THE IMPACT OF UNTREATED SYPHILIS in pregnancy is substantial; 49% to 67% of pregnant women with active syphilis have adverse pregnancy outcomes, including abortion, stillbirth, preterm delivery, low birth weight, and congenital syphilis.1,2 In an African study, 16% of all adverse pregnancy outcomes could be attributed to congenital transmission of syphilis.3 These adverse outcomes can be averted through early detection and treatment of syphilis in pregnancy.4–7 Routine antenatal screening and treatment to prevent congenital syphilis has been shown to be cost-effective when compared with other important public health interventions.8
South African national policy calls for syphilis screening as part of antenatal care; however, consistent, reliable screening and treatment are largely restricted to urban centers.9 In South Africa’s Eastern Cape Province, most antenatal clinics rely on offsite laboratories for processing antenatal specimens using a quantitative rapid plasma reagin (RPR) assay and Treponema pallidum hemagglutination assay (TPHA).10 For women with active syphilis, a single injection of benzathine penicillin can prevent or treat infection in the fetus.6 However, successful diagnosis and treatment of pregnant women with syphilis is impeded by lack of refrigeration, unreliable transport for specimens and test results to and from laboratories, and poor follow up for treatment by antenatal clinic attendees.11
Onsite screening for syphilis at antenatal clinics with immediate treatment for infected women could dramatically improve treatment rates and reduce the burden of disease for both mother and fetus. The RPR test can be a useful and inexpensive screening tool, but the sensitivity and specificity of the test are highly user-dependent.12–16,22 Rapid treponemal tests, using immunochromatographic strip (ICS) technology, are a simpler but more expensive alternative for onsite syphilis testing.17
We implemented onsite rapid syphilis screening in rural clinics providing antenatal care in Eastern Cape Province, South Africa, to determine its impact on maternal syphilis treatment rates. We compared onsite RPR, onsite ICS, and the standard practice offsite RPR/TPHA screening in terms of effectiveness (for diagnosing and treating active maternal syphilis), overtreatment (resulting from previously treated infections or false-positive results), and cost-effectiveness (see companion paper18).
Materials and Methods
Eight rural primary care clinics providing antenatal care in 4 districts of Eastern Cape Province, South Africa, were selected by the Eastern Cape Provincial Department of Health (ECPDoH) for implementation of onsite antenatal syphilis screening (“intervention clinics”). Clinics were chosen to represent the range of rural Eastern Cape clinics in terms of patient volume, distance from the offsite (district) laboratory, and access to utilities. For comparison, an additional 18 clinics in the same districts were similarly selected by the ECPDoH as “standard practice clinics.” These were visited before the study period; only 4 were currently conducting offsite RPR/TPHA antenatal syphilis screening and were retained in the study. The ECPDoH selected a further 4 similar clinics; these were conducting antenatal syphilis screening and were recruited into the study.
The final 8 standard practice clinics received no intervention and no visits during the study period. Standard practice clinic records were examined retrospectively to determine the percent of pregnant women who had blood drawn for syphilis testing, who received the results of their tests, and who received one, 2, or 3 weekly penicillin injections. Intervention and standard practice clinics did not differ in terms of patient volume, distance from the offsite (district) laboratory, or access to utilities: electricity, water, and phone.
A train-the-trainer approach was used for teaching the methods of onsite syphilis screening. Nurse clinicians were taught to perform both a qualitative RPR test (Macro-Vue RPR Card Tests; Becton, Dickinson and Co., Franklin Lakes, NJ) and an ICS test (Determine Syphilis TP; Abbott Laboratories, Abbott Park, IL). The nurses then returned to their clinics and trained several more nurses at each clinic to perform the RPR and ICS.
All pregnant women aged 18 years and over seeking antenatal services at the intervention clinics without a prior syphilis test during that pregnancy were offered onsite syphilis screening. Clinics alternated the 2 onsite screening methods weekly. A trained volunteer educated the clients about syphilis and explained the purpose of the study. Informed consent was obtained by a nurse clinician. For women tested by ICS, fingerstick blood was obtained. All clients had venous blood drawn for testing in the reference laboratory. A short questionnaire recorded the woman’s obstetric history and history of syphilis diagnosis and treatment. Women declining onsite testing were provided the standard offsite RPR/TPHA testing, but transport of specimens and results was ensured.
