Table 11 Impact of syphilis testing and treatment on stillbirth and perinatal mortality

Walker et al. 2001 [124]

None.

Meta-analysis (Cochrane). 29 RCTs and quasi-RCTs reviewed, 0 studies included.

Assessed the impact of antibiotic treatment for syphilis during pregnancy.

No data, 0 eligible studies.

Intervention studies

Bique et al. 2000 [130]

Mozambique, ANC clinics.

Case-control study. 4 suburban ANC clinics (2 intervention clinics, 2 control clinics).

Pregnant women (N = 929; N = 453 intervention, N = 476 controls) with positive RPR.

Compared the impact of an intervention offering RPR testing with immediate on-site treatment of seropositive cases with 2.4 million IU benzathine penicillin by specially trained nurse-midwives. The study offered to treat partners of intervention group as well. Controls were offered routine syphilis screening protocol requiring testing at a separate lab, return visits, and payment for treatment (< 50% compliance), with no partner treatment option.

PMR: 13/1000 vs 34/1000 in intervention vs. control groups, respectively (P = 0.03).

Fetal death (miscarriage+SB): 13/1000 vs. 26/1000 (P = 0.159) [NS]

Donders et al. 1997 [128]

South Africa (Pretoria), hospital-based.

Prospective cohort study. HIV-, RPR+ black urban pregnant women (N = 212; N = 135 received ≥ 1 injection)

Assessed dosage impact of 0–3 weekly IM injections of benzathine penicillin G on perinatal outcomes.

PMR secondary to congenital syphilis: RR = 20.5 (95% CI: 2.3–184, P = 0.0015).

[0 injections: 7/55 (13%)

1 injection: 2/19 (11%)

2 injections: 1/24 (4.2%)

3 injections: 2/82 (2.4%)]

Adjusted when Treponemicidal coverage was < 3 wks

Myer et al. 2003 [123]

South Africa (Hlabisa district, rural KwaZulu Natal), PHC clinics.

Cluster RCT. 7 pairs of clinics. Pregnant women (N = 549).

Compared the impact of on-site syphilis testing complemented by laboratory confirmation vs. laboratory testing alone.

PMR: adj. RD: -0.9%; 95% CI: (-) 4.4-2.7, P = 0.31) [NS]

[33/1000 (18/549) vs. 51/1000 (8/157) in intervention clinics vs. control clinics, respectively.]

Rotchford et al. 2000 [129]

South Africa (Hlabisa district, rural KwaZulu Natal), ANC clinics.

Cluster RCT. 12 clinics. Pregnant women (N = 1783) screened for syphilis (N = 158 RPR+; 9% prevalence) at first ANC visit (mean: 24 wks); RPR+ women followed for pregnancy outcome (data available for N = 142 (90%); N = 30 had no treatment; N = 96 had all 3 doses penicillin)

Assessed impact on PMR of inadequate maternal syphilis treatment in presence of adequate screening.

PMR: 15/142

[0 or 1 dose penicillin: 11/43 (260/1000)

≥ 2 doses penicillin: 4/99 (40/1000)]

Dose-response relationship observed. (P = 0.0001)

PMR risk reduction:

1 dose: 41% reduction

2 doses: 65% reduction

3 doses: 79% reduction

Watson Jones et al. 2002 [127]

Tanzania (Mwanza), ANC clinic.

Case-control study. Pregnant women (N = 1688; N = 133 high-titre [RPR titre ≥ 1:8, TPHA/FTA+]; N = 249 low-titre [RPR titre < 1:8, TPHA/FTA+], N = 950 seronegative controls).

To examine the effectiveness of treatment for maternal syphilis with single-dose IM benzathine penicillin (2.4 million units).

Birth outcomes were compared SBR: 23/1000 vs. 25/1000 in treated high-titer women vs. seronegative women.

LBW: 6.3% vs. 9.2% in treated high-titer women vs. seronegative women.

Adverse pregnancy outcome (combined SBR+LBW): OR = 0.76 (95% CI: 0.4–1.4) [NS].

Observational studies

Delport et al. 1993 [199]

South Africa (Pretoria), ANC clinic.

Descriptive study. Kalafong Hospital. Pregnant women (N = 1237) attending ANC.

Assessed the sensitivity, specificity, negative and positive predictive values of the RPR test at ANC compared with gold-standard laboratory Treponema pallidum haemagglutination test.

RPR test:

Sensitivity: 92.8%

Specificity: 96.3%

Negative predictive value: 99.5%

Positive predictive value: 64.7%.

Guinness et al. 1988 [125]

Swaziland (Mbabane), Public health unit

Prospective cohort study. Pregnant women (N = 283) tested at ANC enrollment: N = 37 (13.1%) TPHA+ and RPR+; N = 87 (30.7%) TPHA+ and RPR-.

Assessed the impact of antenatal screening on perinatal mortality attributable to syphilis. Mothers were tested prenatally and again at delivery; prenatal test found to have sensitivity = 36% and predictive accuracy = 48%.

PMR (untreated active syphilis): 219/1000 (7/32).

12/172 seronegative women had active syphilis (late seroconversion or false negative prenatal test results): 4/12 experienced perinatal death.

PMR: 46/1000 (4/87) vs. 28/1000 (4/415) in TPHA+/RPR- vs. syphilis-seronegative women.

Screening reduced expected syphilis-attributable PMR from 3.5% to 2.3% (65% of mothers with active syphilis missed treatment; sexual partners were not treated).

Temmerman et al. 2000 [126]

Kenya (Nairobi), maternity hospital.

Prospective case control study. Women (N = 12414) delivering at Pumwani Hospital were RPR tested (3%, N = 377 were RPR-positive). TPHA testing confirmed syphilis infection (N = 296). Equal numbers of seronegative women also enrolled; records examined for syphilis testing and treatment during pregnancy.

Assessed the impact of an antenatal syphilis control programme on pregnancy outcome.

Adverse obstetric outcome (LBW or SB): OR = 4.1 (95% CI: 2.3–7.5, P < 0.001).

[22.5% vs. 6.6% in untreated syphilis-positive vs. uninfected mothers, respectively.]

LBW: OR = 4.0 (P < 0.0001) in untreated syphilis-seropositive mothers vs. uninfected mothers, respectively.

SBR: OR = 3.3 (P = 0.028) in untreated syphilis-seropositive mothers vs. uninfected mothers, respectively.

OR = 2.5 in treated syphilis-seropositive mothers vs. uninfected mothers, respectively (P < 0.05).