Objective
To estimate whether bacterial vaginosis, as defined by either Nugent score or vaginal pH, predicts gestational age at delivery in women at risk for recurrent preterm birth.
Study Design
Planned secondary analysis of a randomized cerclage trial in women with prior spontaneous preterm birth 17 0/7 -33 6/7 weeks. Vaginal Gram stain and pH were collected at the initial study visit. Women not assigned to cerclage, either because they did not experience cervical shortening <25 mm or because they were randomly assigned to no cerclage, were studied.
Results
Seven hundred eighty-six women had complete delivery gestational age, Gram stain, and pH results. The diagnosis of bacterial vaginosis by either Nugent score ≥ 7 or by pH ≥ 5 was not associated with earlier birth.
Conclusion
The presence of bacterial vaginosis at 16-21 6/7 weeks does not predict gestational age at birth in women at risk for recurrent preterm birth.
Prematurity is the single most important and persistent problem in obstetrics. In the United States, preterm birth (PTB) has been increasing and now complicates 12.5% of pregnancies. Preterm birth is the leading cause of perinatal morbidity and mortality, responsible for 75-90% of all neonatal deaths not due to congenital malformation, and 50% of neurologic disabilities, including cerebral palsy, blindness, and deafness. Amniotic fluid infection caused by a variety of microorganisms located in the genital tract has been implicated as a cause of PTB. Bacterial vaginosis (BV) is the most common genital tract syndrome in reproductive age women and as many as one-third of pregnant women have BV. BV is characterized by a complex change in the vaginal flora with a reduction in lactobacillus, an increase in the vaginal pH, and an overgrowth of other organisms, especially anaerobes. BV has been associated with an increased risk of PTB.
Diagnostic criteria for BV include clinical findings and laboratory tests. Amsel criteria include a grayish-white discharge that smoothly coats the vaginal walls, vaginal pH >4.5, positive whiff-amine test, and clue cells on wet mount ; 3 of these 4 are required for the diagnosis. Vaginal pH alone is also considered diagnostic if ≥5. Vaginal pH is 88.3% sensitive and 58.6% specific for the diagnosis of BV when using Gram stain as the reference method. Gram stain of the vaginal fluid is a reliable and reproducible method for the diagnosis of BV and is often used in clinical research. Nugent’s Gram stain criteria for the diagnosis of BV has a high sensitivity (89.1%) and specificity (83%), when Amsel’s criteria is used as the reference method. BV is diagnosed by a Nugent score ≥7.
Our aim was to evaluate the association between BV as determined by either Gram stain Nugent score or vaginal pH in the midtrimester and gestational age at delivery in women at high risk for recurrent PTB because of prior spontaneous PTB 17-33 6/7 weeks’ gestational age (GA).
Materials and Methods
Study patients
This is a planned secondary analysis of the NICHD-sponsored randomized controlled trial of cerclage for PTB prevention. The detailed study protocol and study patients have been previously described. This analysis considered women who began vaginal ultrasound screening between 16 0/7 and 21 6/7 weeks’ gestation to indentify candidates for randomization to cerclage or no cerclage based on short cervix <25 mm. Our study population was comprised of women who maintained cervical length ≥25 mm and those with short cervix randomly assigned to no cerclage.
Collection of specimens
Before the initial sonographic cervical length (CL) evaluation an unlubricated sterile speculum examination was performed to collect vaginal fluid for pH and Gram stain. A dry cotton-tipped applicator or Dacron swab was used to harvest vaginal fluid from the upper third of the vaginal sidewalls. The swab with vaginal fluid was touched to the reagent block on the pH indicator strip. Only colorpHast indicator strips, stored in darkness, were used (ColorpHast; EMD Chemicals, Gibbstown, NJ). A second swab with vaginal secretions was spread in a thin layer on a dry glass slide by gently rolling the swab across the entire slide. Vaginal fluid slides were allowed to air dry. Study specimens were collected before other routine tests (eg, Papanicolau [Pap] smear test, cultures) to minimize blood contamination.
Evaluation of vaginal specimens
The pH was read after the color stabilized, but before the paper dried completely. The vaginal pH was recorded as indicated by the pH strips. If the observed hue fell between 2 values on the chart, the recorded value was rounded up. The vaginal fluid slides were stored and processed at the completion of the study. The slides underwent Gram staining and were evaluated under magnification (×1000) for the presence of Lactobacillus , Gardnerella vaginalis , Bacteroides , and Mobiluncus morphotypes, Gram stains were scored by Nugent criteria and a score of 0-10 was assigned. A Nugent score of 0 indicated the most Lactobacillus -predominant flora, whereas a score of 10 represented flora largely replaced by Gardnerella , Bacteroides , and Mobiluncus species.
Diagnosis of BV
Women with a vaginal pH ≥5 or Nugent score ≥7 were considered to have BV. The vaginal fluid pH and Nugent score were not disclosed to the study participants or the providers, nor were investigators performing the pH and Gram staining interpretation aware of patients’ birth outcome.
Statistical analysis
The univariate relationships between BV diagnosed by either Nugent score ≥7 or pH ≥5 and PTB before 28 weeks, 35 weeks, and 37 weeks of gestation were assessed with χ 2 . GA at delivery was also modeled as a continuous outcome variable in linear regression models. Possible confounders considered in multiple logistic and linear regression models, included maternal age, race, smoking, urinary tract infection, as well as GA and sonographically measured CL at study specimen collection.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