The onsite test was performed while the client waited. Each client received posttest counseling with her test result. Women with a positive onsite test were given the option of awaiting the reference laboratory results before receiving treatment but were encouraged to accept immediate treatment. Women choosing immediate treatment were given 2.4 million units benzathine penicillin divided equally into 2 gluteal injections and were asked to return weekly for 2 weeks to receive 2 further treatments according to the South African treatment guidelines for syphilis.9 Before penicillin was given, all participants were asked about a history of allergic reactions to injections. Partner notification cards were given to clients with a positive onsite test to encourage partners to come to the clinic for free treatment. Women with a negative onsite syphilis test were counseled to return in 1 week to receive the reference laboratory RPR/TPHA results and treatment (if indicated).
Serum from each patient was refrigerated onsite and collected weekly for testing in the reference laboratory at University of the Transkei (UNITRA). All sera were tested at UNITRA by quantitative RPR, ICS, and TPHA (Cellognost Syphilis H; Dade Behring, Deerfield, IL). Results from the reference laboratory were returned to the clinics within 1 week of the blood draw. Women who had a negative onsite test but a positive RPR and TPHA in the reference laboratory had treatment initiated at their next clinic visit. At the conclusion of the study, quality assurance testing was performed and a random 20% sample of the study sera were tested by quantitative RPR, ICS, and TPHA by the Reference Center for STDs at the National Health Laboratory Services (NHLS) in Johannesburg.
Six weeks into the study, interim analysis of the data revealed that the onsite ICS had poor sensitivity for detecting low-titer syphilis (RPR <1:8). Investigation revealed that insufficient blood was being obtained from the fingerstick. Nurse clinicians were provided with additional training on how to properly obtain fingerstick blood, and automatic lancets and calibrated, heparinized capillary tubes for drawing up the blood were introduced. The onsite RPR similarly had poor sensitivity, but for both low- and high-titer syphilis. Problems occurred at all intervention clinics but no systematic error was identified. District laboratory technologists provided enhanced onsite observation and instruction for nurses performing the RPR at each clinic. In the final analysis, we compared the performance of the onsite RPR throughout the study period with the performance of the onsite ICS achieved after the ICS refresher training.
Syphilis stage of disease was defined using the reference laboratory results. Women with a positive TPHA and a positive RPR were considered to have active syphilis. Women with a positive TPHA and an RPR titer ≥1:8 were considered to have high-titer active syphilis, whereas women with a positive TPHA and a positive RPR of titer <1:8 were considered to have low-titer active syphilis. For the purposes of our analysis, high-titer active syphilis was presumed to indicate primary, secondary, or early latent syphilis, but patients were not staged through examination or questions about symptoms. Low-titer syphilis was considered to be late syphilis or past, untreated syphilis, although it could represent recently acquired primary syphilis. A positive TPHA with a negative RPR was considered to be past or previously treated infection. A negative TPHA with a positive RPR was considered to be a biologic false-positive RPR, although these could also represent early syphilis infections. Women with a negative TPHA and negative RPR were considered to be uninfected. The reference laboratory was used as the standard in determining the accuracy of the onsite test results.
We compared the intervention clinics and the standard practice clinics in terms of the percent of women with a positive syphilis test who received one, 2, or 3 appropriately timed weekly doses of penicillin.
At the close of the study, structured questionnaires were administered to the nurse clinicians who participated in the study to assess the acceptability of the onsite testing approaches and their preferred testing method.
The study protocol was approved by the Institutional Review Boards of the Centers for Disease Control and Prevention and the University of Witwatersrand, Republic of South Africa, the latter review board providing ethical oversight for both the Department of Health of Eastern Cape Province and the University of Transkei.
