摘要
Objective:To apply a regression equation to calculate the gray zone range for detecting Treponema pallidum antibody(TP-Ab)in blood samples using the ARCHITECT i2000SR fully automated microparticle chemiluminescence analyzer.Methods:A total of 180 blood samples initially screened for syphilis antibodies in our hospital were collected,which were tested using both the Treponema pallidum chemiluminescent microparticle immunoassay(TP-CMIA)on the ARCHITECT i2000SR electrochemical luminescence analyzer and the Treponema pallidum particle agglutination assay(TPPA).The optimal regression equation was established based on the TPPA cut-off index(COI)values and positive predictive probability,which was then used to determine the gray zone range for TP-CMIA.Results:This study analyzed 180 serum samples using a binary logistic regression model to evaluate the predictive value of COI values determined by the TP-CMIA method for syphilis.The Hosmer-Lemeshow test yielded P=0.332,further confirming the model's goodness-of-fit.The COI value was identified as a sensitive predictor of syphilis,with the cubic curve model demonstrating optimal fitting performance(R^(2)=0.925).When the positive predictive probability reached 95%,the corresponding COI value was 5.353,indicating that samples with COI values exceeding 5.353 had a true-positive probability>95%.Consequently,the gray zone range for TP-CMIA COI values was defined as 1.000-5.353,with values below this range classified as a false-positive interval.The model accurately classified 98.7%of syphilis cases and 87.0%of non-syphilis cases,achieving an overall correct classification rate of 97.2%.Conclusion:The gray zone range for COI values of TP-Ab detected by the TP-CMIA method(ARCHITECT i2000SR)was established as 1.000-5.353 using a regression equation,with samples showing COI values≤5.353 carrying a potential false-positive risk.
