Bacterial Vaginosis: An Update on Diagnosis and Treatment

Hans Verstraelen; Rita Verhelst


Expert Rev Anti Infect Ther. 2009;7(9):1109-1124. 

In This Article

Treatment of BV

Despite numerous therapeutic research efforts, treatment of BV remains cumbersome and clinicians are currently rather poorly armed to treat BV properly in the long run. Moreover, appraisal of the available evidence on treatment options for BV is hindered by methodological concerns.

Rationale for Treatment of BV

Treatment of BV is primarily targeted at resolving or alleviating the presenting symptoms, most commonly profuse, foul-smelling vaginal discharge. In the absence of treatment BV may resolve spontaneously; however, it often recurs over an extended period of time. As recurrence of BV is common, treatment of BV is ideally also targeted at prophylaxis of BV recurrences. On a secondary level, BV treatment aims to prevent infectious complications associated with BV, and hence also involves treatment of asymptomatic BV.

Definition of Cure

Assessment of therapeutic efficacy of a given treatment depends on the clinical context in which such an evaluation is to be made. In clinical practice, cure of BV will therefore typically be considered a fact if a patient reports the resolution of the symptoms for which therapy was instigated on the occasion of a control visit. Contrary to this symptomatic approach, a more objective manner of establishing cure is to make a reassessment by use of the Amsel's criteria[1] or Nugent's criteria.[3]

This approach has not been properly defined, however, and hence some ambiguity prevails. If cure is evaluated through Amsel's criteria, cure could be accepted as long as no more than two signs defined by Amsel et al.[1] persist following therapy, since it takes at least three out of four signs to build the diagnosis. Many researchers agree however that so-called clinical cure is reflected by the resolution of all Amsel's signs. Similarly, so-called microbiological cure as assessed through Nugent's criteria[3] has not been unequivocally defined: does microbiological cure indicate that the condition of BV has resolved, that is, a Nugent score below 7, or does it entail the reconversion to a normal vaginal microflora, that is, a Nugent score of less than 4? Both definitions have been applied throughout the literature. The distinction between both situations is particularly important, as the former definition, reconversion to a vaginal microflora with a Nugent score of less than 7, may implicate intermediate microflora (Nugent score 4–6) as the normalized state, while intermediate microflora are significantly more likely to shift to BV than normal microflora defined by a Nugent score of less than 3.

Another as yet undefined aspect with regard to the assessment of cure relates to the time frame relative to therapy. Many studies have reported cure rates through the assessment of patients after treatment cessation or at 1-week follow-up. Only a small number of studies typically report on cure as defined through Amsel's or Nugent's criteria at 3–4 weeks of follow-up. As more than 50% of woman may have persistence or recurrence within 2 months of antibiotic treatment,[52] assessment of cure at such extended follow-up might be an even better indicator of therapeutic efficacy, although very few studies have actually adopted this approach. At present, the sole guideline on the assessment of cure following treatment of BV comes from a draft guidance document that was launched by the US FDA.[53] According to this guidance, cure is defined as the absence of all four Amsel's signs and a Nugent score of less than 4 at a test-of-cure visit 21–30 days after the first day of treatment. This stringent approach is now increasingly gaining attention as a standard to define cure of BV by researchers in this field.

Treatment Modalities

Antibiotics have served as the mainstay of BV treatment over the past six decades. Several alternative approaches also deserve attention, however, and can grossly be categorized as treatment with antiseptics and disinfectants, with vaginal acidifying or buffering agents, and with probiotics, which have been used as a standalone therapy or in conjunction with antibiotics.

Standard Treatment with Antibiotics

Soon after the advent of antibiotics and as early as in the 1950s, attempts were made to identify the optimal antibiotic treatment for BV. A number of antibiotics passed the revue in the following decades, although many of the alleged early therapeutic successes in this era could not be repeated later.[54]

Eventually, Pheifer et al. reported, in their 1978 New England Journal of Medicine paper, on an oral regimen of 500 mg metronidazole twice daily for 7 days that eradicated Haemophilus vaginalis in all 81 patients treated, with resolution of symptoms and signs in 80 patients.[54] A total of 10 years later, Greaves et al. found an oral regimen of 300 mg clindamycin twice daily to be equally safe and effective as the metronidazole regimen.[55] Since then, metronidazole and clindamycin have been the drugs of choice in the treatment of BV, as also recommended by the most recent sexually transmitted disease treatment guidelines by the US CDC (Box 1).[56]