The study was conducted from December 2001 through March 2002. Of 1,285 women enrolled, complete reference laboratory results were obtained for 1,250 women: 555 women were screened using the onsite RPR and 695 women using the onsite ICS. From the reference laboratory results, 79 women (6.3%) had active syphilis: 21 (1.7%) had high-titer active syphilis and 58 (4.6%) had low-titer active syphilis. Eighty-eight women (7.0%) had past or previously treated syphilis (positive TPHA only). Thirteen women (1.0%) had a biologically false-positive RPR with negative TPHA. At standard practice clinics, 1,456 women were tested for syphilis at the offsite district laboratory. Of these, 54 women had no results recorded, 17 women had no specimen or an unlabeled specimen taken, results were provided to a private doctor instead of the clinic for 5 women, and 1,380 women had results returned to the clinics. For the 79 women from intervention clinics with active syphilis, the mean gestational age at presentation and treatment was 26.6 weeks (median, 28 weeks).
The sensitivity and specificity of the onsite tests compared with the reference laboratory RPR and TPHA results are shown in Table 1. The sensitivity of the ICS for detecting high-titer active syphilis remained 100% (95% confidence interval [CI] = 77–100%) throughout the study, whereas the sensitivity for detecting low-titer syphilis was immediately improved by the interim refresher training. For the RPR, sensitivity remained poor even after intensified supervision by the laboratory technologists.
RPR and TPHA quality assurance testing at the NHLS showed highly concordant results with the study reference laboratory: 249 (95.4%) of 261 samples agreed. Three samples classified as low-titer syphilis in the reference laboratory were classified as past or previously treated syphilis at the NHLS, and these women may have received penicillin unnecessarily. The 9 other discordant samples were all RPR-negative samples that differed by the TPHA; 6 were TPHA-negative in the reference laboratory but positive in the NHLS and the reverse occurred for the other 3. Because the reference laboratory results differed minimally from the NHLS results, no overall adjustment to the reference laboratory results was made in the final determination of the sensitivities and specificities of the onsite tests. Overall, the onsite ICS results agreed with the reference laboratory ICS results for 657 (94.5%) of 695 tests. After the retraining, 336 (98.5%) of 341 tests agreed (there were 4 false-negative and one false-positive onsite ICS results).
The percent of women with a positive syphilis test who received one, 2, or 3 appropriately timed weekly doses of penicillin is shown in Table 2. Significantly more women received at least one dose of penicillin at the intervention clinics, where onsite testing allowed immediate treatment. For the intervention clinics, 93 (89.4%) of 104 eligible women received at least one dose of penicillin; no adverse reactions occurred. Of the 11 women who received no penicillin, one absconded, one refused treatment, one elected to wait for the reference laboratory results but did not return to the clinic, 2 did not receive onsite treatment when it was indicated as a result of nurse error, and 6 had negative onsite tests and positive reference laboratory results but never returned to the clinic to learn their results. Results from the offsite laboratories were returned to the standard practice clinics for 1,380 (94.8%) of the women who presented for syphilis testing; 79 (5.7%) of women with results returned had a positive RPR and TPHA, and 48 (60.8%) of those received at least one dose of penicillin.
Of the 3 methods, the onsite ICS resulted in the greatest percentage of pregnant women correctly diagnosed and subsequently treated for active syphilis. The onsite ICS had 100% sensitivity (95% CI = 77–100%) for detecting high-titer syphilis and 89.4% of women diagnosed onsite received at least one injection of penicillin. Therefore, in the ICS group, 89.4% of women with high-titer syphilis received curative treatment that would also prevent or treat syphilis infection in the fetus. For the purpose of our analysis, we assumed that the offsite testing laboratories used by the standard practice clinics provided a correct diagnosis 100% of the time. Nevertheless, only 60.8% of women were effectively treated for high-titer syphilis because of the poor client return rates. The onsite RPR performed poorly because the sensitivity of the test was only 71.4% (95% CI = 29–96%) for high-titer active syphilis and therefore only 63.9% of women and their fetuses were effectively treated.
