RESUMO
OBJECTIVE: We evaluate diagnostic accuracy of the ARCHITECT chemiluminescent immunoassay (CIA) screening test in pregnancy, and evaluate pregnancy outcomes among screen-positive women. STUDY DESIGN: Samples from routine prenatal rapid plasma reagin (RPR) tests were collected between June 22 and August 18, 2017 and frozen. Samples were batch-tested with the Abbott ARCHITECT syphilis TP immunoassay (CIA, index test). We calculated sensitivity, specificity, predictive value, and false positivity. We compared pregnancy and neonatal outcomes among screen-positive women. RESULTS: Of 1,602 specimens, 35 (2.2%) were RPR + ; of those, 24 (69%) were CIA +/Treponema pallidum particle agglutination assay (TPPA)+ and 11 (31%) were CIA-/TPPA-. Of 1,567 RPR- specimens, 14 (0.9%) were CIA + ; of those, 13 (93%) were TPPA + , and one (7%) had a false positive CIA test. Sensitivity of the CIA (95% CI) was 100% (90.5-100%), specificity 99.9% (99.6-100%), positive predictive value 97.4% (86.2-99.9%), and false positive rate 0.06% (0.002-0.4%) for current or past syphilis. Among 37 CIA +/TPPA+ women, seven (19%) had RPR-negative status (Group 1), 11 (30%) had previously treated syphilis (Group 2), and 19 (51%) had active infection (Group 3). One stillbirth occurred in a woman with early, active syphilis identified at delivery; no adverse perinatal outcomes occurred among women in Groups 1 or 2. CONCLUSION: The ARCHITECT syphilis TP immunoassay accurately diagnoses current or past syphilis in pregnancy. Clinical history and staging remain essential using a reverse algorithm.