A Laboratory-Based Evaluation of Four Rapid Point-of-Care Tests for Syphilis

Background: Syphilis point-of-care tests may reduce morbidity and ongoing transmission by increasing the proportion of people rapidly treated. Syphilis stage and co-infection with HIV may influence test performance. We evaluated four commercially available syphilis point-of-care devices in a head-to-head comparison using sera from laboratories in Australia. Methods: Point-of-care tests were evaluated using sera stored at Sydney and Melbourne laboratories. Sensitivity and specificity were calculated by standard methods, comparing point-of-care results to treponemal immunoassay (IA) reference test results. Additional analyses by clinical syphilis stage, HIV status, and non-treponemal antibody titre were performed. Non-overlapping 95% confidence intervals (CI) were considered statistically significant differences in estimates. Results: In total 1203 specimens were tested (736 IA-reactive, 467 IA-nonreactive). Point-of-care test sensitivities were: Determine 97.3%(95% CI: 95.8-98.3), Onsite 92.5%(90.3-94.3), DPP 89.8%(87.3-91.9) and Bioline 87.8%(85.1-90.0). Specificities were: Determine 96.4%(94.1-97.8), Onsite 92.5%(90.3-94.3), DPP 98.3%(96.5-99.2), and Bioline 98.5%(96.8-99.3). Sensitivity of the Determine test was 100% for primary and 100% for secondary syphilis. The three other tests had reduced sensitivity among primary (80.4-90.2%) compared to secondary syphilis (94.3-98.6%). No significant differences in sensitivity were observed by HIV status. Test sensitivities were significantly higher among high-RPR titre (RPR >= 8) (range: 94.6-99.5%) than RPR non-reactive infections (range: 76.3-92.9%). Conclusions: The Determine test had the highest sensitivity overall. All tests were most sensitive among high-RPR titre infections. Point-of-care tests have a role in syphilis control programs however in developed countries with established laboratory infrastructures, the lower sensitivities of some tests observed in primary syphilis suggest these would need to be supplemented with additional tests among populations where syphilis incidence is high to avoid missing early syphilis cases.