3 Apr, 13 | by Leslie Goode, Blogmaster
Congenital syphilis (syphilis transmitted from mother to child (MTCT)) remains a scourge in many low- and middle- income countries (LMIC) causing stillbirth and neonatal death. This is despite the existence of inexpensive, cost-effective and feasible testing and treatment – and despite the fact that a majority of pregnant women, even in LMIC, attend antenatal clinics (ANC), where such testing and treatment could, in principal, be offered. A recent multi-country study, for example, found levels of testing of 1.7% and 17.8% respectively in rural ANCs in Uganda, and 51 health facilities in Tanzania (http://blogs.bmj.com/sti/2012/06/25/sustained-health-systems-strengthening-holds-the-key-to-prevention-of-mother-to-child-transmission-of-syphilis/). The year 2007 saw the launch of a WHO initiative for the Global Elimination of Congenital Syphilis, with the goal that by 2015 ≥90% of pregnant of women should be tested.
One complicating factor of any public health policy where syphilis is concerned is the difficulty of establishing definitive diagnosis. This complicates the task of quantifying MTCT and assessing progress in eliminating it. It also makes it difficult to evaluate screening tools, such as the relatively recent point-of-care tests (POCT) that promise to extend effective diagnosis to countries without a robust laboratory infrastructure. In both cases – the assessment of the scale of MTCT, and evaluating screening tools – researchers have had to find a way round the absence of a good reference standard. Two recently-published papers have addressed themselves to these problems. One is an analysis of global surveillance data (Newman & Broutet) using modelling to estimate total global adverse outcomes from syphilis, and the impact on these outcomes, of testing in ANCs for the year following launch of the WHO initiative (2008) (http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.1001396). The other (Jafari and Pai) addresses itself to evaluating the diagnostic accuracy of POCT, employing the technique of Bayesian analysis to get round the problem of an inadequate reference standard for syphilis (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0054695).
Newman & Broutet estimate global adverse events for 2008 at 521,000, including 212,000 stillbirths and 92.000 neonatal deaths. This provides a baseline for evaluating subsequent progress. Of the adverse events, 66% are estimated to involve women who attended ANCs – which indicates the potential improvements that might be achievable through ANCs. As things were in 2008, testing and treatment by ANCs was reckoned to have averted a quarter of the adverse events that might otherwise have been expected; the model predicts that achievement in 2008 of the 2015 WHO target for testing in ANCs would have eliminated over half of these events.
So much then for the potential for improvement. Such future developments will be dependent on the success of POCTs, which promise both an inexpensive form of screening without the need for laboratory infrastructure, and the possibility of same day treatment for ANC attendees. There remains the question of their effectiveness compared with traditional methods of diagnosis. Here again the difficulty of establishing definitive diagnosis has been a problem. The absence of a reference standard has, according to Jafari & Pai, impeded the comparative evaluation of POCTs. A recent meta-analysis of electronic databases using a Bayesian hierarchical latent class model has attempted to put a figure on sensitivity and specificity of four globally used POCTs with both serum and whole blood samples. These estimates are comparable of those for lab-based testing: with serum (sensitivity/specificity) Determine 90.04/94.15; SD Bioline 87.06/95.85; VisiTect 85.13/96.45; Syphicheck 74.48/99.14; with whole blood Determine 86.32/95.85; SD Bioline 84.5/97.95; VisiTect 74.26/99.43; Syphicheck 74.47/99.58. As it turns out, these estimates are close to those of Mabey, Zheng et al., using treponemum pallidum haemagglutination assay or fluorescent treponemal antibody absorbed as a reference standard (http://sti.bmj.com/content/82/suppl_5/v13.full.pdf+html).
POCTs have also been shown to be cost effective (http://sti.bmj.com/content/82/suppl_5/v38.abstract?sid=82a53292-0a5a-46d1-a01d-9cf6bbade9ae). So a substantial reduction of syphilis related adverse events seems achievable even in LMICs – if they can meet the training and organizational challenges to deployment of POCTs in ANCs. According a recent paper of Mabey & Peeling, it is this latter area of health service organization that we find the greatest obstacle to delivering the promise of POCTs (http://blogs.bmj.com/sti/2012/06/25/sustained-health-systems-strengthening-holds-the-key-to-prevention-of-mother-to-child-transmission-of-syphilis/).