Bulletin of the World Health Organization

Untreated maternal syphilis and adverse outcomes of pregnancy: a systematic review and meta-analysis

Gabriela B Gomez, Mary L Kamb, Lori M Newman, Jennifer Mark, Nathalie Broutet & Sarah J Hawkes

Volume 91, Number 3, March 2013, 217-226

Table 1. Characteristics of studies included in a systematic review and meta-analysis to determine the frequency of adverse pregnancy outcomes (APOs) among untreated women with syphilis and women without syphilis

Study Country Study group,a design and follow-up n Syphilis prevalence among mothers (%) Syphilis diagnostic test Comment
Harman, 191716 United Kingdom Group 1: Study of historical data to assess the frequency of APOs among women whose reproductive history was recorded before the existence of syphilis treatment. The follow-up period was birth to several years after delivery (but was not defined). Cases: 1001; controls: 826 4.5b Wasserman test (assumed) This study assessed all pregnancy outcomes (over a lifetime) in a group of 150 women with syphilis and a group of 150 “healthy” women from a similar neighbourhood. Women had 1–18 pregnancies. The average was 7 pregnancies among women with syphilis. Timing of syphilis was not reported: some APOs may have preceded T. pallidum infection and some may have occurred many years after infection. The study did not differentiate between stages of syphilis.
Wammock, 1945–4817 United States of America Group 2: Retrospective study to assess the frequency of APOs among women attending an ANC clinic in which syphilis treatment was not available. The follow-up period was birth to ≥ 28 days after delivery. Cases: 61; controls: 5596 1.5c Kahn test This study assessed best practices in providing penicillin to pregnant women to prevent congenital syphilis and was done shortly after penicillin became available. Background data on APOs among untreated women who were seroreactive for T. pallidum infection were reported; uninfected women (historical controls) were used as a comparison group. Of 61 cases of early syphilis, 7 were “symptomatic” (which seems to correspond to primary or secondary infection) and resulted in no “normal living infants” (i.e. 2 stillbirths, 1 neonatal death and 4 infants with congenital syphilis). The remaining 54 cases had “early latent” syphilis (defined as “symptomless” seropositivity for T. pallidum infection of < 4 years’ duration): these cases are combined with the 61 early cases in the meta-analysis. There were 14 cases with “late latent syphilis” (defined as symptomless infection of > 4 years’ duration). Of these, 1 had a stillbirth and 13 had “normal living infants”.
Ingraham, 1940–4918 United States of America Group 2: Study to assess the frequency of APOs among women attending an ANC clinic in which treatment was not available. The follow-up period was birth to ≥ 60 days after delivery and, for at least 70% of the cohort, lasted for 6 months after delivery. Cases: 220; controls: 10 323 1.5c Kahn test (assumed) This study described pregnancy outcomes among asymptomatic women with and without syphilis at two public hospitals in Philadelphia, United States of America. Asymptomatic women consisted of subjects who were seroreactive for T. pallidum infection and either untreated, treated with arsenicals or treated with various doses of penicillin. Subjects were followed for at least 60 days after delivery (77% were followed for 6 months). On the basis of clinical history, 220 of 302 women had “early” syphilis (defined as untreated syphilis of < 4 years’ duration) and 82 had “late” syphilis (defined as untreated syphilis of > 4 years’ duration). APOs were stillbirths for 10, neonatal deaths for 7, prematurity for 2 and congenital syphilis for 2; 61 women had “normal full-term infants”.
Hira et al., 19906 Zambia Group 3: Study to assess the frequency of APOs among women attending an ANC clinic in which recruitment was at the time of delivery. There was no follow-up after delivery. Cases: 230; controls: 2647 6.5d RPR, with FTA-Ab confirmation This study tested the effectiveness of an intervention to reduce APOs due to syphilis. The intervention was implemented in the context of an existing screening and treatment programme for syphilis and involved new health education methods and prenatal screening for syphilis. Overall, 8.0% of women were confirmed to be seroreactive for T. pallidum infection and there was no difference in seroprevalence between the intervention centres and the non-intervention centres. Uptake of screening and treatment at the intervention centres was suboptimal.
McDermott et al., 1987–9019 Malawi Group 2: Study to assess the frequency of APOs among women attending an ANC clinic in which treatment was not available because routine syphilis screening was not performed. The follow-up period was birth to ≥ 1 year after delivery. Cases: 130; controls: 2890 3.6 VDRL or RPR, with MHA-TP confirmation In a study of APOs due to malaria, women were enrolled prospectively from four rural ANC clinics. All subjects received “routine ANC care”, including monthly visits, tetanus toxoid vaccination, iron supplementation and malaria chemoprophylaxis, but routine syphilis screening was not provided through the ANC clinic system. Women were followed monthly and had blood drawn for later analysis of malaria outcome.
Watson-Jones et al., 1998–200020 United Republic of Tanzania Group 3: Retrospective study of data from antenatal health cards in which recruitment was at the time of delivery. There was no follow-up after delivery. Cases: 73; controls: 233 8.0 RPR, with FTA-Ab or TPHA confirmation “Relatively representative” pregnant women in urban and rural settings were recruited at delivery and tested for syphilis. Unscreened women, defined as those who had no ANC care or who did not have records of syphilis screening on their ANC card (17%), were invited to participate in screening at delivery (participation rate, 51%). Two consecutive women with an unreactive T. pallidum serologic test were selected as controls for each case, defined as a woman with high- or low-titre T. pallidum infection. The study assessed APOs in 73 women with high-titre T. pallidum infection (defined as an RPR titre of ≥ 1:8 and a positive result of a confirmatory test), 27 women with low-titre T. pallidum infection (defined as an RPR titre of < 1:8 and a positive result of a confirmatory test), 9 women with past or currently treated syphilis (defined as a negative result of an RPR test and a positive result of a treponemal test) and 233 women who were seronegative for T. pallidum infection. As soon as possible after delivery, treatment was administered to women with high- or low-titre T. pallidum infection and their infants. Study authors defined “stillbirth” (i.e. dead fetus > 22 weeks’ gestation), “intrauterine fetal death” (i.e. fetal death ≤ 22 weeks’ gestation) and “low birth weight” (i.e, < 2500 g). There were no intrauterine fetal deaths. Among unscreened women, the population attributable fraction of APOs due to high-titre T. pallidum infection was 51% for stillbirth, 24% for prematurity, 5% for intrauterine growth retardation and 12% for low birth weight. Overall, 17% of all APOs among unscreened women were attributable to syphilis.

ANC, antenatal care; ; FTA-Ab, fluorescent treponemal antibody absorption test; MHA-TP, microhaemagglutination assay for Treponema pallidum; ; RPR, rapid plasma reagin test; SB: stillbirth; T. pallidum, Treponema pallidum; TPHA, T. pallidum haemagglutination test; VDRL, Venereal Disease Research Laboratory test.

a Study groups were defined on the basis of study design, as follows: group 1, studies calculating the frequency of APOs among women whose reproductive history was recorded before the existence of penicillin treatment; group 2, studies in which the frequency of APOs was calculated for women attending an ANC clinic in which screening and treatment for syphilis were unavailable; and group 3, studies of the frequency of APOs among women attending ANC in which study recruitment was done at the time of delivery (and syphilis treatment was only available at that time).

b In Glasgow’s Royal Maternity & Women’s Hospital in 1922, 4.5% of umbilical cord samples had positive results on Wasserman tests.

c The rate of APOs was not directly reported in these papers but was 3.4% in Philadelphia’s largest delivery hospital during the same period.21

d Data are from the report by Powell et al.22