Accuracy of on-site screening for syphilis among women attending a rural mobile antenatal clinic

Objective. To determine the accuracy and cost of on-site antenatal clinic syphilis screening, compared with laboratory-based screening. Setting. Mobile primary care clinic providing antenatal care in remote rural areas in Hlabisa health district; the prevalence of syphilis is 8.5%. Methods. 528 consecutive women screened for syphilis by a rapid plasma reagin (RPR) test on plasma derived from whole blood by gravity. A battery-powered, solar-charging rotator mixed antigen with plasma. The RPR test was repeated on serum in a reference laboratory for comparison. Results. The on-site test was highly sensitive (86.7%; 95% confidence interval (Cl) 72.5 - 94.5%) and highly specific (88.2%; 95% Cl 84.9 - 90.9). The positive predictive value (40.6%; 95% Cl 30.9 - 51.1) was low owing to a large number of samples reported as 'slightly positive' on site being reported as 'negative' in the laboratory. There were very few false-negative results on site (negative predictive value 98.6%; 95% Cl 96.8 - 99.4). The cost of the rotator is R3 521 and reagent costs are 81 c per woman screened. Conclusions. On-site RPR screening for syphilis is highly accurate, cheap and easy to do. This strategy may reduce perinatal mortality by increasing the proportion of pregnant women with syphilis who are adequately treated.