Measurement uncertainty in medical analysis laboratories

Abstract:

A brief historical review of the measurement uncertainty estimation models that have been proposed for the clinical laboratory in recent years as a way to satisfy requirement 5.5. 1.4 ISO 15189.2012 is carried out. The compilation begins with the generic model GUM, in the «Bottom-Up» mode, adapted to the chemistry by EURACEHM, which when required to evaluate the uncertainty from its basic components, made it complex, difficult to apply to the routine of laboratories and without practical application. This is followed by simplified models of the «Top-Dowm» type, one proposed by the Royal College of Australasian Pathologists (RCPA) and then by the Spanish Society of Laboratory Medicine (SEQCML), which prioritize the contribution of intermediate precision in the medium term and add other contributions such as the uncertainty of the calibrator and eventually the one corresponding to the correction factor of the systematic error. The experimental application of this model demonstrates that expanded uncertainty remains virtually constant over the long term and does not reflect the behavior of analytical processes when evaluating data obtained from intercomparisons. These concepts are present in ISO Guide DTS 20914: 2018 which is in a very preliminary stage of development. The uncertainty estimation mechanism proposed by Nordtest is then evaluated, which includes the intermediate precision data obtained from the internal quality control routine and the bias obtained from the results of the intercomparison tests, experimentally showing that with respect to the other models, it more realistically reflects the behavior of the long-term analysis …

Año de publicación:

2019

Keywords:

Fuente:

google