Bacterial vaginosis occurs when the normal lactobacillus species in the normal vaginal flora is replaced with high concentrations of various anaerobic organisms. Bacterial vaginosis is the most common cause of vaginal discharge in women during their reproductive years, and women with bacterial vaginosis are usually asymptomatic. The standard treatment for bacterial vaginosis is a single, 2-g oral dose of metronidazole. Four weeks after treatment, this therapy has a cure rate of 70 to 80 percent; however, this cure rate tends to diminish over time. Studies on the treatment of bacterial vaginosis are difficult to interpret because of differences in study populations and treatment regimens, and in the definition of resolution of bacterial vaginosis. Klebanoff and colleagues determined the effectiveness of treatment over time for bacterial vaginosis and the probability of spontaneous resolution with placebo in women who are pregnant.

The trial was a randomized study of pregnant women who had an established diagnosis of bacterial vaginosis. Inclusion criteria included being at eight to 22 weeks of gestation with no genital itching, burning, or malodor, and no contraindication for receiving metronidazole. Women were excluded from the study if they had a major medical or obstetric complication during the current pregnancy, had taken antibiotics within the past 14 days, or had trichomoniasis, syphilis, gonorrhea, or Chlamydia. At the next visit, patients who agreed to participate in the study were randomized to receive a 2-g oral dose of metronidazole or placebo. This dose was repeated 48 hours later. Patients had a follow-up visit to assess the presence or abence of bacterial vaginosis no sooner than two weeks after the intervention.

A total of 658 women received metronidazole therapy, and 683 women received placebo. Of the women who received metronidazole, 78 percent had successful treatment. This success rate was relatively consistent if patients were seen within two to four weeks or more than 10 weeks after the treatment. The women who received placebo had a 13 percent rate of spontaneous resolution of bacterial vaginosis when evaluated at two to four weeks after the intervention and a 36 percent rate if assessed 10 weeks or more after the intervention. Women in the placebo group who had a lower vaginal pH or lactobacillus on the Gram stain were more likely to have spontaneous resolution of bacterial vaginosis.

The authors conclude that the effectiveness of metronidazole therapy for bacterial vaginosis persists for 10 or more weeks after treatment. They add that women who received placebo also had a significant spontaneous resolution of bacterial vaginosis over a period of 10 or more weeks. The spontaneous resolution appears to be directly related to the severity of the vaginal micro-biologic disarray.