Evaluation of the Bio-Rad Syphilis IgG test performed on the CODA system for serologic diagnosis of syphilis

The performance of the Bio-Rad Syphilis IgG EIA test as a "screen for syphi lis" [testing first by EIA and then by the rapid plasma reagin (RPR) assay if the EIA was positive or equivocal] and as a confirmatory test was evalua ted by comparing results to those obtained by CAPTIA Syphilis-G. Discrepanc ies were resolved by repeating both EIAs and/or the SeroDia TP-PA (a partic le agglutination assay that replaced the microhemagglutination Treponema pa llidum test). Both EIAs were totally automated, the Bio-Rad test using the AutoPrep instrument for pipetting and the CODA system to perform all of the steps required to complete the EIA and interpret results, and the CAPTIA t est using the LabOTech(R) to accomplish both functions. Of 449 unselected s era submitted to "screen for syphilis," both EIAs agreed for 432 (96.2%) sp ecimens: 395 negative, 36 positive, and one equivocal. Fifty-four specimens were positive or equivocal by one or both EIAs; 41 of these were RPR react ive. Three of these 41 were incorrectly called negative by Bio-Rad (sensiti vity 92.7%), and there was I false-negative result by CAPTIA (sensitivity, 97.6%) (P, not significant). To further evaluate the Bio-Rad assay as a con firmatory test, 144 known RPR-reactive specimens were tested by both EIAs. Results agreed for 134 (93.1%): 123 positive, 11 negative. After resolving discrepancies, there were 3 false-negative and no false-positive results by Bio-Rad (sensitivity 97.8%, specificity 100%), and with CAPTIA there were no false-negative results and 1 false-positive (sensitivity 100%, specifici ty 91.7%) (P, not significant). The sensitivity of the Bio-Rad assay could be improved, without altering specificity, by lowering the cut-off value fo r equivocal results. In summary, the Bio-Rad Syphilis IgG EIA performed usi ng the AutoPrep instrument and CODA system is a reliable, efficient method of syphilis testing. (C) 2000 Elsevier Science Inc. All rights reserved.