The primary aim of this study is to determine the impact of community-based screening and treatment of abnormal vaginal flora and urinary tract infections in early pregnancy (13-19 weeks) on preterm live birth in Sylhet district, Bangladesh.

Hypothesis 1: Community-based screening and treatment of abnormal vaginal flora (Nugent score >4) and urinary tract infections in early pregnancy (13-19 weeks) will reduce the population rate of preterm live birth by at least 15%.

The secondary aims of this study are:

• To determine the impact of community-based screening and treatment of abnormal vaginal flora and urinary tract infections on the:

  • proportion of pregnancies with outcomes occurring prior to 37 weeks (late miscarriage, preterm still birth and preterm live birth); and
  • proportion of babies with early onset neonatal sepsis.

• To determine the prevalence of abnormal vaginal flora and urinary tract infections, including asymptomatic bactiuria, among pregnant women in Sylhet district, Bangladesh.

• To evaluate the accuracy of simple, low-cost, point of care diagnostic tests for detecting bacterial vaginosis and urinary tract infections by community health workers in a rural, developing country setting.

Burden of Maternal Genitourinary Tract Infections: Genital and urinary tract (GU) infections may be due to endogenous or sexually transmitted pathogens, and are estimated to affect up to 41% of women of reproductive age globally, although there is wide regional, country, and population specific variation (Table 1). These estimates, however, may underestimate the burden in developing countries, as 60-80% of GU infections are asymptomatic in pregnant women [1], and furthermore many women never receive appropriate medical care in resource poor settings.

Several epidemiologic studies have been conducted in Bangladesh reporting the prevalence of GU infections in different populations [2-15]. Genital tract infections, particularly sexually transmitted infections are common among urban and high risk populations, i.e. commercial sex workers. In rural Bangladesh, the burden of diseases associated with bacterial vaginosis (BV) (5.9-18.9%) [3, 4] and asymptomatic bactiuria (12%) are high [12]. In urban areas, mostly Dhaka, the prevalence of BV is higher; one study reported a prevalence of 28% [8]. There is no known data on the prevalence of intermediate vaginal flora in Bangladesh. The prevalence of sexually transmitted infections including Gonorrhea, Chlamydia, Trichomonas and Syphilis are high among high risk urban populations of Bangladesh but generally low in rural areas. Given these prevalence data, we postulate that among pregnant women of rural Sylhet district in Bangladesh, BV and UTI are the most prevalent GU infections.

Rationale for screening and treating asymptomatic women The rationale for treating asymptomatic bacterial vaginosis (Nugent score 7-10) and asymptomatic intermediate flora (Nugent score 4-6) is based on data: 1) showing their association with adverse pregnancy outcomes[38], and 2) several promising trials showing that treatment may reduce preterm birth[39, 40]. Up to 84% of bacterial vaginosis cases are asymptomatic [23]. In a meta-analysis of 32 studies in developed and developing countries by Leitich et al, asymptomatic BV (Nugent score 7-10) was associated with a 6.32 times elevated risk of late miscarriage (95% CI 3.65-10.94) and 2.16 times (95% CI 1.56-3.00) increased risk of preterm birth. The association with preterm birth was higher when BV was detected in early pregnancy (<16 weeks, OR 2.97, 95% CI 1.48-5.98)[38]. Among women with a prior history of preterm birth, Hauth and colleagues found that screening and treatment of asymptomatic BV (Nugent score 7-10) with metronidazole and erythromycin at 22 weeks gestation significantly reduced the incidence of preterm birth from 46% in the placebo group to 31% in the treatment group [41]. In the multi-center NICHD BV trial, 1953 women with asymptomatic BV between 16-24 weeks of gestation were randomized to receive two doses of metronidazole (2g) or placebo; however, treatment did not significantly affect preterm delivery or other adverse perinatal outcomes [42].

Intermediate vaginal flora (Nugent score 4-6) is a heterogeneous condition which has been also associated with elevated risk of preterm birth and neonatal infections [43-46]. Intermediate vaginal flora comprises 15% of all abnormal vaginal flora (Nugent score >=4) [39]. In a recent trial, Ugwumadu et al reported that early (12-22 weeks of gestation) screening and treatment for abnormal vaginal flora (Nugent score >4) with 5 days of oral clindamycin resulted in a significant reduction in spontaneous preterm birth rate (12% in placebo vs. 5% in treatment group) and late miscarriage (13-24 weeks; 4% in placebo vs. 1% in treatment group) [39]. Similarly, Lamont et al reported that early (13-20 week) treatment of abnormal vaginal flora (Nugent score >4) with intravaginal clindamycin reduced the incidence of preterm birth by 60% [40]. Potential explanations for the treatment effect in the 2 later trials may include: 1) the earlier timing of treatment, prior to the amniotic membranes sealing the uterus at 20 weeks [47], which may thus prevent early ascension of bacteria into the intrauterine cavity; 2) antibiotic choice: 5-7 day course of clindamycin, which has greater activity against Mobiluncus and atypical Mycoplasma species vs. 2 days of metronidazole [26]; and 3) treatment of abnormal vaginal flora in Ugwumadu et al and Lamont et al, vs. treatment of BV only in the NICHD trial. A Cochrane meta-analysis concluded that the risk of preterm birth was significantly reduced by treatment of abnormal vaginal flora (Nugent score >4) (2 trials, 894 women; OR 0.51, 95% CI 0.32-0.81). Thus, in low-resource settings such as in rural Bangladesh, where both BV and preterm birth are prevalent, treatment of abnormal vaginal flora in early pregnancy may hold promise in reducing the incidence of preterm birth, and an evaluation in well-conducted community-based randomized trials is needed.

Inclusion Criteria:

All women who become pregnant in the study area will be eligible to participate in the trial. We will identify women with missed periods through monthly surveillance by health workers and identify pregnant women by positive pregnancy test conducted by CHWs. Thus enrollment may begin as soon as 5 weeks gestation and will continue until 19 weeks gestation. A sample of women who are enrolled early in gestation (<12 weeks), will be considered for enrolment in the gestational age sub-study. The first CHW visit for screening for abnormal vaginal flora and urinary tract infection will occur after the initial enrollment between 13-19 weeks. Their infants will enrolled as well.

Exclusion Criteria:

Subjects will be excluded from the study if they have no recall or uncertain report of LMP (due to lactational amenorrhea, recent discontinuation of contraceptive or irregular menses), history of irregular bleeding due to injectable depoprovera, or history of severe chronic disease based on their self-reported history on a medical history checklist. For women reporting a missed period during pregnancy surveillance, the CHW will perform a urine pregnancy test to confirm the pregnancy and keep a log of all positive pregnancy tests and LMP.

Pregnant women who are uncertain or refuse to participate at the initial visit will be allowed time to consider and enroll at the next study visit if they still meet inclusion criteria. Women will be given information regarding the study and CHWs will answer questions for both themselves and the family. If a woman is not certain whether she would like to participate, she will be allowed 1 week to consider participation, and the CHW will return in 1 week to revisit participation and answer any additional questions. If the woman refuses to participate at any visit, she will no longer be approached to participate.

Families can withdraw from the study at any time, and study staff will be available to answer questions families may have at any time.

Open-label