Of the 695 women who had an onsite ICS, 74 (10.6%) had a positive result. Only 31 (41.9%) of these had active syphilis (positive RPR and positive TPHA in the reference laboratory). The remaining 43 women had a positive ICS result but a negative RPR or TPHA in the reference laboratory. Of these, 36 women had no history of syphilis or treatment for syphilis, suggesting that the positive ICS should not be attributed to a previously treated infection and that the women warranted treatment with penicillin. The remaining 7 women (1.0% of the 695 women screened by ICS) had a history of syphilis and/or treatment with penicillin for a sexually transmitted disease and may therefore have received penicillin onsite unnecessarily.
The onsite testing strategy was highly acceptable to the 22 nurse clinicians in the intervention clinics. Onsite testing was favored over offsite testing because prompt diagnosis, patient education, and immediate treatment were possible. All of the nurse respondents found the onsite ICS fast, easy to perform, and reliable; 21 (95%) felt the onsite RPR was moderately difficult to perform stating it was time consuming, tricky to read, and unreliable. Overall, 15 of 22 nurses (68.2%) preferred the onsite ICS.
Onsite screening for syphilis at antenatal clinics is a practicable and effective means of diagnosing and treating maternal syphilis. The onsite ICS test was the most effective screening strategy with 89.4% of women with high-titer active syphilis being diagnosed and effectively treated. The onsite RPR was less effective because of the poor sensitivity of the test for detecting both high- and low-titer syphilis. The RPR is inherently difficult to perform and interpret for inexperienced personnel.
The effectiveness of the offsite RPR/TPHA suffered from poor client return rates, the primary reason that eligible women did not receive therapy. Despite previously identified problems with transportation of samples in Southern Africa,14,19 94.8% of women in standard practice clinics had results returned to the clinic. Assuming that test results returned to the clinic had the same proportion of positives as test results that were not returned, then 94.8% of missed penicillin treatments at standard practice clinics were the result of failure of antenatal clinic clients to return to the clinic for follow up. Barring a means of improving client return rates, accurate onsite diagnosis and treatment will retain its advantage over offsite testing for effectively treating maternal syphilis and preventing congenital syphilis. For many African clinics, successful transportation of specimens or laboratory results is likely less efficient than our observed 94.8%, and the benefits of onsite testing would therefore be even greater.
We encountered unanticipated difficulties with obtaining sufficient blood volume from the fingerstick. After providing refresher training and introducing calibrated, heparinized capillary tubes, the sensitivity of the ICS improved dramatically. By using heparinized tubes, the tested sample may have been plasma rather than whole blood. Whether this contributed to the improved results is difficult to determine. In contrast, performance of the RPR did not improve over the course of the study even with increased supervision; nurses consistently found the test difficult to conduct and interpret.
Use of a treponemal test for screening generated discussion locally about overtreatment of women whose positive onsite ICS was the result of past or previously treated syphilis. The rate of unnecessary treatment will depend both on the overall prevalence of syphilis (TPHA-positive women) and on the relative proportion of active versus previously treated syphilis in different settings (proportion of RPR-positive vs. RPR-negative women among those with a positive TPHA). In time, the proportion of women who were once infected but have been appropriately treated will increase. These women may still test positive with a treponemal test such as the ICS and thus the proportion of women receiving unnecessary treatment will increase over time if they are repeatedly tested using the ICS. However, in view of the significant morbidity and mortality averted through antenatal syphilis screening, the benefits of effective treatment certainly outweigh the risks of overtreatment for infected women in our and similar settings. New rapid tests under development that combine treponemal and nontreponemal components on one strip will circumvent this problem.
