[Estimate of uncertainty of measurements in clinical laboratories and in environmental, occupational, and preventive medicine].

According to recently issued (UNI CEI EN ISO/IEC 17025, UNI CEI ENV 13005 and prEN ISO 15189) standards, to assure the quality of analytical results and their comparability in time and in different places, testing and clinical laboratories must demonstrate to use validated methods, guarantee the traceability of their measurements and state the measurement uncertainty associated with each result. For some SI quantities, such as the mass, the traceability to SI and the estimate of measurement uncertainty are warranted by established methods. In the clinical laboratory and especially in preventive, environmental and occupational laboratory medicine, specific difficulties arise to warrant the traceability to the mole. On one hand, the whole concept of measurement uncertainty is new in laboratory medicine, on the other hand, its application faces practical difficulties, because of the wideness and the complexity of the analytical repertory, the lack of officially validated methods and matrix-specific reference materials traceable to SI. In this paper we discuss briefly the concept of measurement uncertainty and its meaning in comparison with other parameters used to define the performance of analytical methods. In addition, we describe the procedures recommended by international organisations for estimating measurement uncertainty and interpreting analytical results with an associated measurement uncertainty in comparison with limit values.