From a clinician's perspective of diligent diagnosis and proper treatment, bacterial vaginosis may be the most innocently ignored and improperly managed vaginal infection. Diagnosis involves consideration of microbiologic, cytologic, clinical and chemical factors. The most sensitive chemical parameter (abnormally elevated pH) is never considered by 95 percent of health care professionals. In contrast with other types of vaginitis, simple microscopic recognition of a unique pathogen does not confirm bacterial vaginosis infection.

No one knows what causes the virtual banishment of Lactobacillus acidophilus from the vagina and the preponderance of a diverse mixture of opportunistic microorganisms within the vagina. Because the true etiology of bacterial vaginosis remains unknown, therapeutic agents are not universally effective. As a consequence of imprecisely directed treatment (i.e., treating the expression of bacterial vaginosis and not the exact cause), the recurrence of bacterial vaginosis infection is exceedingly common.

Yet, data from studies of pregnant women with bacterial vaginosis demonstrate substantial twofold or greater increased risk for serious complications, including preterm birth, low-birth-weight infants, premature rupture of the membranes, amniotic fluid infection, chorioamnionitis, and post-cesarean and postpartum endometritis. Although bacterial vaginosis has been shown to be an independent risk factor for these complications, many health care professionals still consider bacterial vaginosis more of a nuisance than a genuine fetal-maternal threat. Even though widely disseminated educational efforts should rectify misunderstandings with regard to pregnancy outcome and bacterial vaginosis, many other critical questions concerning this topic remain unanswered. Should asymptomatic pregnant women with bacterial vaginosis be treated? Is there a role for screening pregnant women for bacterial vaginosis? Which antibiotics effectively reduce the risk for pregnancy-related complications of bacterial vaginosis?

Clearly, a casual approach to pregnant women with symptoms or clinical signs of bacterial vaginosis appears unjustified. Symptomatic pregnant women with confirmed bacterial vaginosis should be treated.¹ Whether asymptomatic pregnant women deserve, and would benefit from, therapy is less well defined. More than one half of all women with bacterial vaginosis have no symptoms of the lower genital tract; nevertheless, they are still afflicted. One could easily argue that treatment should be implemented for pregnant women because of the increased probability of serious complications, regardless of the absence of symptoms.

Several researchers have demonstrated that treatment of bacterial vaginosis in pregnant women reduces the rate of preterm birth.^2,3 Hauth and colleagues² showed that pregnant women who had bacterial vaginosis and an increased risk for preterm delivery (previous history of preterm delivery or low prepregnant weight of less than 50 kg [110 lb]), and who received metronidazole (Flagyl) and erythromycin therapy, had significantly reduced rates of preterm delivery when compared with women given placebo (39 percent versus 57 percent). Morales and colleagues³ demonstrated that oral metronidazole therapy reduced the rate of preterm births for pregnant women with bacterial vaginosis and a previous history of preterm birth when compared with women who received placebo (18 percent versus 39 percent).

Of particular importance is the fact that both of these studies^2,3 demonstrated therapeutic efficacy for only pregnant women considered at high risk for preterm delivery. It is not clearly understood if low-risk pregnant women with no underlying increased risk factors for preterm birth would benefit equally as well from treatment of bacterial vaginosis. Similarly, claims of benefit from treatment of asymptomatic pregnant women with bacterial vaginosis have not been properly substantiated. Recent guidelines from the Centers for Disease Control and Prevention (CDC) recommend treating asymptomatic high-risk pregnant women with bacterial vaginosis.⁴ It may also be prudent to treat symptomatic low-risk pregnant women with bacterial vaginosis to eliminate their symptoms.⁴

The effectiveness of screening patients to prevent potentially serious sequelae depends on the prevalence of the condition in the population at risk, readily available and accurate diagnostic tests, consequences of the untreated condition, effective therapy and the overall cost-effectiveness of such intervention. In support of screening pregnant women for bacterial vaginosis are the following factors: the condition is exceedingly common (a prevalence rate of up to 30 percent); reasonably reliable tests are available; the potential consequences of not treating bacterial vaginosis are severe for the mother and the fetus; and moderately effective therapy is available. While these factors bolster screening, many questions regarding screening pregnant women for bacterial vaginosis exist.

