Background To determine if addition of vaginal clindamycin or an oestrogen-containing vaginal probiotic, to current recommended therapy for bacterial vaginosis (BV), oral metronidazole for 7 days, reduces 6 month recurrence rates.

Methods Three arm randomised double-blind placebo controlled trial of 450 women (150 per arm): [MetPlac] oral metronidazole (7 days)/vaginal placebo (12 days), [MProb] oral metronidazole (7 days)/vaginal probiotic(12 days) and [MetClin] oral metronidazole (7 days)/vaginal clindamycin (1 g 2% nocte,7 days). Symptomatic 18–50-year-old females with BV on vaginal swab by the Nugent method were enrolled at Melbourne Sexual Health Centre, Australia. Participants underwent initial examination & STI screen and completed a detailed behavioural questionnaire at 0, 1, 2, 3 & 6 months. At each interval participants were posted a kit containing swabs and a slide for self-collection and a questionnaire. Principle study outcome: Nugent score of 7–10. Cumulative BV recurrence rates were calculated and compared using χ² and survival analyses using SPSS and STATA.

Results 450 women with BV were recruited from December 2007 to May 2010. Median age was 27 years (range 18–49), 210 (48%) reported a past history of BV; there were no significant differences in participant demographic or behavioural characteristics between arms. Adherence to study medication did not differ between arms: 382 (91%) took all or most oral metronidazole and 330 (80%) all or most vaginal therapy. Retention rates were high, with 77 (17%) lost to follow-up over 6 months, and did not differ between arms. Participants contributed 153.7 person years of follow-up to analyses. On exit survey 88% of participants did not know or correctly guess the vaginal therapy they had received. Six month cumulative BV recurrence rates did not differ between study arms by per protocol analysis: MetPlac (32%, 95% CI 24% to 41%) MetProb (33%, 25% to 42%) and MetClin (34%, 26% to 42%), p>0.05 (Abstract O3-S5.06 figure 1), or intention-to-treat analysis (noncompleter=recurrence) [recurrence range 44-9%].

• Download figure
• Open in new tab
• Download powerpoint

Abstract O3-S5.06 Figure 1

Conclusions The addition of vaginal clindamycin or a vaginal probiotic to oral metronidazole does not improve 6 month BV recurrence rates. This is the first RCT to evaluate the efficacy of combination clindamycin/metronidazole for BV treatment, and has important implications for clinical practice. Combination therapy is often used in patients with recurrent BV, but evidence to support this practice has not been available.