratory.
Keywords: bias; internal quality control; measurement uncertainty; proficiency testing; reproducibility.