Objective: To determine the prevalence of bacterial vaginosis (BV) in the second trimester of pregnancy in a Danish population using the Schmidt criteria and to examine whether BV was associated with subsequent preterm delivery, low birthweight or perinatal infections.
Design: Prospective cohort study.
Setting: Department of Obstetrics and Gynaecology at a University Hospital, Denmark.
Population: Three thousand five hundred and forty pregnant women aged 18 years or more.
Methods: A smear from the vagina was obtained from all women, air-dried and stored for subsequent diagnosis of BV. After rehydration with isotonic saline, the smear was examined in a phase-contrast microscope at 400x, and the numbers of lactobacilli morphotypes and small bacterial morphotypes were counted. A score for BV was calculated according to the method described by Schmidt. The outcome of pregnancy from 20 weeks of gestation was examined in the 3262 singleton pregnant women who were included in this study before 20 weeks of gestation. The relationship between BV and adverse outcome of pregnancy was examined by univariate and multivariate analyses.
Main outcome measures: Prevalence of BV, preterm delivery (<37 weeks), low birthweight (<2500 g), preterm delivery of a low-birthweight infant and clinical chorioamnionitis.
Results: The prevalence of BV was 16%, and the rate of preterm delivery was 5.2% in the study population of 3262 singleton pregnant women who were included before 20 weeks of gestation. Mean birthweight was significantly lower in infants of women with BV than in infants of women without BV (3408 versus 3511 g, P < 0.01). Univariate analyses showed that BV was marginally associated with preterm delivery but significantly associated with low birthweight, preterm delivery of a low birthweight infant, indicated preterm delivery and clinical chorioamnionitis. Multivariate analyses, which adjusted for previous miscarriage, previous preterm delivery, previous conisation, smoking, gestational diabetes, fetal death and preterm premature rupture of membranes, showed that BV was significantly associated with low birthweight (OR 1.95, 95% CI 1.3-2.9), preterm delivery of a low-birthweight infant (OR 2.5, 95% CI 1.6-3.9), indicated preterm delivery (OR 2.4, 95% CI 1.4-4.1) and clinical chorioamnionitis (OR 2.7, 95% CI 1.4-5.1).
Conclusions: The prevalence of BV determined using the Schmidt criteria in the early second trimester of pregnancy was similar to that found in similar studies. The presence of BV before 20 weeks of gestation was an independent risk factor for delivery of an infant with low birthweight, preterm delivery of a low-birthweight infant, indicated preterm delivery and clinical chorioamnionitis.