Screen-and-Treat for Bacterial Vaginosis Fails to Reduce Preterm Births

Molecular screening for and treatment of bacterial vaginosis (BV) did not significantly reduce preterm birth rates compared with usual care, according to the randomized AuTop trial out of France.

In an intention-to-treat analysis involving over 6,600 women, the rate of preterm birth before 37 weeks’ gestation was 3.8% in the screen-and-treat group and 4.6% in the control group (risk ratio [RR] 0.83, 95% CI 0.66-1.05, P=0.12), reported Florence Bretelle, MD, PhD, of La Conception Hospital, Assistance Publique Hopitaux de Marseille in France, and co-authors.

There were also no significant differences in total costs, they noted in JAMA Pediatrics.

On average, the intervention cost €203.60 (U.S. $218) per patient, and the average total cost was €3,344.30 (U.S. $3,580.50), while the cost of usual care on average was €3,272.90 (U.S. $3,504.10).

Notably, a post hoc subgroup analysis showed that the screen-and-treat strategy was significantly more effective than usual care for nulliparous women (RR 0.62, 95% CI 0.45-0.84, P for interaction=0.003). However, there was not a significant difference among multiparous participants (RR 1.30, 95% CI 0.90-1.87).

“BV increases the risk of preterm birth from 2-fold to 7-fold according to the gestational age at diagnosis,” Bretelle and team wrote, adding that there is higher risk with earlier gestational ages. Preterm birth affects about 5% to 11% of births worldwide.

Cynthia Gyamfi-Bannerman, MD, department chair of obstetrics, gynecology, and reproductive sciences at UC San Diego Health, told MedPage Today that this study confirmed what ob/gyns already knew: “that a general strategy of screening and treating was not effective at reducing preterm birth, even using newer molecular technology to appropriately diagnose BV.”

“This study uses updated diagnostic criteria to identify and treat BV, showing a possible reduction in preterm delivery in the subgroup of nulliparous participants,” she noted, calling this the “most noteworthy finding in this study,” which “requires further study as it was not the primary question addressed.”

Bretelle and colleagues also noted that future research should further evaluate nulliparous and high-risk multiparous women specifically.

For the open-label superiority AuTop trial conducted from March 2015 to December 2017 across 19 perinatal centers in France, the researchers included 6,671 low-risk pregnant women prior to 20 weeks’ gestation who did not have previous preterm births or late miscarriages. Both parous and nulliparous women were eligible. Exclusion criteria included extrauterine pregnancies, nonprogressive pregnancies, and use of antibiotics during the previous week. Mean age was 30.6 years, and mean gestational age was 15.5 weeks.

The women were randomized in a 1:1 ratio to the intervention, undergoing molecular screening and treatment for BV using self-collected vaginal swabs, or to usual care, receiving no screening. The molecular diagnostic tool used quantitative real-time polymerase chain reaction assays to quantify the DNA levels of Atopobium vaginae and Gardnerella vaginalis.

There were no significant differences in secondary outcomes between groups, including preterm birth rates before 26, 28, and 32 weeks’ gestation; premature rupture of membranes; endometritis; length of hospital stay; and intrauterine growth restriction.

Bretelle and team noted that it wasn’t possible to make the study completely blinded because women with positive BV test results had to receive treatment, including vaginal swabs, because it would be unethical to not treat everyone who had BV. However, 44 of 242 patients who had BV didn’t receive treatment. “This reflects the difficulties of a randomized trial nested within routine practice in the care of pregnant women and may lower the effectiveness of the screen and treat strategy,” they wrote.

Gyamfi-Bannerman also had a few reservations about the study.

“The investigators had many prespecified secondary outcomes, but the analysis comparing nulliparous with multiparous participants was done post hoc, limiting the conclusions that can be drawn from those findings,” she said. “Further, there are many important secondary outcomes, like, for example, the incidence of PPROM [preterm premature rupture of membranes], which the sample size is too small to draw conclusions. It raises the question of whether the findings would have been different in a U.S. population, where the incidence of preterm birth is double.”