Vaginal lactobacilli and preterm birth

Objective: To assess the relationship between the absence of vaginal lactobacilli and preterm birth at < 33 weeks of gestation. Methods: A prospective study of the vaginal flora in the second trimester was undertaken in 1958 women with singleton pregnancies. The contribution of various microorganisms to preterm delivery was analyzed using a multivariate-logistic regression model. Results: Lactobacillus species were not cultured from 28% of 118 women who delivered at < 33 weeks, 10% of 224 women who delivered between 33 and 36 weeks, and 5% of 1616 women who delivered at > 37 weeks of gestation. Lactobacilli (odds ratio and 95% confidence interval: 0.15 [0.09 to 0.24]), Mycoplasma hominis (2.3 [1.0 to 5.4]), and glucose non-fermentative gram-negative rods (2.1 [1.0 to 4.2]) were identified as independent risk factors for preterm delivery at < 33 weeks of gestation. Absence of lactobacilli (sensitivity and positive predictive value: 28% and 25%) was a better predictor of preterm delivery at < 33 weeks of gestation than the presence of Mycoplasma hominis (7% and 13%, respectively) or glucose non-fermentative rods (9% and 11%). Conclusions: Although this was not a cohort study, results suggest that tests for determining the presence of vaginal lactobacilli may be clinically useful tools for identifying women at an increased risk of preterm delivery at < 33 weeks of gestation.