Syphilis continues to exert a high burden of disease globally 
, and in several parts of the world rates of congenital syphilis are rising 
. The extremely high rates of adverse outcomes of pregnancy associated with in-utero transmission can be substantially reduced through relatively simple and cost-effective interventions which are feasible to administer in most settings 
, include serological or whole blood screening, and treatment with penicillin for women (and their partners) found to be seropositive. Recently, WHO has called for the global elimination of mother-to-child transmission of syphilis, through commitment to and implementation of four strategic pillars – advocacy and political commitment; early access to maternal and neonatal health care; screening all pregnant women; surveillance, monitoring and evaluation 
. Pillar 2 of the WHO global strategy explicitly mentions a goal of “increasing the percentage of pregnant women attending….facilities early
in pregnancy” 
. However, to date, there has been no review of the effectiveness of early interventions to prevent mother to child transmission of syphilis.
This review represents the first systematic evaluation of the impact of screening and treating women early in pregnancy compared to later in pregnancy. We have defined “early” in its broadest terms – the first and second trimesters (i.e. up to and including week 27 of the pregnancy). This was mainly done for pragmatic reasons – antenatal care is often not sought till at least the second trimester in many settings 
, therefore we assumed a priori
that there would be insufficient data to allow us to analyse the impact of screening and treatment in the first trimester alone.
We identified five studies which compared women seeking and receiving interventions (screening and treatment for syphilis) in the first and second trimesters (reference group) compared to those women not seen until the third trimester 
. Overall prevalence of any adverse pregnancy outcomes was higher among women seen in the third trimester compared to the reference group for all studies – resulting in an odds ratio of 2.24 (for adverse outcomes among women seen later in their pregnancies). There was an important heterogeneity observed between studies, and results might have been affected by individual study designs. According to the results of our meta-analyses by sub-groups, the main characteristic that helped explain the heterogeneity observed was the type of outcome reported.
Screening and treatment earlier in the pregnancy had an impact, in general,on the risk of all adverse outcomes (OR
2.24 (95% CI 1.28, 3.93)), and in particular, on the risk of ‘congenital syphilis’ (i.e. an infant with evidence of infection; OR
2.92 (95% CI 0.66, 12.87)). However, only the studies reporting on congenital syphilis events 
were statistically similar and their estimates could be confidently pooled. Among those studies reporting on all adverse outcomes, there were some methodological differences. For instance, Watson-Jones et al 
included low birth weight, stillbirth, and preterm birth in their definition of all adverse outcomes, and presented detailed data on individual pregnancy outcomes. Zhu et al 
included congenital syphilis (i.e. an infected infant), foetal death, and neonatal death in their definition, and included mothers with primary, secondary, and tertiary syphilis in their analysis. Both of these studies might have underestimated the rate of adverse outcomes. Meanwhile, Carles et al 
looked at low birth weight, preterm birth, and intrauterine death in a population accessing services very late during the pregnancy or at birth.
Overall, there was an increased risk of prematurity in mothers presenting late to ANC, with an odds ratio of 2.09 (95% CI 1.09–4.00). For stillbirth, we found an odds ratio of 0.71 (95% CI 0.21–2.48; not significant), probably due to the small number of events observed (n
15 stillbirths overall, only three in women presenting to ANC during the third trimester).
As noted, antenatal care is now accessed at least once in pregnancy by a majority of women. This is a significant public health achievement and a reflection of the highest level of global commitment to ensure that all pregnant women can access the antenatal care they need 
, which has helped to mobilise resource allocation in this area. However, the global goals and commitments for antenatal care do not, at present, include a target for ensuring that women seek care early enough in pregnancy for interventions to be most effective.
We still know little about “what works” in terms of successfully enabling women to seek and receive antenatal care early in their pregnancies. As mentioned, women face numerous obstacles to early antenatal care, including cultural barriers 
, economic barriers 
, and misconceptions about the perceived and actual benefits of early antenatal care 
. We have little empirical or trial evidence on how to increase early antenatal care attendance. A non-randomised controlled trial conducted in Zambia in the late 1980s showed that a multi-pronged intervention including behaviour change messages aimed at whole communities as well as specific groups in those communities had the effect of increasing the percentage of pregnant women who sought ANC in the first trimester from 9.4% (68/723) to 42.5% (194/457) 
. Components of the successful intervention included provision of health information about the importance of early antenatal care to groups identified as sexually active (for example, women at family planning clinics, men and women at out-patient clinics), and to elderly community members who have a specific role as community leaders and influencers. Health education messages were developed in collaboration with local communities, and were given repeatedly. However, additional high quality (for example, randomised controlled trial) evidence is currently lacking in this area.
Our review has clearly shown that the timing of antenatal care interventions to prevent adverse outcomes due to syphilis makes a significant difference in outcome rates. Women who sought care in the first two trimesters of their pregnancy, and received the appropriate intervention, were more likely to have a healthy infant, than those who waited until the third trimester before seeking care. Such findings carry important implications for antenatal care programmes, and whether or not similar findings are observed with other antenatal care interventions (such as preventing mother to child transmission of HIV) needs further exploration.
Encouraging ALL pregnant women to seek care in the first two trimesters of their pregnancy to avoid preventable adverse outcomes should be a priority for health programmes. This is likely to be achievable, in part, with strengthening of health systems so that health workers are able to provide high quality comprehensive antenatal care for all women. Screening for syphilis and HIV during the first ANC visit is recommended. However, the timing of this visit in addition to whether these tests were performed and treatment received should be seen as indicators of high quality antenatal care. Early access to ANC will also be enhanced by community engagement programmes that work with women, their families and their communities to enable pregnant women to seek antenatal care as early as possible for interventions to be effective – the mechanisms and methods by which to achieve this deserve further attention.