An assessment of analytical performance using the six sigma scale in second-trimester maternal prenatal screening practices in China

Jinming Zhang,Xingtong Chen,Jiaming Wu,Penghui Feng,Wei Wang,Kun Zhong,Shuai Yuan,Yuxuan Du,Chuanbao Zhang,Falin He
DOI: https://doi.org/10.1016/j.plabm.2024.e00422
2024-07-23
Abstract:Objectives: We aimed to evaluate the analytical performance of second-trimester maternal serum screening in China, and to compare if there are differences in sigma levels across different methods and months. Methods: A retrospective study was conducted to assess the analytical quality levels of laboratories by calculating the Sigma metrics with prenatal screening biomarkers: AFP, Total β-hCG, free β-hCG, uE3. Data from 591 laboratories were selected. Sigma metrics were computed using the formula: Sigma metrics(σ) = (%TEa - |%Bias|)/%CV. The Friedman test and Mann-Whitney test were used to compare differences across various methods and different months. The Hodges-Lehmann was used for determining 95 % confidence intervals of pseudo-medians. Results: Only uE3 showed significant monthly variations in sigma calculations. However, around 8 % of laboratories across all four analytes demonstrated sigma levels both above 6 and below 3 in different months. Laboratories utilizing time-resolved fluorescence methods significantly outperformed those using chemiluminescence in sigma level. For AFP, the pseudo-median difference between these methods lies within a 95 % confidence interval of (-3.22, -1.93), while for uE3, it is at (-2.30, -1.40). Notably, the median sigma levels for all analytes reached the 4-sigma threshold, with free β-hCG even attaining the 6-sigma level. Conclusion: With current standards, China's second-trimester maternal serum screening is of relatively high analytical quality, and variations in sigma levels exist across different months and methods.
What problem does this paper attempt to address?