Practical application of biological variation and Sigma metrics quality models to evaluate 20 chemistry analytes on the Beckman Coulter AU680

Abstract Objectives This study aimed to evaluate the imprecision and bias data generated for 20 routine chemistry analytes against both the biological variation fitness for purpose (FFP) and Sigma metrics (SM) criteria. Design and method Twenty serum/plasma analytes were evaluated on the Beckman Coulter AU680. Third party commercial lyophilized internal quality control samples of human origin were used for day-to-day imprecision calculations. Commercial external quality assurance (EQA) samples were used to determine the systematic error between the test method result and the instrument group mean result from the EQA program for each analyte. Biological variation data was used to calculate the minimum , desirable and optimal imprecision and bias for determination of FFP. The desirable total allowable error was determined from biological variation data and applied to the SM calculation. The outcomes of both quality approaches were then compared. Results The day-to-day imprecision of most tested analytes (except sodium and chloride) were smaller than the allowable imprecision (ranging from minimum to optimum ). Most analytes achieved at least minimum bias. The SM varied with analyte concentration with six analytes producing low Sigma values. Comparing the quality processes eleven analytes produced a green light for both FFP and SM. There was some difference seen in interpretation for the other nine analytes. Conclusions The individual interpretation of bias and imprecision using FFP criteria allowed for the clear determination of the major source of error. Whereas, SM provided a summative evaluation of method performance. But the selection of total allowable error (TEa) is fundamental to this interpretation and harmonisation of the TEa calculation is needed.