一种白细胞介素-6荧光免疫层析检测方法的临床评价

Clinical Evaluation of a Fluorescence Immunochromatographic Assay for Interleukin-6 Detection

目的 对一种通过荧光免疫层析法检测白细胞介素-6(interleukin-6, IL-6)水平的新方法进行性能评价。方法 收集2021年12月北京协和医院疑似感染患者和体检正常人群的104份临床剩余血清样本,以飞测IL-6荧光免疫层析法(广州万孚生物技术股份有限公司,方法A)、IMMULITE1000化学发光免疫分析法(德国西门子公司,方法B)为参照,对待评价方法[荧光免疫层析法,伊诺达生物科技(太原)有限公司]检测血清IL-6的试剂等效性进行评价,同期收集39对同源配对的血浆和血清标本,评估待评价方法检测不同类型样本IL-6含量的等效性。结果 定量分析时(2份样本由于超出方法B检测上限被排除),待评价方法与方法A、方法B检测血清IL-6水平的回归方程分别为Y=-7.0950+1.1924X(R^(2)=0.9448)和Y=-2.6143+1.3072X (R^(2)=0.9391),Pearson相关系数分别为0.9720和0.9691;分别以方法A、方法B为参照时,待评价方法的Bland-Altman偏倚分别为-3.0、8.0,在医学决定水平处(7 ng/L)的预期偏倚分别为-1.44(95%CI:-5.37-2.50)ng/L和1.97(95%CI:-2.09-6.01)ng/L,95%CI均包含允许误差(±15%,-1.05-1.05 ng/L)。定性结果显示,待评价方法与方法A、方法B之间的总符合率分别为84.6%、83.7%。待评价方法对39对同源配对血浆和血清样本IL-6检测结果的一致性良好(P<0.0001),Bland-Altman偏倚分析显示,仅1个(2.6%,1/39)数据位于最大允许误差范围之外。结论 本文评价方法与目前临床应用的相同或不同原理的2种方法,对于IL-6水平的检测均具有良好的一致性和相关性,且检测结果几乎不受血清/血浆样本类型的影响,一定程度上可满足临床快速检测需求。

Objective To evaluate the performance of a new fluorescence immunochromatographic assay for interleukin-6(IL-6) detection. Methods A total of 104 serum samples from patients with suspected infections and normal population were collected at Peking Union Medical College Hospital in December, 2021. Assay A(Wanfu biofluorescence immunochromatography assay) and B(Siemens chemiluminescence assay) were used as references to evaluate the equivalence of the Innova assay for detecting serum IL-6. Meanwhile, 39 homologous paired plasma and serum were collected to evaluate the consistency of the Innova assay in detecting IL-6 content in different types of samples. Results In quantitative analysis, two samples were excluded because the content was above the limit of detection for method B. Compared with the assay A or B, the regression equations of the Innova assay were Y=-7.0950+1.1924X(R^(2)=0.9448), Y=-2.6143+1.3072X(R^(2)=0.9391), and the Pearson correlation coefficients were 0.9720 and 0.9691, respectively. With assay A or B as reference, the Innova assay showed the Bland-Altman bias values of-3.0 and 8.0, while the expected bias at the medical decision level(7 ng/L) was-1.44 ng/L(95% CI:-5.37 to 2.50), 1.97 ng/L(95% CI:-2.09 to 6.01), respectively, with 95% CI intervals all included permissible errors(±15%,-1.05 to 1.05 ng/L). The qualitative results showed the high total coincidence rate between the Innova assay and assay A(84.6%), and assay B(83.7%). In addition, the consistency of the Innova assay was good for 39 homologous pairs of plasma and serum samples(P<0.0001). Only one data(2.6%, 1/39) was identified by Bland-Altman bias analysis to be outside the maximum allowable error range.Conclusions Compared with the two reference assays with the same or different principles used in clinical practice, the Innova assay for IL-6 detection has good consistency and correlation, and its test results are almost not affected by the type of serum/plasma samples. To a certain extent, the Innova assay meets the needs of rapid detection in clinical practice.