应用ROC曲线确定化学发光法检测梅毒特异性抗体的最佳临界值

Determination of optimal cut off value of chemiluminescent immunoassay for treponemal specific antibody by receiver operating characteristic curve analysis

目的应用受试者操作特性(ROC)曲线,确定化学发光法(CLIA)检测梅毒特异性抗体的最佳临界值。方法收集经CLIA检测的临床血清标本共计261份,全部标本使用酶联免疫吸附试验(ELISA)和梅毒螺旋体明胶凝集试验(TPPA)进行复检。以TPPA作为参考标准,计算CLIA和ELISA检测梅毒抗体的敏感性、特异性、阳性预测值和阴性预测值;应用ROC曲线确定CLIA检测梅毒抗体的最佳临界值。结果 CLIA和ELISA检测梅毒抗体的敏感性分别为100.0%和97.1%,特异性为86.3%和91.1%,阳性预测值为72.9%和80.0%,阴性预测值为100.0%和98.9%。ROC曲线分析表明,当CLIA的临界值设定为2.41时,其方法的敏感性和特异性分别为100.0%和96.0%,与TPPA的总符合率为98.0%。结论当CLIA检测梅毒特异性抗体的临界值设定为2.41时,能够有效提高该方法的特异性。

Objective Receiver operating characteristic （ROC） curve was used to determine the optimal cut off value of ehemiluminescent immunoassay （CLIA） for treponemal specific antibody. Methods 261 serum samples were collected and tested by CLIA from clinic laboratory; all samples were retested by treponema pallidum particle agglutination assay （TPPA） and enzyme-linked immunosorbent assay （ELISA）. The sensitivities, specificities, positive predictive values and negative predictive values of CLIA and ELISA were evaluated by the results of TPPA. ROC was used to determine optimal cut off value of CLIA for treponemal specific antibody. Results The sensitivity, specificity, positive predictive value and negative predictive value of CLIA were 100.0%, 86.3%, 72.9%, and 100.0%, respectively. The sensitivity, specificity, positive predictive value and negative predictive value of ELISA were 97.1%, 91.1%, 80.0%, and 98.9%, respectively. ROC curve analysis demonstrated that the sensitivity and specificity of CLIA would be 100.0% and 90.0% if the cut off value was 2.41. Conclusion The cut off value of CLIA for treponemal specific antibody was set to 2.41 that could improve the specificity.