If screening is done, when should it be conducted, second or third trimester?

It appears that infection with bacterial vaginosis in early pregnancy (second trimester) conveys a greater risk for complications than infection with bacterial vaginosis in late pregnancy.⁵ However, a positive test for bacterial vaginosis in early pregnancy may be a poor predictor for the development of preterm birth, preterm labor and premature rupture of the membranes (positive predictive values: 4 percent to 11 percent).⁶ Based on increased risk, current CDC guidelines recommend screening early in the second trimester.⁴

Should all pregnant women be screened, or only women considered to be at increased risk for fetal-maternal complications?

While treatment of high-risk pregnant women with bacterial vaginosis reduces the risks for fetal-maternal complications,^2,3 no data are available to compare bacterial vaginosis treatment outcomes for both women at low risk and women at high risk of fetal-maternal complications. Therefore, currently only women determined to be at high risk for preterm delivery should be considered candidates for screening for bacterial vaginosis.^1,4

How should screening be performed—using simple clinical evaluation and Amsel's criteria, Gram stain, gas-liquid chromatography (GLC) or fibronectin assay?

Screening by Amsel's criteria (three of four findings: pH of more than 4.5, amine odor on adding KOH, presence of clue cells, adherent off-white vaginal discharge) is readily available to all clinicians but is also the least accurate method of diagnosing bacterial vaginosis. The other tests are considerably more expensive and are not readily available, but they offer more accurate diagnosis or assessment of increased risk for potential complications. Whether screening would be cost-effective is controversial,^7,8 because data based on outcomes from randomized controlled screening trials are not available. Screening guidelines may evolve from the BV/TV trials sponsored by the National Institutes of Health, but initial data will not be available until 1999. As such, no guidelines currently recommend universal screening of pregnant women for bacterial vaginosis.

Some limited data are available to help guide clinicians in the selection of antibiotics for bacterial vaginosis in pregnant women. Topical clindamycin vaginal cream is ineffective in reducing the rates of preterm birth.^9,10 In fact, such treatment actually increases the presence of vaginal Escherichia coli, an organism known to increase the risk for preterm birth. Topical metronidazole gel (Metrogel) has not been evaluated in the context of bacterial vaginosis during pregnancy. Topical antibiotics usually eradicate local bacterial vaginosis infection, but do not reduce prematurity sequelae because of the lack of access to the upper genital tract. Therefore, systemic antibiotics are probably required to adequately reduce the risk of pregnancy-related complications.

Oral metronidazole and metronidazole combined with erythromycin have been shown to reduce pregnancy complications associated with bacterial vaginosis.^2,11 But because metronidazole use is contraindicated during the first trimester, only women in mid to late pregnancy should be treated with the drug. Alternatively, oral clindamycin (Cleocin) could be used, but limited data are available on its use,¹² particularly in the context of treating women without a current or past history of pregnancy-related complications. A test of cure evaluation one month following treatment of bacterial vaginosis may be beneficial because treatment failures are common.^4,13

In summary, women with bacterial vaginosis during pregnancy should be aggressively evaluated and effectively treated. This is particularly true for women considered at high risk for pregnancy-related complications. Asymptomatic pregnant women with bacterial vaginosis may also benefit from therapy. Systemic antibiotics appear to afford both effective treatment for bacterial vaginosis and minimization of pregnancy-related complications.

Universal screening of pregnant women for bacterial vaginosis is not currently recommended, but women at high risk for preterm birth may benefit from early second trimester screening for bacterial vaginosis. Casual clinical recognition of bacterial vaginosis in asymptomatic pregnant women should prompt proper diagnosis and treatment. Although pregnant women with bacterial vaginosis obviously have an increased risk for pregnancy-related complications, it is unknown whether therapeutic intervention decreases the rate of specific fetal-maternal problems for all pregnant women. Evidence-based guidelines for proper management of pregnant women with bacterial vaginosis await the outcomes of clinical trials currently being conducted.