The prevalence of syphilis in pregnant women in our study (6.3%) likely underestimates the true prevalence of syphilis in pregnant women, because women with early miscarriages resulting from syphilis never present for antenatal care. In our study, the mean gestational age at presentation and treatment for infected women (26.6 weeks) was similar to that seen in other antenatal studies in Africa.4,5,8 There is conflicting evidence as to whether gestational age at treatment or time between treatment and delivery affects success of treatment.4,5,20 In one study, an interval of ≤30 days from treatment to delivery was associated with treatment failure.4 In our study, 99.6% of women received their first dose of penicillin at or before 36 weeks gestation. Screening can never realize its full benefit unless women attend antenatal care earlier in pregnancy. Hopefully, with the advent of onsite testing and immediate treatment, women will come to realize the benefit of antenatal care and will present earlier in pregnancy.
Ninety-four percent of pregnant women in South Africa attend antenatal clinics,21 but antenatal syphilis screening was uncommon in the province. In such a setting, the impact of any successfully implemented onsite ICS screening program would be a major improvement over the current practice. Also, women were tested only once for syphilis as part of this study. One study found the incidence of reinfection during pregnancy to be 0.4%, although the prevalence of syphilis at the first antenatal visit was only 1.9%.4 Ideally women should be screened twice during pregnancy, and partner notification is an important prevention strategy both for the community and for preventing reinfection of the woman.
Standard practice clinics were, by necessity for the study design, already conducting syphilis screening. There may be other unidentified differences between the populations served by the intervention versus standard practice clinics, which may have biased the results. However, the impact of onsite testing would be even greater had we accounted for the fact that only 22% (4 of 18) of the originally selected standard practice clinics were conducting antenatal syphilis testing. The ICS is clearly an effective means of improving diagnosis and treatment of antenatal syphilis in the study area and the test has performed similarly in other settings.23 However, how much treatment rates would improve with use of the ICS in other areas or countries depends on which factors, transportation of samples and results, client return rates, lack of antenatal testing, or other factors, currently hinder diagnosis and treatment.
Onsite ICS testing at rural antenatal clinics provides a feasible, practical, and effective approach to improving the diagnosis and treatment of syphilis during pregnancy. Treatment rates were 89.4% at intervention clinics compared with 60.8% at standard practice clinics with only 1% of screened women receiving unnecessary treatment. In Eastern Cape Province, onsite ICS testing has continued in intervention clinics and expansion to antenatal clinics throughout the province is in progress guided and funded by the Eastern Cape Provincial Department of Health. Given the limited extent of syphilis screening being conducted before this survey, reductions in congenital syphilis should be substantial. The onsite ICS is also cost-effective (companion paper18) warranting broader implementation of the test for antenatal screening in underresourced areas with moderate to high prevalence of maternal syphilis.
R. Bronzan implemented the study, analyzed and interpreted the data, and wrote the paper. D. Mwesigwa-Kayongo assisted with study site preparation, implemented the study, and assisted with analysis and interpretation of the data. D. Narkunas assisted with questionnaire development, implemented and monitored the study process, and facilitated cooperation between collaborators. G. Schmid helped to design the study and develop the protocol and provided critical review of the manuscript. G. Neilsen assisted in drafting the initial study protocol and facilitated early discussions of the protocol among collaborators, identified Eastern Cape Province as the study site, coordinated site visits, and reviewed the manuscript. P. Karuhije, J. Ddamba, E. Nombekela, and G. Hoyi conducted laboratory training and implemented and managed the laboratory component of the study. N. Makwedini provided Provincial Department of Health (DoH) support for the study, including coordination of provincial DoH nurses and supervisors, selection of clinics, and dissemination of results within the province. P. Dlali assisted with implementation and entered and managed study data. C. Ryan and J. Blandford assisted with study monitoring, interpretation of data, and critical review of the manuscript.
1. Hira SK, Bhat GJ, Chikamata DM, et al. Syphilis intervention in pregnancy: Zambian demonstration project. Genitourin Med 1990; 66:159–164.
2. Schulz KF, Cates W Jr, O’Mara PR. Pregnancy loss, infant death, and suffering: Legacy of syphilis and gonorrhoea in Africa. Genitourin Med 1987; 63:320–325.
3. Watson-Jones D, Changalucha J, Gumodoka B, et al. Syphilis in pregnancy in Tanzania. I. Impact of maternal syphilis on outcome of pregnancy. J Infect Dis 2002; 186:940–947.
