Bacterial vaginosis and infection with Trichomonas vaginalis have been associated with pre-term labor and preterm birth. Although the mechanism is unclear, it has been suggested that the infection may spread from the vagina to the uterus. Three meta-analyses have recommended that women at high risk for preterm birth be screened and appropriately treated for bacterial vaginosis and T. vaginalis infection. Nevertheless, these meta-analyses concluded that screening all pregnant women would not be beneficial, and the most recent Cochrane review cautioned that treatment of T. vaginalis with metronidazole (Flagyl) may increase the risk of preterm birth. Because more studies have been published since the meta-analyses, Okun and colleagues conducted a systematic review of the evidence to determine whether antibiotic treatment of bacterial vaginosis or T. vaginalis infection during pregnancy reduces the risk for preterm birth.

The authors searched electronic databases for randomized controlled trials of antibiotic treatment of bacterial vaginosis or T. vaginalis infection during the second or third trimester of pregnancy. All trials included preterm birth (i.e., earlier than 37 weeks’ gestation) as a primary outcome measure. Each study was reviewed independently by two reviewers. Of the 1,888 studies originally identified, only 14 met the criteria for inclusion. These studies generally were of high quality, with good randomization practices and less than 5 percent loss to follow-up in most cases, and all but one trial were placebo controlled.

Studies of bacterial vaginosis indicated that antibiotic therapy significantly reduced the risk of persistent vaginosis. Nevertheless, the pooled results from 11 studies showed no difference in the risk of preterm birth. The same result was found in a subgroup analysis of women at high risk of preterm birth and in women treated with metronidazole. Antibiotic treatment had no significant effect on other outcomes such as low birth weight, admission to a newborn intensive care unit, perinatal death, or maternal peripartum infection.

In the two studies of antibiotic treatment for T. vaginalis infection, the pooled data indicated a significant reduction in persistent infection after therapy. However, treatment with metronidazole was associated with a significant increase in preterm birth. This risk also was found in subgroup analysis of high-risk mothers. No effect was found on very low birth weight or other significant outcomes.

The authors conclude that antibiotic therapy appears to reduce bacterial vaginosis but does not reduce the risk of preterm birth or other adverse outcomes. Similarly, they conclude that antibiotic treatment of T. vaginalis infection is not beneficial and that the use of metronidazole may be harmful. Killing the organisms may cause inflammation or release a virus that increases the risk of pre-term birth.