Expert Commentary

Focus on Gynecologic Infections: Clindamycin treatment of bacterial vaginosis reduces preterm deliveries

Author and Disclosure Information



To investigate whether treating abnormal genital tract flora with clindamycin vaginal cream in gravidas before 20 weeks’ gestation prevents preterm delivery.

Methods and results

This randomized, double-blind, placebo-controlled tricenter study included 409 women with abnormal genital tract flora presenting to antenatal care clinics at 13 to 20 weeks’ gestation. Infection or colonization consistent with bacterial vaginosis was defined as decreased lactobacilli and increased numbers of “other” bacterial morphotypes and was identified by Gram stain performed on secretions obtained from the upper portion of the vagina. Women who tested positive for infection or colonization were treated with a 3-day course of vaginal clindamycin cream (n = 208) or placebo (n = 201).

If infection or colonization persisted 3 weeks later, a second, 7-day course of the drug or placebo was given, in accordance with the original randomization.

Compared with controls, women treated with clindamycin had a statistically significant reduction in the incidence of preterm delivery (4% versus 10%, P = .03). Consequently, admission to the neonatal intensive care unit also was significantly reduced among babies born to women in the treatment group.

Expert commentary

Preterm delivery remains the bane of the obstetrician’s existence. Treatment of clinically evident preterm labor can delay delivery sufficiently to allow for administration of antenatal steroids, but only rarely is established labor abolished.

Given our limited effectiveness in combating premature labor, one alternative is identifying the woman at risk. Unfortunately, the majority of pregnancies complicated by preterm delivery have no obvious risk factors. The study by Lamont et al is important because it describes a means of identifying and successfully treating infection that might otherwise remain undiagnosed until preterm labor becomes established and essentially untreatable. Indeed, the essence of a good screening method is its ability to identify risk in those who exhibit no ostensible risk—that is, the population at large.

While this study is one of many to consider the role of bacterial vaginosis in preterm labor, the use of a Gram stain to identify the abnormal bacterial morphology is clever and deserves consideration. Once risk is identified, the next logical step is finding a means to facilitate its reduction—and the study succeeds here as well. If the risk of preterm delivery can be suitably diminished—as it was in the women given clindamycin—the potential to lower the preterm delivery rate is greater than with traditional interventions for clinically apparent preterm labor.

My practice is inner city, where preterm deliveries occur for a variety of reasons and the degree of prematurity is on the severe end of the scale. Thus, an approach that clearly lowers admissions to the neonatal intensive care unit would be valuable. In my opinion, this approach is worth a trial.

Bottom line

In the low-risk population studied here, identifying infection by Gram stain and treating it with intravaginal clindamycin cream had a marked impact on the goal of reducing preterm delivery. This is an elegant application of a simple, direct, and inexpensive means to a most valued end.

Next Article: