Investigation of precision and trueness of results from the imported commercial systems for measurement of gamma-glutamyitransferase in serum / 中华检验医学杂志

ABSTRACT

Objective To investigate the precision and trueness of results from six imported commercial systems for measurement of gamma-glutamyltransferase (GGT) in serum in order to provide reference for the clinical laboratories to verify the target accuracy. Methods Five fresh frozen human serum samples that differed in catalytic concentration were analyzed in two candidate domestic reference laboratories and the target values for GGT were assigned using IFCC reference measurement procedure. The same samples were tested by six commercial systems which were calibrated using the matched calibrator. Each system consisted of five instruments in five laboratories, which had been well maintained before measurement. The data was collected. Precision from the same manufacturer and different manufacturers and biases between target values and mean values from each system were calculated. Results The differences of the mean values for five levels of commercial systems varied from 16. 1% to 35.4%. For the five levels, the coefficients of variation (CVs) of the results from all measurement system were from 5.3% to 8. 8% , and CVs from each level were A 2. 17%- 5.07%, B 4. 21%-10. 98%, C 0. 52%-2. 38%, D 1.35%-2. 59%, E 0. 23%-1..54%-), F 1.83%-2. 38%. Biases between the mean values of each commercial systems and the target values were A 0. 43% -8.41% ), B -1.49% - -13.04% ), C 11.20% -17.73% ), D 0. 19% -4. 62% ), E -0. 30% - -2. 63% ), F 4). 46%-7.90%, respectively. The investigation showed that biases of two manufacturers were less than a quarter of the total allowable error (TAE) of The Clinical Laboratory Improvement Amendments of 1988 (CLIA'88) in the whole range of investigated concentrations and the other two manufacturers' biases could meet a quarter of TAE in a relative limited range. The biases of two manufacturers were near or more than half of TAE in most levels. It also revealed that the biases of more than half of manufacturers were more than a quarter of TAE in the low or high level of investigated concentrations. Conclusions The mean values of each manufacturer were significantly different. The variances of commercial systems from different manufacturers were much higher than those from the same manufacturer. Some imported commercial systems for measurement GGT should be better calibrated with the reference method, especially in the whole measurement linearity.