Issues and challenges in applicability of measurement uncertainty estimation in medical laboratories
The ISO 15189 has endorsed the use of measurement uncertainty (MU) to estimate the quantity of measurement error of a given method, although the total error (TE) approach, originally postulated by Westgard et al. in 1974, is still widely used and appreciated in laboratory medicine. Both MU and TE present some struggles, especially regarding the estimation of both bias and imprecision. In particular, bias may be seen as a complicating factor because it needs specification of a reference quantity, which may be obtained by different sources. Different formulations of MU have been proposed by different guidelines so far, and each approach presents advantages or limitations. These methods are not comparable and there are considerable differences not only in the theoretic approach but also in MU formulation. Moreover, only few practical results have been reported in the literature, thus limiting comprehension and applicability of these methods, ultimately supporting the hypothesis that implementation of MU in medical laboratories remains in embryo. Additional research and developments in this area are hence advisable for establishing consensus on practical methods for medical laboratories.