So how strong is the evidence in support of the CDC guidelines and, hence, how effective are the recommended regimens? There are over 100 published clinical trials on treatment of BV with antibiotics, although this bulky material has never been subjected to a meta-analysis, presumably due to the notorious methodological heterogeneity across studies as indicated earlier. The only evidence-based approach comes from a narrative systematic review that is regularly updated within the framework of the BMJ Clinical Evidence series,[52] which at present provides a systematic appraisal of the evidence of interventions for BV published before July 2006, of which the main findings on strength of evidence are summarized in Table 1. From this evidence-based review, it was concluded that all CDC recommended regimens are likely to beneficial in the treatment of BV in terms of short-term benefit, with no obvious differences between the different recommended regimens, albeit the overall quality of evidence is actually rather low.[52] In addition, few studies have assessed long-term cure rates, but it is generally accepted that recurrence of BV even after a proper treatment course is high, with some 30–50% of women experiencing a BV relapse within 2–3 months.

Therefore, even with the drugs-of-choice, it appears as if antibiotic treatment largely fails to prevent the relapse of BV and hence clinicians are currently rather poorly armed to properly treat BV in the long run. Much as a result of the deploring long-term cure rates with standard antibiotic treatment of BV, several alternative antibiotics and regimens have recently been evaluated for their clinical efficacy in eradicating BV.

Of particular relevance here is the renewed interest in 5-nitroimidazole derivates other than metronidazole, such as secnidazole,[57,58] ornidazole[57,59–61] and tinidazole.[59,62–78] Overall, the metronidazole relatives, secnidazole, ornidazole and tinidazole seem to be as effective as metronidazole itself.

Other alternative antibiotics and regimens have also recently been considered. In a recent large randomized, controlled trial the search for alternative antibiotic regimens was directed towards an extended-release metronidazole regimen and to azithromycin as yet another candidate in BV treatment. In this rigorously designed clinical trial, Schwebke and Desmond randomized 568 women with symptomatic BV to one of four arms; that is, a regimen of extended-release metronidazole (750 mg orally) once per day for 7 days was compared with a regimen of extended-release metronidazole (750 mg orally) once per day for 14 days, with both metronidazole regimens additionally being evaluated with and without azithromycin 1 g orally on days 1 and 3 as an adjuvant antibiotic.[79] As compared with the standard 7-day regimen, the extended 14-day administration regimen for metronidazole was associated with an increased cure rate at 7 days post-treatment (82.5 vs 73.8%), although not at 21 days post-treatment (65.9 and 75.6% respectively). Furthermore, there was no therapeutic effect at all associated with the administration of azithromycin in addition to metronidazole.[79]

Antibiotic Resistance with BV Treatments

There are few reports on antimicrobial resistance of BV-associated anaerobes, except for a vast number of studies on metronidazole susceptibility of G. vaginalis, with widely varying rates of metronidazole resistance having been reported. A single study that was addressed in two reports is of particular interest here.[80,81] In this randomized, clinical trial, 119 nonpregnant women diagnosed according to Amsel's criteria and with a Nugent score of 4 or more were randomized to receive either intravaginal metronidazole gel for 5 days or intravaginal clindamycin ovules for 3 days. Quantitative vaginal cultures were then performed at three follow-up occasions over 90 days and isolated anaerobes were assessed for metronidazole and clindamycin susceptibility, respectively. Emerging resistance to metronidazole following therapy was a rare phenomenon, occurring with merely 0.3% of Gram-negatives evaluated for susceptibility. By contrast, approximately half of the anaerobic isolates tested in the clindamycin arm rapidly developed antibiotic resistance to clindamycin, primarily involving Prevotella spp.[80,81] It was concluded from this study that emerging clindamycin resistance of vaginal anaerobic bacteria is a matter of concern, especially since clindamycin and other macrolides are widely applied in obstetrics and gynecology.

The recent discovery of several difficult-to-culture bacteria with BV through molecular techniques has, however, shed a new light on the resistant nature of BV in response to treatment with metronidazole. In particular, a considerable number of strains of the BV index species A. vaginae have been found to elicit pronounced metronidazole resistance in subsequent studies,[40,82–84] while the limited number of strains evaluated thus far consistently showed proper susceptibility to clindamycin.[84]

The picture has become even more intricate, as BV has recently been found to involve the development of a dense bacterial biofilm consisting of G. vaginalis and A. vaginae.[48] Swidsinski et al. indeed documented that the biofilm persisted but temporarily switched to a metabolically latent state with standard treatment with oral metronidazole and then rapidly regained activity following treatment cessation.[85]

So when accounting for the complexity of the bacterial community dynamics with BV, it may be questioned whether antibiotic susceptibility testing in vitro is actually of any value to the study of BV.

