[Establishment of an algorithm for serological testing of syphilis identification]

Serological methods are widely used for the laboratory diagnosis of syphilis or for screening purposes. The aim of this study was to determine an algorithm for the application of laboratory tests that will provide accurate diagnosis of syphilis in a cheap, fast and practical way. A total of 162 serum samples were evaluated by the following tests: VDRL (Venereal Disease Research Laboratory; Omega Diagnostic, UK), TPHA (Treponema pallidum Hemagglutination Test; Omega Diagnostic, UK), ELISA IgG + IgM (Enzyme Linked Immunosorbent Assay; DiaPro Diagnostic Bioprobes, Italy), FTA-ABS (Fluorescent Treponemal Antibody-Absorption; IgG, Euroimmun, Germany) and WB (Western Blot; IgG Euroimmun, Germany). When the gold standard was considered as FTA-ABS test, the sensitivity, specificity, positive and negative predictive values for VDRL were 77.1%, 100% 100% an 80.6%; for TPHA were 92.8%, 98.7%, 98.7% and 92.9%, for ELISA 98.8%, 98.7%, 98.8% and 98.7%, and for WB 98.8%, 100%, 100% and 98.7%, respectively. When the results of screening with VDRL together with TPHA were compared with FTA-ABS, it was observed that if both VDRL and TPHA results were positive, then there was 100% concordance between the tests. However, when both of the test results were negative, 1.3% of them yielded positive result with FTA-ABS. If either one of VDRL or TPHA results were positive (n = 24), 95.8% (n = 23) was positive with FTA-ABS. Therefore, inconsistent results obtained by VDRL and TPHA requires verification by another method. When ELISA or WB tests were used, the borderline results need verification, however, positive or negative results would be reported. The determination of an algorithm for laboratory tests also depend on the number of patients, cost, cost per positive patient and workload of the laboratory. Thus, ELISA could be selected when the number of cases is high and the results should be reported unless they are suspicious. When the number of cases is low, VDRL/TPHA should be selected, and the results should be verified if they are inconsistent. However, the demographic characteristics of patient groups are also important in test selection and work flow. False positive results are troublesome in case of marriage pre-screening and false negative results in sex workers. When all these factors are taken into consideration it may be suggested that either ELISA or VDRL together with TPHA should be performed and the results should be confirmed by a reference test in case of borderline results in ELISA or inconsistency between VDRL and TPHA results. Although screening for syphilis in the setting of blood banking is a matter of debate, if it is to be performed, then ELISA would be better since the work load is high. In case of pregnancy inconsistent VDRL and TPHA results should be verified since no risk could be afforded.