Background: This study was performed to investigate the necessity of the second treponemal test and to evaluate the diagnostic performance of the first treponemal test in the reverse algorithm of syphilis diagnosis.
Methods: Abbott Architect Syphilis TP assay, a chemiluminescence immunoassay (CIA), was used as the first step treponemal test. Treponema pallidum haemagglutination assay (TPHA) test results of reactive samples from the first test were recorded. TPHA test result was considered as confirmatory. TPHA test results were grouped according to their Abbott Architect Syphilis TP assay results and they were compared with Mann-Whitney U test. For Abbott Architect Syphilis TP assay, a cutoff value with 100% specificity was determined via a ROC curve analysis which would render TPHA test unnecessary.
Results: Out of 146,800 samples 2,646 were reactive in the first step. Of those, 2,002 had a TPHA test result. Of the 2,002 TPHA tests, 1,706 were positive and 296 were negative. TPHA positive and negative groups have significantly different CIA signal/cutoff values. Using a ROC curve built for evaluation of the first-step test, the maximum Youden’s index value was found as 5.26. If this value would be accepted as cutoff, it would have a specificity of 85%. The specificity of 100% can be reached if a new cutoff value is set to 27.83.
Conclusions: Calculated cutoff value with 100% specificity is not practically applicable. It achieves saving of TPHA test in only six percent of reactive samples. Architect Syphilis TP assay is advantageous in large laboratories but is not enough to lead diagnosis without a second treponemal assay. Therefore it was decided to continue the reverse algorithm with dual treponemal assays.