O tratamento de vaginose bacteriana na gestação pode reduzir a incidência de trabalho de parto prematuro? : uma revisão sistemática e metanálise

Abstract

Preterm delivery is one of the most frequent obstetric complications, with up to 40% of cases appearing to be related to intrauterine infection, ascending via the lower genital tract. Among the cervicovaginal agents related to the onset of preterm labor (PTL), we must also mention those associated with bacterial vaginosis (BV). Thus, it is logical to consider the use of antibiotics in order to prevent prematurity. Many studies have been done to prove this hypothesis; however, they obtained contradictory results, likely due to methodological failures. Objectives: To evaluate if the treatment of pregnant women with BV reduces the incidence of PTL. Methods: The study consisted of a systematic review of the literature and a meta-analysis. Bibliographic research was carried out in the electronic databases: Pubmed, Bireme, Google Schoolar, Scopus, Web of Science, Science Direct, Scielo, Nature, New England Journal of Medicine, Repository Portugal, and Omics. Published studies were conducted between January 1, 1990 and December 20, 2017, using many designs, including the population of pregnant women with vaginosis (Amsel criteria or bacterioscopy), without other known risk factors for PTL. The intervention was likely metronidazole or clindamycin (oral or vaginal), used up to 28 gestational weeks, compared to placebo or no treatment. The assessed outcome was the incidence of PTL. All studies were evaluated using the Jadad score. A measure of probability of analysis (OR), calculated with 95% confidence interval (CI), and values of p < 0.05 were considered statistically significant. Heterogeneity between studies was assessed by I2-static, and all statistical analyses were performed with Software Review Manager 5.3 (Copenhagen, Denmark). Results: Nine studies were selected for systematic review and 8 were submitted to meta-analysis. All were randomized clinical trials with a Jadad score ≥ 3. The meta-analysis, which included 2 articles with oral metronidazole and 6, with vaginal clindamycin, showed that these treatments did not decrease the incidence of PTL (OR = 0.97, 95% CI 0.79-1.19, I2=0%). Only one study was identified using oral clindamycin, which showed the medication’s benefit in decreasing prematurity (OR 0.31, 95% CI 0.15-0.64). With an analysis of the funnel plot, this assay was excluded from the meta-analysis, as it would generate too much heterogeneity. Conclusions: The current study identified a promising outcome using oral clindamycin for treatment of vaginosis during pregnancy to reduce preterm labor. This finding should be interpreted with caution, however, since it was based on only one trial. Nonetheless, there is support in the literature for such a hypothesis, which justifies subsequent studies with the same drug.