BV Recurrence

It remains unclear at present whether BV recurrence reflects resistance, recurrence and/or reinfection.[86] The surge of molecular studies is likely to clarify this issue in the nearby future. Recent studies indicated, for instance, that the presence of A. vaginae with BV is an indicator of treatment failure,[41] as well as indicator of BV recurrence.[87] It has been further postulated that following the resolution of BV, Lactobacillus iners is the Lactobacillus species most likely to replenish the vagina in appreciable amounts, which in turn may render patients more vulnerable to a new episode of BV, considering the rather moderate colonization resistance offered by L. iners.[49] Albeit based on very small numbers of observations, the prevailing hypothesis is that recuperation of a stable Lactobacillus microbiota following treatment is crucial to the long-term outcome of BV treatment, and that the abundance of L. iners with BV and following BV cure may compromise this evolution.

Alternative Treatment with Antiseptics & Disinfectants

Antiseptics have been applied for over half a century in the treatment of vaginal infections. Essentially, antiseptics are applied from the same perspective as antibiotics, which is, the eradication of the vaginal microflora with BV, following which recolonization with indigenous lactobacilli is expected. Antiseptics generally have a very broad spectrum as they act nonspecifically on bacteria through mechanisms such as bacterial cell membrane disruption. In accordance, there are very few reports on antimicrobial resistance with these agents. Antiseptics are also generally regarded as safe for mucosal application when administered in appropriate concentrations. Hence, antiseptics may be unduly overlooked in contemporary gynecological practice, as they may provide a valuable alternative without systemic exposure to antibiotics.

We are currently performing a systematic review on the efficacy and safety of antiseptics and disinfectants in the treatment of BV. To this purpose, we conducted a systematic literature search on antiseptics and disinfectants listed in the Anatomical Therapeutic Chemical (ATC) Classification System under the code D08A ('antiseptics and disinfectants'), with the exceptions of hydrogen peroxide, which is listed in the ATC classification system under the code S02AA for 'anti-infectives' and benzydamine which is listed under ATC code G02CC03 for 'anti-inflammatory products for vaginal administration'. Through this thorough search we identified a total of 12 clinical trials that are further considered for evidence-based appraisal.

Antiseptics that have been administered to women with BV as vaginal suppositories, bioadhesive gel formulations and occasionally loaded on pessaries in these clinical trials include benzydamine,[88] chlorhexidine,[89,90] dequalinium chloride,[91] polyhexamethylene biguanide,[92,93] povidone iodine[65,88,91,94–96] and hydrogen peroxide.[97,98]

Most of these studies were actually rather well-designed, however, most were single-blinded and none was placebo-controlled. Although possibly prone to publication bias, most studies documented cure rates that were at least as effective as the antibiotics they were weighted against. However, as each particular antiseptic, except for povidone iodine, has been studied only once or twice, no firm conclusions on these antimicrobial agents can be drawn, although as a whole, this approach definitely warrants further scrutiny as it may offer a superior means – without involving systemic exposure to antibiotics – of treatment and prevention of BV recurrence.

Alternative or Adjuvant Treatment with Acidifying Agents

A basic component of the healthy vaginal ecosystem is the maintenance of an acidic vaginal environment at an average pH of 4 ± 0.5 – inhospitable to most bacteria and viruses – through the enzymatic conversion of epithelial cell-derived glycogen primarily into lactate.[99,100] Alkalinization of the vaginal milieu as induced, for example, by menses or sperm leads to decreased epithelial adherence of the lactobacilli and gives a free rein to the overgrowth of typical BV-associated microorganisms. From this perspective, it has been postulated that actively acidifying the vagina with naturally occurring acids like lactate or buffering the vagina against alkalic exposures may enhance lactobacillary colonization and prevent anaerobic overgrowth. Hence, vaginal acidifying or so-called buffering agents might serve as either a therapeutic or preventive means to BV.

