American Journal of Obstetrics and Gynecology | 2019
Syphilis diagnosis in pregnancy: is the reverse algorithm better than the traditional algorithm?: 637
Abstract
637 Syphilis diagnosis in pregnancy: is the reverse algorithm better than the traditional algorithm? Jodie Dionne-Odom, Akila Subramaniam, Alexander Boutwell, Glenda Corley-Topham, Barbara Van Der Pol, Alan T. Tita, Edward W. Hook III Division of Infectious Diseases, University of Alabama at Birmingham, Birmingham, AL, Center for Women’s Reproductive Health, University of Alabama at Birmingham, Birmingham, AL, Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, AL, Division of Infectious Diseases, University of Alabama at Birmingham, Birmingham, AL, Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, AL, Center for Women’s Reproductive Health, University of Alabama at Birmingham, Birmingham, AL, Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, AL OBJECTIVE: Primary and secondary syphilis rates in the US are rising and congenital syphilis reports to CDC have increased 87% since 2012. Many hospital laboratories have switched to the reverse algorithm for syphilis testing with an automated treponemal screening test (syphilis enzyme immunoassay (EIA) or IgG) instead of a nontreponemal screen (RPR). Our goal was to compare performance characteristics of the traditional and reverse algorithms for syphilis testing in pregnancy. STUDY DESIGN: This retrospective cohort analysis included pregnant women tested for syphilis who delivered at our center between 11/ 2012 and 12/2017. The cohort was separated into two groups based on the lab algorithm switch date: “traditional algorithm” (11/201210/2014) and “reverse algorithm” (11/2014-12/2017). The treponemal screening test for the reverse algorithm was EIA (Trinity). Non-treponemal testing used RPR and follow-up treponemal testing used the T. pallidum particle agglutination assay (TPPA). Algorithm cohorts were compared for rates of screen positivity. RESULTS: Serologic testing for syphilis was performed at our institution for 8,790 pregnant women using the traditional algorithm and 11,403 women with the reverse algorithm. Maternal age, race, marital status, and location of testing (outpatient vs. inpatient) were similar for both groups: approximately 50% of women were Black and 80% were tested as outpatients. Under the traditional algorithm, 0.5% (95% CI 0.4-0.7%) (46/8790) were RPR screen +, and 50% (18/36) of those with reflex testing were TPPA+ (0.25% true infection). (Fig 1) Under the reverse algorithm, 0.7% (95% CI 0.6-0.9%) (83/11403) were EIA screen +, and 30% (25/83) were RPR+ (0.21% true infection). Screen positivity was similar in both groups (p1⁄40.08). In the reverse algorithm, only 41.4% (24/58) of EIA+/ RPRsamples had confirmatory TPPA testing, but 33.3% (8/24) were TPPA negative (EIA+/RPR-/TPPA-), indicating EIA false positivity. (Fig 2) CONCLUSION: In a diverse US region with high STI prevalence, the traditional and reverse syphilis testing algorithms in pregnancy have similar rates of screen positivity. Use of the reverse algorithm may be limited by an elevated proportion of discordant results (EIA+/RPR-) that require additional testing and can reflect previously treated infection. False-positive screens are a limitation of both algorithms and new strategies to detect active syphilis in pregnancy are needed.