4. Alexander JM, Sheffield JS, Sanchez PJ, et al. Efficacy of treatment for syphilis in pregnancy. Obstet Gynecol 1999; 93:5–8.
5. Sheffield J, Sanchez PJ, Morris G, et al. Congenital syphilis after maternal treatment for syphilis during pregnancy. Am J Obstet Gynecol 2002; 186:569–573.
6. Watson-Jones D, Gumodoka B, Weiss H, et al. Syphilis in pregnancy in Tanzania. II. The effectiveness of antenatal syphilis screening and single-dose benzathine penicillin treatment for the prevention of adverse pregnancy outcomes. J Infect Dis 2002; 186:948–957.
7. Temmerman M, Gichangi P, Fonck K, et al. Effect of a syphilis control programme on pregnancy outcome in Nairobi, Kenya. Sex Transm Infect 2000; 76:117–121.
8. Terris-Prestoholt F, Watson-Jones D, Mugeye K, et al. Is antenatal syphilis screening still cost effective in sub-Saharan Africa. Sex Transm Infect 2003; 79:375–381.
9. Republic of South Africa Department of Health. Protocols for the Management of a Person With a Sexually Transmitted Disease. South Africa: Department of Health, Directorate: HIV/AIDS and STDs, 1998.
10. Republic of South Africa Department of Health. Eastern Cape Report on Health Status. South Africa: Department of Health, Directorate: HIV/AIDS and STDs, 1999.
11. Beksinska ME, Mullick S, Kunene B, et al. A case study of antenatal syphilis screening in South Africa. Sex Transm Dis 2002; 29:32–37.
12. Patel A, Moodley D, Moodley J. An evaluation of on-site testing for syphilis. Trop Doct 2001; 31:79–82.
13. Delport SD. On-site screening for maternal syphilis in an antenatal clinic. S Afr Med J 1993; 83:723–724.
14. Van Dyck E, Van de Velden L, Ndoye I, et al. Evaluation of the RPR ‘teardrop’ card test for screening for syphilis in field conditions. Sex Transm Dis 1993; 20:194–197.
15. Bique Osman N, Challis K, Folgosa E, et al. An intervention study to reduce adverse pregnancy outcomes as a result of syphilis in Mozambique. Sex Transm Infect 2000; 76:203–207.
16. Fehler HG, Ballard RC. Pilot study to evaluate the feasibility of on-site RPR screening at antenatal and dedicated sexually transmitted disease clinics in South Africa. South Afr J Epidemiol Infect 1998; 13:22–25.
17. West B, Walraven G, Morison L, et al. Performance of the rapid plasma reagin and the rapid syphilis screening tests in the diagnosis of syphilis in field conditions in rural Africa. Sex Transm Infect 2002; 78:282–285.
18. Blandford JM, Gift TL, Dlali P, et al. Cost-effectiveness of on-site antenatal screening to prevent congenital syphilis in rural Eastern Cape Province, Republic of South Africa. Sex Transm Dis. In press.
19. Gloyd S, Chai S, Mercer MA. Antenatal syphilis in sub-Saharan Africa: Missed opportunities for mortality reduction. Health Policy Plan 2001; 16:29–34.
20. McDermott J, Steketee R, Larsen S, et al. Syphilis-associated perinatal and infant mortality in rural Malawi. Bull World Health Organ 1993; 71:773–780.
21. South African Demographic and Health Survey. MACRO International, 1998.
22. Myer L, Wilkinson D, Lombard C, et al. Impact of on-site testing for maternal syphilis on treatment delays, treatment rates, and perinatal mortality in rural South Africa: A randomised controlled trial. Sex Transm Infect 2003; 79:208–213.
23. Montoya PJ, Lukehart SA, Brentlinger PE, et al. Comparison of the diagnostic accuracy of a rapid immunochromatographic test and the rapid plasma reagin test for antenatal syphilis screening in Mozambique. Bull World Health Organ 2006; 84:97–104.