In exploratory studies not involving BV, a tampon lubricated with a lactate-buffered gel was not found effective in one study,[101] whereas in another study an intelligent tampon comprising a polymeric delivery system that upon absorption of menstrual fluid gradually releases lactic acid and citric acid convincingly leveled off the pH increase associated with menses.[102]

However, in recent clinical trials, the acidifying approach did not unequivocally prove effective. In a placebo-controlled randomized study by Holley et al., a 0.92% acetic acid-based gel applied twice daily for 7 days was not superior to a placebo gel in the cure of vaginosis.[103] In another study, the acid-buffering ACIDFORM gel was also significantly less effective than a 10% metronidazole gel.[104] Intravaginal vitamin C has been proposed as yet another potentially beneficial acidifying agent.[105]

Two studies evaluated the therapeutic efficacy of gels that contain polycarbophil, a weak polyacid that it is able to stick on the vaginal epithelial cells until they turnover (up to 3 to 5 days) and that is assumed to buffer the vaginal secretions. A noncontrolled study found polycarbophil to be moderately effective in treating BV, despite the lack of the alleged pH-controlling effect,[106] whereas in a randomized, double-blind, placebo-controlled trial, a polycarbophil–carbopolol-based vaginal gel was found to be very effective at 1 week following a 5-week treatment course.[107]

Interestingly, a couple of studies documented promising results with vaginal acidifiers for the long-term treatment of recurrent BV. Almost two decades ago, Andersch et al. treated 42 women with recurrent BV with a lactate gel for 7 days and then randomized these study participants to receive either prophylactic lactate gel 3 days monthly for 6 months or a placebo gel.[108] Patients treated prophylactically with lactate gel for 6 months had – albeit not defined by conventional criteria – an 83% cure rate as compared with 16% of the women in the control group, although it must be acknowledged that this comparison might be biased, as 1-week cure rates (following a 6-month treatment course) in the active treatment group were compared with 6-month treatment rates in the placebo group. In turn, Wilson et al. enrolled 61 women with recurrent BV, of whom 49 were followed for a mean duration of 18 months.[109] Following standard antibiotic treatment, patients were instructed to use a commercial vaginal gel containing 0.94% glacial acetic acid either nightly or after trigger factors such as menses and coitus. BV recurrences were treated with standard antibiotic treatment. Although there was no control group, the authors concluded that the use of the 'maintenance acetic acid vaginal gel' was associated with a significant decrease in the number of women having any further relapses, as well as in the rate of recurrences among the remainder, the overall recurrence rate being reduced from 4.4 to 0.6 recurrences per woman/year.[109] Clearly, more rigorously designed trials are needed to evaluate the potential role of vaginal acidifying and buffering agents in the treatment and particularly in the prevention of recurrent BV.

Alternative or Adjuvant Treatment with Probiotics

Probiotics have been defined by the United Nations Food and Agriculture Organization and the WHO in 2001 as live microorganisms, which when administered in adequate amounts confer a health benefit to the host.[110] Somewhere down the line, the presence of lactobacilli in fermented dairy products and yoghurt in particular has led to the idea that yoghurt could also offer an alternative treatment for vaginal infections. Eating yoghurt or inserting yoghurt into the vagina, for instance, by soaking a tampon in it, has therefore been a popular alternative treatment for vaginal infections over the past decades. Only one clinical trial has actually studied oral yoghurt therapy – in an open, crossover design – for the prophylaxis of recurrent BV.[111] The latter study suffered, however, from such a huge attrition bias – merely 25% of study participants completing the study protocol – that any conclusion is prevented from being drawn. In another two trials, the effect of intravaginal administration of fermented dairy products on BV was assessed. Fredricsson et al. used a fermented milk product and found that at 4 weeks, only one out of 13 patients had been successfully treated with fermented milk, comparing unfavorably with 13 out of 15 patients being successfully treated with metronidazole.[112] Neri et al. enrolled 84 women with BV diagnosed during early pregnancy and obtained a cure rate of 87.5% in the intravaginal Lactobacillus acidophilus yoghurt group as compared with a cure rate of 37.5% in the group who applied a vaginal tampon soaked in 5% acetic acid.[113]

The yoghurt approach is questionable, however, taking into consideration the dissimilarity of lactobacilli pertaining to food products and the vaginal Lactobacillus species. Yoghurt and other fermented milk products rely on the mandatory addition of two lactic acid bacteria, Streptococcus salivarius subsp. thermophilus and Lactobacillus delbrueckii subsp. bulgaricus, to initiate the fermentation process in milk. These elementary bacteria are often cocultured with other lactic acid bacteria, including L. acidophilus, Lactobacillus casei and Bifidobacterium species, but not with typical vaginal Lactobacillus species.

