Possibility of harmonization for immune serum items: First report

Abstract

In immune serum test items, the actual differences among devices and reagents have not been sufficiently elucidated owing to the matrix effect of samples for external accuracy control. In this study, we performed factor analyses of differences among reagents and evaluated the appropriateness of test samples. Devices from four equipment manufacturers were investigated using a prostate-specific antigen (PSA), ferritin, α-fetoprotein (AFP), and carcinoembryonic antigen (CEA) as the test items. For samples, 30 panel sera and 16 test samples were used. Regarding the total average as the standard, the measured values of the panel sera were evaluated by regression analysis. The measured values were noted to be proportional among the methods used for PSA, ferritin, and AFP, and no value showed deviation. However, the measured value of CEA differed among the methods, showing no consistent tendency. By setting the criterion to the 95% confidence interval determined by the regression analysis of the panel sera, the reactivity of the test samples was evaluated. Samples prepared with pooled sera met the evaluation criterion, suggesting the usefulness of the pooled serum as a test sample.