Comparison of three syphilis algorithms in West China

The study was designed to analyze three different algorithms with the implementation of treponemal tests for detecting suspected syphilis in west China where no syphilis algorithms guideline exists.

We retrospectively collected data to reanalyze three syphilis testing algorithms: the classical reverse algorithm, and two options of syphilis screening in European syphilis guideline. The kappa (κ) coefficients were used to compare the concordance between algorithms using different syphilis assays during two periods. A receiver operating characteristic curve was used to determine the optimal S/CO ratios of InTec EIA or Lumipulse® G TP-N assay (CLIA) to predict confirmatory TPPA results.

The agreements between the reverse algorithm and the EU option 1 algorithms were perfect irrespective of EIA or CLIA used (all κ > 0.9). But the agreements between the EU option 2 algorithms and reverse or EU option 1 were not good (κ < 0.4). >50% cases confirmed syphilis infections by reverse or EU option 1 algorithm were missed by EU option 2. There is no very need for a confirmatory TPPA assay in EU Option 1 when S/CO is above 6.12 for CLIA (3.67 for EIA). The false-positive rate was 0.26% above this cutoff level.

EU Option 1, involving a reactive TT, followed by another TT of a different type and a quantitative NTT if second TT is positive, which is the same as the 2015 UK syphilis algorithm, is recommended. We also propose that when S/CO of the CLIA or EIA is high enough, TPPA confirmation can be omitted from the testing algorithm, and costs would be significantly reduced.