Measurement uncertainty for clinical laboratories – a revision of the concept

Abstract

The uncertainty of a measurement result is a fundamental concept in metrology indicating the range within the "true" value of a measurement should lie. Although not commonly reported with results, the calculation of measurement uncertainty (MU) has become common in routine clinical laboratories. Interpretation of numerical pathology results is made by comparison with data from other measurements. As MU is aimed at assisting with result interpretation, it should be related to the specific comparison being made. There are three basic type of comparators: a previous result from the same patient, a population reference interval, or a clinical decision point. For each comparison, the "true" value is that which would have been obtained from the instrument used to make the comparator measurements if it was measured without uncertainty. The MU is the range of likely deviations from this true value due to the method used to produce the result under interpretation. For patient monitoring, if the two measurements were made on the same analyzer, the uncertainty is the imprecision of the assay over the relevant time frame. In comparing with a manufacturer-specific reference interval, the MU is deviation from the manufacturer's master calibrator. For clinical decision points produced with the assays traceable to international references, the MU is related to deviation from that reference standard. For optimal use of MU in the clinical laboratory, it may be necessary to consider the use of the test result and the concept of a single MU for each result may need to be revised.