Hence, it was not until more recently that a limited number of well-documented Lactobacillus strains have been subjected to study of their colonization potential upon oral or vaginal administration, such as the L. crispatus strain CTV-05,[114] the Lactobacillus rhamnosus strain GR-1[115] and the Lactobacillus reuteri strain RC-14.[116] As a matter of fact, 100 years after the probiotic principle was launched by Metchnikoff, treatment of BV with probiotics is regarded as the most promising therapeutic perspective in this particular area. A number of potentially probiotic strains are currently under study, that is, vaginal Lactobacillus are carefully selected based on their properties relating to mucosal colonization and microbial antagonism, including adhesion to vaginal epithelial cells, hydrogen peroxide production, bacteriocin production, co-aggregation with pathogens, and overall inhibitory or antagonistic activity towards BV-associated microorganisms.

In order to assess the current evidence on the safety and efficacy of probiotics in the treatment of BV, we conducted a systematic review in collaboration with the Cochrane Collaboration.[117] We performed a thorough search for published and unpublished trials conducted or published before 2008 on any probiotic used alone or in conjunction with antibiotics in treating BV diagnosed according to Nugent's or Amsel's criteria. The primary outcome measure set forward in the study protocol was BV cure according to Nugent's criteria at 21–30 days post-treatment. Overall, 16 studies published between 1992 and 2006 were identified, of which four met the criteria for inclusion in the systematic review.

In the study by Parent et al., microbiological cure on day 28 was documented in seven out of eight women in the arm who received one Gynoflor® tablet (10 million viable 'L. acidophilus bacteria' and 0.03 mg estriol per tablet) daily for 6 days compared with one out of seven women in the placebo arm (OR: 0.02; 95% CI: 0.00–0.47%).[118] It may be added here that the so-called L. acidophilus labeled by the Gynoflor manufacturer is actually a L. crispatus strain, based on a product analysis we performed [Verhelst R, Verstraelen H, Unpublished data]. In the study by Eriksson et al. 225 subjects were initially enrolled in the open part of the trial and all were treated with 100 mg clindamycin ovules inserted vaginally once daily for 3 days and then randomized to one of two arms.[119] During the first menstrual period following clindamycin treatment, patients received either lactobacilli-impregnated tampons (1 × 108 freeze-dried L. rhamnosus, Lactobacillus gasseri and Lactobacillus fermentum cells per tampon) or placebo tampons, and participants had to use at least five tampons to be included in the efficacy analysis. After the second menstruation, during which the women used their normal catamenial protection, and according to an intention-to-treat analysis, 75 out of 108 women in the treatment group and 80 of the 109 women in the placebo arm presented with normalized Nugent scores (OR: 0.82; 95% CI: 0.46–1.49%). The authors explained these results by the fact that the Lactobacillus strains were not properly released from the tampons. In a first study by Anukam et al., patients were randomized in a double-blind manner and given oral metronidazole 500 mg twice daily for 7 days, plus either oral L. rhamnosus GR-1 and L. reuteri RC-14 (1 × 109 cells per capsule) twice daily for 30 days starting on day 1 of metronidazole treatment or identical-looking placebo capsules.[120] In the per-protocol analysis, BV cure at day 30 according to Nugent's criteria, occurred with 43 of the 49 women in the antibiotic/probiotic treatment arm and with 23 of the 57 women in the antibiotic/placebo arm (OR: 0.09; 95% CI: 0.03–0.26%). In a second randomized, controlled trial by Anukam et al., 40 women diagnosed with BV were randomized to receive either two dried capsules containing L. rhamnosus GR-1 and L. reuteri RC-14 each night for 5 days, or 0.75% metronidazole gel, applied vaginally twice a day.[121] Follow-up at days 6, 15 and 30 showed cure of BV in significantly more probiotic-treated subjects (16/20, 17/20 and 18/20, respectively) compared with metronidazole treatment (9/20, 9/20 and 11/20, respectively; p = 0.016 at day 6, 0.002 at day 15 and 0.056 at day 30). Hence, in this study the probiotic regimen proved superior to metronidazole gel on 1- and 2-week cure rates, but marginally missed significance on 3-week cure rates (OR: 0.27; 95% CI: 0.07–1.10%), which was the primary outcome of the Cochrane Review.

In brief, two of the four studies, one involving vaginally administered lactobacilli combined with estriol[118] and one study involving orally administered lactobacilli,[120] documented a highly beneficial effect on the resolution of BV at 21–30 days post-treatment, whereas in the other two studies involving the administration of intravaginal lactobacilli, the probiotic treatment was not significantly better compared with placebo[119] or compared with intravaginal metronidazole gel.[121] Therefore, based on a limited number of studies that were selected based on their level of evidence, it must be concluded that there is insufficient evidence available as yet to recommend the use of probiotics in addition to antibiotics in the treatment of BV, although the results from some studies were particularly promising.

