Clare Calvert

Background Bacterial vaginosis is a common and distressing condition associated with serious comorbidities. Antibiotic treatment is usually clinically effective in the short term, but recurrence is common and side effects can occur. Objectives The objective is to assess whether or not intravaginal lactic acid gel is clinically effective and cost-effective for treating recurrent bacterial vaginosis compared with oral metronidazole (Flagyl, Sanofi). Design This was an open-label, multicentre, parallel-arm, randomised (1 : 1) controlled trial. Setting This took place in one general practice and 19 sexual health centres in the UK. Participants Women aged ≥ 16 years with bacterial vaginosis symptoms and one or more episode(s) within the past 2 years took part. Interventions The interventions were 5 ml of intravaginal lactic acid gel taken once daily for 7 days (intervention) or 400-mg oral metronidazole tablets taken twice daily for 7 days (control). Main outcome measures The primary outcome was the resolution of bacterial vaginosis symptoms 14 days after randomisation. The secondary outcomes were time to first recurrence of symptoms; number of recurrences and treatment courses over 6 months; microbiological resolution on microscopy of vaginal smears at week 2; time to resolution of symptoms; tolerability, adherence and acceptability of the treatment; prevalence of concurrent sexually transmitted infections; quality of life; and cost-effectiveness. Results Recruitment stopped prior to reaching the target of 1900 participants on recommendation from the Data Monitoring Committee and Trial Steering Committee after a planned review of the results indicated that the research question had been answered. Overall, 518 participants were randomised and primary outcome data were available for 409 participants (79%; 204 in the metronidazole arm, 205 in the lactic acid gel arm). Participant-reported symptom resolution at week 2 was higher with metronidazole (143/204; 70%) than with lactic acid gel (97/205; 47%) (adjusted risk difference –23.2%, 95% confidence interval –32.3% to –14.0%). Recurrence in 6 months in a subset of participants who had initial resolution and were available for follow-up was similar across arms (metronidazole arm: 51/72, 71%; lactic acid gel arm: 32/46, 70%). A higher incidence of some side effects was reported with metronidazole than with lactic acid gel (nausea 32% vs. 8%; taste changes 18% vs. 1%; diarrhoea 20% vs. 6%, respectively). At week 2, the average cost per participant with resolved symptoms was £86.94 (metronidazole), compared with £147.00 (lactic acid gel). Some participants preferred using lactic acid gel even if they perceived it to be less effective than metronidazole. Limitations Loss to fol