Evidence from a systematic review suggests that both intravaginal clindamycin and intravaginal metronidazole are more effective than placebo at curing bacterial vaginosis (BV). Evidence from a systematic review suggests that there is no significant difference in efficacy among a 7-day regimen of oral metronidazole, intravaginal clindamycin, or intravaginal metronidazole. This is consistent with findings from subsequent randomized controlled trials (RCTs). Evidence from one RCT suggests that there is no difference in efficacy between a 7-day regimen of metronidazole and a 7-day regimen of oral clindamycin; further evidence regarding the efficacy and safety of clindamycin is needed. Evidence from a systematic review suggests that a single dose of metronidazole is significantly less effective than a 7-day regimen. CKS found no RCTs directly comparing oral metronidazole or oral clindamycin with tinidazole.

Oral antibiotic compared with placebo:

• CKS found no RCTs comparing oral antibiotic treatment with placebo in women who are not pregnant.

Intravaginal antibiotic compared with placebo:

• CKS found one systematic review (search date: 1996) comparing intravaginal antibiotic treatment with placebo [Joesoef et al, 1999]. The systematic review included two studies comparing clindamycin 2% intravaginal cream with placebo (n = 247) and two studies comparing metronidazole 0.75% intravaginal gel with placebo (n = 159). P-values and confidence intervals were not reported:
• Cumulative cure rates for clindamycin (5 g at bedtime for 7 days) were 82% 25–39 days after completion of treatment, compared with 35% for placebo.
• Cumulative cure rates for metronidazole (5 g twice a day for 5 days) were 71% 28–32 days after completion of treatment, compared with 50% for placebo.

Intravaginal antibiotic compared with oral antibiotic:

• CKS found one systematic review (search date: 1996) comparing intravaginal antibiotic treatment with oral antibiotic treatment [Joesoef et al, 1999]. The review included three RCTs comparing intravaginal clindamycin with oral metronidazole (n = 528), one study (published as an abstract) comparing intravaginal metronidazole with oral metronidazole (n = 112), and one study comparing all three treatments (n = 101):
• Pooling data from these studies showed that there was no significant difference in cure rates among oral metronidazole (500 mg twice a day for 7 days), clindamycin 2% vaginal cream (5 g at bedtime for 7 days), and metronidazole 0.75% vaginal gel (5 g twice a day for 5 days).
• The cumulative cure rates 5–10 days after finishing treatment were 85% with intravaginal clindamycin, 81% with intravaginal metronidazole, and 86% with oral metronidazole.
• The cumulative cure rates 4 weeks after finishing treatment were 82% with intravaginal clindamycin, 71% with intravaginal metronidazole, and 78% with oral metronidazole.
• CKS found one subsequent RCT (n = 399) comparing intravaginal antibiotics with oral antibiotics [Paavonen et al, 2000]:
• Women with BV were randomized to receive clindamycin 100 mg ovules intravaginally for 3 days plus placebo tablets for 7 days, or metronidazole 500 mg twice a day for 7 days plus placebo intravaginal ovules for 3 days:
• Of the 399 women enrolled, only 233 were evaluated for efficacy. There was no significant difference in cure rate between clindamycin ovules (68%) and oral metronidazole (67%; p = 0.81).
• Treatment-related adverse effects were reported to be more frequent in the metronidazole group.

Different regimens of intravaginal antibiotics:

• CKS found one RCT (n = 119) comparing intravaginal clindamycin with intravaginal metronidazole [Austin et al, 2005]:
• Women with symptomatic BV were randomized to receive either clindamycin vaginal ovules once a day for 3 days or metronidazole vaginal gel once a day for 5 days.
• There were no significant differences in cure rates between the two treatment groups:
• At 7–12 days, the cure rates were 79% for metronidazole and 88% for clindamycin (p = 0.3).
• At 35–45 days, the cure rates were 62% for metronidazole and 55% for clindamycin (p = 0.5).
• CKS found one RCT (n = 514) comparing once-daily with twice-daily dosing of intravaginal metronidazole gel [Livengood et al, 1999]:
• There was no significant difference in cure rates between once-daily and twice-daily dosing.
• In intention-to-treat analysis, after 1 month, the cure rates were 57% with once-daily gel and 62% with twice-daily gel.

Different regimens of oral antibiotics:

• CKS found one systematic review (search date: 1992) comparing different regimens of oral antibiotics [Joesoef and Schmid, 1995]:
• Single-dose metronidazole versus 7 days of metronidazole:
• From the data presented, it is not clear how many studies directly compared single-dose metronidazole (2 grams) with the 7-day treatment regimen (500 mg twice a day).
• After data from eight RCTs were pooled, the cure rate for the 7-day treatment regimen was 92% (295 of 322 women treated). After data from 11 RCTs were pooled, the cure rate for the single-dose regimen was 87% (369 of 426 women treated); the difference between the two groups was statistically significant (p < 0.05).
• The relapse rates in individual studies were 13–44% for the 7-day regimen and 0–53% for the single-dose regimen.
• When the analysis included only studies that directly compared the single-dose regimen with 7-day regimens, the cumulative cure rates 3–4 weeks after completion of treatment were 62% for the single-dose regimen and 82% for the 7-day regimen (p < 0.05).
• Oral clindamycin versus oral metronidazole:
• The systematic review included one RCT (n = 144) comparing oral clindamycin (300 mg twice a day for 7 days) with oral metronidazole (500 mg twice a day for 7 days).
• There was no significant difference in cure rates (94% with oral clindamycin vs. 96% with oral metronidazole).
• Oral tinidazole:
• CKS found no RCTs directly comparing oral metronidazole or oral clindamycin with tinidazole.
• A more recent review also concluded that a 7–day course of metronidazole was the optimum treatment for BV [Koumans et al, 2002].