The incidence of dengue worldwide is increasing rapidly. A better understanding of dengue transmission may help improve interventions against this major public health problem. The virus is mostly transmitted by vectors. There are, however, other modes of transmission, notably mother-to-child transmission or vertical transmission. We studied a prospective cohort of 54 women who had dengue while pregnant during the 2012-2013 epidemic in French Guiana to estimate the mother-to-child transmission rate and assess the clinical and biological presentation of neonatal dengue. The rate of vertical transmission was between 18.5% (95% confidence interval [CI]: 9.25-31.4) and 22.7% (95% CI: 11.5-37.8), depending on the calculation method used. Mother-to-child transmission occurred both in early and late pregnancy. There were 52 births, including three newborns who presented neonatal dengue with warning signs requiring platelet transfusion. This quantification of the mother-to-child transmission of dengue highlights three points: first, vertical transmission of dengue is not negligible; second, it is more frequent when maternal dengue occurs late during pregnancy near delivery; and third, reliable diagnostic tests must be used to allow the diagnosis of vertical transmission. Our findings indicate that if there is a known history of maternal dengue during pregnancy, or if there is fever during the 15 days before term, cord blood and placenta should be sampled after delivery and tested for the virus, and the newborn should be closely monitored during the postpartum period.

Dengue fever is an increasing problem worldwide, but consequences during pregnancy remain unclear. Much of the available literature suffers from methodological biases that compromise the validity of clinical recommendations. We conducted a matched cohort study during an epidemic in French Guiana to compare events and pregnancy outcomes between two paired groups of pregnant women: women having presented with symptomatic dengue during pregnancy (n = 73) and women having had neither fever nor dengue during pregnancy (n = 219). Women in each arm were matched by place of follow up, gestation weeks at inclusion, and place of residence. Dengue infection was considered to be confirmed if viral RNA, N S1 antigen, the seroconversion of IgM antibodies or the presence of IgM was detected in collected samples. According to the 2009 WHO classification, 27% of the women with symptomatic dengue had at least one clinical or biological warning sign. These complications occurred after the 28th week of gestation in 55% of cases. The medical history, socioeconomic status and demographic characteristics were included in multivariate analysis. Exposure to dengue during pregnancy was not significantly associated with prematurity, small for gestational age infants, hypertension or emergency caesarian section. Maternal dengue with warning signs was a risk factor for peripartum hemorrhage with adjusted relative risk = 8.6(95% CI = 1.2–62). There was a near significant association between dengue and in utero death (p = 0.09). This prospective comparative study underlined the importance of taking into account potential confounders between exposure to dengue and the occurrence of obstetrical events. It also confirms the need for increased vigilance for pregnant women with dengue, particularly for women who present with severe dengue.

Background In French Guiana, gold miners working illegally represents a major reservoir of malaria. This mobile population, mainly of Brazilian descent, enters the French Guianese forest from neighbouring countries, Suriname and Brazil. A complex and innovative intervention was piloted as a cooperation with the three involved countries involved to control malaria in this specific population. The principle was that health workers called “facilitators” provide the participants with a self-diagnosis and self-treatment kit along with adequate training and material to rapidly manage an episode of malaria symptoms on their own, when they find themselves isolated from health care services. Methods This paper describes the design, development, content of the intervention and players’ organization of this multi-country project, the opportunities and constraints encountered, and the lessons learnt at this stage. Results The choice not to implement the usual “Test and Treat” approach within the community is mainly driven by regulatory reasons. The content of medical messages tends to balance the tension between thoroughness, accuracy and efficacy. The wide range of tools developed through a participatory approach was intended to cope with the challenges of the literacy level of the target population. Despite the difficulties encountered due to language, regulation differences and distance between partners, cooperation was fruitful, due to the complementary of stakeholders, their involvement at all important stages and regular face-to-face meetings. Discussion and conclusion This experience shows the feasibility of an ambitious project of action-research in a border malaria context, involving several countries and with a mobile and undocumented population. It reveals some factors of success which may be transferable in analogous settings.