Background When surveying large groups at risk for STI, the search of anti T pallidum Ab is a valuable tool both for assessing risks associated with syphilis acquisition and, if followed by VDRL test, also for syphilis cases detection. By the other hand testing first with VDRL a large number of samples followed by confirmation with anti TP specific test would be laborious and prone to errors while the data of past cured syphilis would be lost. On 2010 the HIV/AIDS Program of Mexico City and the National Institute of Public Health, Mexico initiated a health survey of around 40 000 inmates of the city for assessing syphilis and other STIs. Presently more than 18 000 inmates have been surveyed for HIV, HBV HVC and syphilis using the automated immunoquimioluminiscence analysis system Abbott Architect i2000. The accuracy of Abbott Architect Syphilis TP (ASTP) for detecting treated and untreated syphilis was reported before thus our study focused in validate ASTP as compared with an accepted treponemic test.

Methods For evaluatig the sensitivity and specificity of ASTP in the context of our HIV clinic we compared ASTP with a test extensively used for syphilis confirmation, Treponema pallidum haemagglutination assay (TPHA). Samples were assayed with ASTP in pools of four sera and each positive pool developed and re-assayed with ASTP to find one or more individual positive samples which were further assayed with tittered VDRL. ASTP− pools were not re-assayed and individual samples were scored as negative for syphilis. We selected 218 ASTP+ and 1920 ASTP− individual consecutive samples to be tested with BioRad Syphilis TPHA.

Results From 218 ASTP+ samples 212 were TPHA+ and all 1920 ASTP− were also TPHA−. ASTP and TPHA detected all 77 VDRL+ samples thus considered diagnostic of latent or active syphilis. Six ASTP+/TPHA− samples and 135 ASTP−/TPHA− were also VDRL− and considered as evidence of treated/cured syphilis. Using TPHA as gold standard ASTP Sensitivity was 100% and Specificity was 99.7%. In two ASTP- pools that showed more than 0.6 but <1 S/CO reading, a sample weakly positive by ASTP and TPHA but VDRL− was found when assayed individually.

Conclusions Reverting the traditional algorithm of syphilis diagnosis by first determining TP specific antibodies with ASTP followed by tittered VDRL of positive samples is highly accurate even if done in pools of four sera. This approach allows also the identification of epidemiologically valuable data of cured syphilis.