摘要
目的为全国人群乙型病毒性肝炎(乙肝)血清流行病学调查,筛选适宜的乙肝病毒表面抗原(Hepatitis B Virus Surface Antigen,HBsAg)检测试剂。方法选择符合国家质量标准,且在全国销量最大的五种(A、B、C、D、E)酶免疫测定法HBsAg检测试剂,平行检测由63份阳性和90份阴性品组成的参比系统各5次,将每个检测的s/n值做log(s/n+1)转换,计算并比较各试剂的可靠性指标组内相关系数(Intraclass Correlation Coefficient,ICC)、变异系数(Coefficient of Variability,CV)和真实性指标受试者工作特征(Receiver Operating Characteristic,ROC)曲线下面积(Area Under Curve,AUC)、部分(partial)ROC AUC(pAUC)、固定特异度下灵敏度[Sensitivity at a Fixed Specificity,Se(FPR=e)]等。结果五种试剂ICC、CV从优到劣的排位顺序均为E、C、D、B、A,且差异有统计学意义(Bootstrap法,P<0.05);五种试剂的AUC均在0.965~0.987,pAUC最大的试剂E、B明显高于最小的试剂A、D(Bootstrap法,P<0.05);Se(FPR=e)最高的试剂E明显高于试剂C(Bootstrap法,P<0.05)。结论五种试剂均有较强的诊断能力,其中试剂E为最佳。
Objective To Screen best EIA kit that fits best for using in serological detection of HBsAg in the national investigation of hepatitis B virus infection. Methods Top 5 kinds of kits on the best seller list of Chinamade HBsAg EIA detection reagents were selected and each was detected the HBsAg panel (63 positive and 90 negative) 5 times. The s/n ratio of each test was recorded and transformed by log(s/n+1). Reliability indices ICC and CV, and validity indices AUC, pAUC and Se(FPR-e) were calculated and compared between each of the kits. Results The indices of ICC and CV of the 5 kits arranged in the same order of E,C, D, B and A, from the best to the worst, and the differences between them were significantly important (Bootstrap method, P〈0.05); The AUCs ranged from 0.965 to 0.987, and for kits E and B which had the biggest pAUCs, differs significantly from kits A and D (Bootstrap method, P〈0.05). Kit E had the highest Se(FPR-e) which was significantly bigger than kit B (Bootstrap method, P〈0.05). Conclusion The 5 kinds of kits are all excellent and kit E is the best.
出处
《中国疫苗和免疫》
CAS
2009年第5期388-391,共4页
Chinese Journal of Vaccines and Immunization
基金
卫生部全国乙型病毒性肝炎血清流行病学调查专项资金
关键词
诊断试验
乙型肝炎病毒表面抗原
评价
受试者工作特征曲线
Diagnostic test
Hepatitis B virus surface antigen
Evaluation
Receiver operating characteristic curve