Several, similarly promising studies have been published since we performed the systematic literature search within the framework of the Cochrane Review thereby including studies published before 2008. Larsson et al. randomized 100 women with BV in two groups of 50 and the participants were given a 7-day course of daily 2% vaginal clindamycin cream directly followed by vaginal gelatine capsules containing 108–9 freeze-dried lactobacilli (a L. gasseri and a L. rhamnosus strain) or placebo capsules of identical appearance, sufficient for 10 days or until menstruation commenced.[122] After each menstruation, treatment with vaginal lactobacilli capsules or placebo was repeated during 10 days for three cycles. Thus, the treatment regime included one treatment course with clindamycin followed by four lactobacilli/placebo courses; one within the same menstrual cycle and the other during the next three consecutive cycles. The initial intent-to-treat analysis for the 1-month cure rate was 64% in the lactobacilli group and 78% in the placebo group (p = 0.8). The 76 cured women were followed for six menstrual cycles or until relapse within that timespan. At the end of the study, 64.9% (24 out of 37) of the lactobacilli-treated women were still BV-free compared with 46.2% (18 out of 39) of the placebo-treated women (p = 0.042). Marcone et al. randomized 84 patients with BV in a single-blinded manner.[123] All patients received oral metronidazole 500 mg twice a day for 7 days, then followed in the treatment group by one vaginal tablet containing freeze-dried L. rhamnosus once a week at bedtime for 2 months starting 1 week after the last antibiotic administration. Follow-up was performed at days 30, 90 and 180, with cure rates in the treatment group of 88, 88 and 83% and in the control group of 81, 71 and 67% at 1, 3 and 6 months, respectively (p = 0.4, p= 0.05 and p = 0.07). In a rather small but well-designed double-blinded randomized, controlled trial, Mastromarino et al. randomized 39 women with BV to receive either a vaginal probiotic tablet (containing at least 109 viable lactobacilli, in particular a Lactobacillus brevis strain, a Lactobacillus salivarius subsp. salicinius strain, and a Lactobacillus plantarum strain) or an identical placebo for 7 days.[124] The 2-week cure rates were 61% (11 out of 18) in the active treatment group as compared with 19% (three out of 16) in the placebo group (p = 0.017).

Treatment of Recurrent BV

As apparent from the aforementioned information, the main challenge with BV remains the quest for long-term cure and hence the prevention of BV recurrence. Standard antibiotic treatment regimens are associated with high relapse rates, and we have very few options for dealing with relapsing BV. As outlined, several alternative treatment options, including long-term treatment with vaginal acidifying agents and probiotics, may offer promise for overcoming this longstanding frustration.

The most rigorously studied approach to the prevention of recurrent BV, however, certainly comes from a US multicenter trial in which women were initially treated during the open-labeled phase with a 10-day course of 0.75% metronidazole gel and in which asymptomatic responders were then randomized to receive twice-weekly 0.75% metronidazole vaginal gel or placebo for 16 weeks and off therapy for 12 weeks.[125] During suppressive therapy, recurrent BV occurred in 13 women (25.5%) receiving metronidazole and in 26 (59.1%) receiving placebo (RR: 0.43; 95% CI: 0.25–0.73%; p = 0.001). During the entire 28-week follow-up, recurrence occurred in 26 (51.0%) on treatment compared with 33 (75.0%) on placebo (RR: 0.68; 95% CI: 0.49–0.93%; p = 0.02). Adverse effects were uncommon, although secondary vaginal candidiasis occurred significantly more often in metronidazole-treated women.

In a recent study, Reichman et al. performed a retrospective analysis of a modified long-term suppressive triple-phase treatment scheme in which patients with recurrent BV were treated consecutively with 7 days of oral nitroimidazole, followed by 21 days of intravaginal boric acid 600 mg/day and if in remission then treated with metronidazole gel twice weekly for 16 weeks.[126] Cure after nitroimidazole and boric acid therapy ranged from 88 to 92%, 7 and 12 weeks after the initial visit, respectively. Cumulative cure at 12, 16 and 28 weeks from the initial visit was 87, 78 and 65%, respectively. A failure rate of 50% was documented by 36 weeks of follow-up. Although very encouraging, this approach requires validation in a prospective randomized, controlled study.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.