Bacterial vaginosis: a narrative review of concepts and best practices

Before addressing the main topic, it is necessary to examine key aspects of the microenvironment in which bacterial vaginosis (BV) develops. Significant distinctions exist between key terms in microbiology. Microbiota refers to the microorganisms, including bacteria, fungi, and viruses, that inhabit a specific environment. In contrast, the microbiome encompasses these microorganisms, their collective genetic material (metagenome), and the substances and functions they produce, thereby representing the entire microbial ecosystem and its activities ( 1 ). The vagina possesses a complex and dynamic microenvironment, including both the microbiota and microbiome, which is influenced by factors such as age, genetics, diet, medications, sexually transmitted microorganisms, and additional variables. Vaginal contents consist of the microbiome, cervical and vestibular secretions, epithelial cell shedding, immune cells, proteins, and transudate. Under physiological conditions, the vaginal epithelium is composed of nonkeratinized stratified squamous cells and lacks glands; consequently, there are no secretions originating from the vagina itself. Therefore, the appropriate term for the material present in the vagina is vaginal fluid or vaginal contents ( 2 ). Since the early studies conducted by Albert Doderlein, it has been widely accepted that lactobacilli constitute the primary bacterial population in the vagina of healthy adult women ( 3 ). In a recent landmark study, Ravel et al. ( 4 ) identified five distinct types of vaginal bacterial communities present in healthy women; these are referred to as community state types (CSTs). CST I is characterized by a predominance of Lactobacillus crispatus , CST II by a predominance of L . gasseri , CST III by L . iners , CST IV by bacteria typically associated with BV, and CST V by a predominance of L . jensenii . Subsequently, CST IV was divided into IV-A and IV-B according to the presence of lactobacilli associated with the BV ( 5 ). CST III ( L . iners ) and CST IV (bacteria associated with BV) are associated with a pro-inflammatory vaginal environment, driven by factors such as macrophage migration inhibitory factor, interleukin-1α, interleukin-18, and tumor necrosis factor-alpha. These factors activate inflammatory responses, causing the vagina to be more susceptible to dysbiosis. Conversely, CST I, characterized by a predominance of L . crispatus , is associated with high production of Dextrogiro-lactate (D-lactate) and bacteriocins, making this type a model of a healthy vagina (6-8). BV is defined as an imbalance in the vaginal bacterial population. Emerging studies are reshaping our understanding of BV, offering a richer definition that goes beyond the loss of lactobacilli and the rise of anaerobic bacteria. This condition commonly affects women of reproductive age ( 9 ). The prevalence of BV varies depending on the diagnostic tools used, affecting approximately 20–30% of women of reproductive age worldwide. BV represents a primary cause of the claim of vaginal discharge ( 10 ). When the vaginal microbiota is out of balance, it can raise the risk of sexually transmitted infections, cervical cancer, and problems during pregnancy ( 10 ). Although vaginosis has long been studied, recent discoveries have raised new questions that require thorough investigation. Current methodologies for studying bacterial populations should extend beyond pathogen identification to include the broader ecological context of these microorganisms. Such approaches must address their roles in the initiation and persistence of BV, as well as their impact on immune responses, susceptibility to sexually transmitted infections, cervical cancer, and overall reproductive health risks. The purpose of this narrative review was to provide an updated and comprehensive understanding of the pathophysiology of BV, as well as the groundbreaking advancements in diagnostic methods and treatment options for the infection. By examining the latest research and developments, we aimed to empower clinicians and researchers with the knowledge necessary to improve care and outcomes for patients with this prevalent condition. We present this article in accordance with the Narrative Review reporting checklist (available at https://gpm.amegroups.com/article/view/10.21037/gpm-2025-1-52/rc).

We searched MEDLINE, Google Scholar, and Scopus for original studies, clinical trials, systematic reviews, meta-analyses, narrative reviews, and integrative reviews published in English prior to September 2025. The keywords used for the search were [vaginal microbiota] OR [vaginal microbiome] OR [vaginosis] OR [bacterial vaginosis] OR [dysbiosis] AND [vagina] AND [diagnosis] AND [treatment]. Additionally, we reviewed the references of the articles identified in our literature searches ( Table 1 ). Texts written in languages other than English, case reports, and opinions were excluded from our literature search.

The vaginal microbiota is a crucial barrier against infections, and maintaining its balance is essential for this function. These beneficial bacteria create a biofilm on the epithelial surface and produce, or stimulate the production of, various antimicrobial components. These include lactic acid, hydrogen peroxide, bacteriocins, the enzyme arginine deaminase, lactoferrin, lysozyme, calprotectin, and cathelicidin LL-37 ( 9 ). Typically, the healthy vaginal microbiota is primarily composed of Lactobacillus species, including L . crispatus , L . gasseri , and L . jensenii . This composition helps to inhibit potentially pathogenic microorganisms, thereby maintaining a balanced microenvironment (11,12). The action of the D-form of lactic acid helps to maintain homeostasis. This form, present in greater quantities than the levogiro form (L-lactate), plays a crucial role in activating the appropriate defense mechanisms and maintaining a pH below 4.5 ( 13 ). In environments in which L . iners is predominant, the ratio of L-lactate is higher, with a greater expression of the extracellular matrix metalloproteinase inducer. Consequently, this can compromise the integrity of the epithelium, leading to a greater risk of dysbiosis and infection with sexually transmitted diseases ( 8 ). D-lactate inhibits pathogenic bacteria more effectively than the L-isomer ( 14 ). Research has shown that environments dominated by L . iners are less able to maintain the pH at less than 4.5 and to produce hydrogen peroxide ( 15 ). BV is a condition characterized by vaginal dysbiosis, in which Lactobacillus is reduced or absent, and anaerobic bacteria—primarily Gardnerella , Megasphaera , Sneathia , Prevotella , Fannyhessea , Ureaplasma , Dialister , and Mobiluncus —are increased (16,17). BV is the most commonly diagnosed condition among women of reproductive age seeking gynecological care. Its prevalence varies worldwide, ranging from 11% to 60%, and is notably higher among black and latina women. In the United States, this condition leads to direct costs of approximately $4.8 billion for diagnosing and treating symptomatic cases; this figure triples when the complications associated with dysbiosis are considered ( 18 ). This condition can present with or without symptoms, and the bacterial composition appears to differ accordingly. The predominance of Gardnerella , Fannyhessea , and Sneathia characterizes asymptomatic cases ( 19 ). This profile clearly indicates a high risk of spontaneous abortions, preterm births, infertility, and cancer, irrespective of the clinical complaints ( 20 ). Of note is that BV occurs more frequently among women who have a history of multiple sexual partners, use vaginal douches, have uncircumcised partners, and utilize intrauterine devices, particularly those that are copper (21-23). Approximately 50% of women with BV experience no symptoms. When symptoms do occur, the most common complaints include a grayish vaginal discharge and an unpleasant odor during intercourse and menstruation. The odor results from the increased growth of anaerobic bacteria, such as Gardnerella and Prevotella species, leading to the breakdown of vaginal peptides. This metabolic process produces volatile biogenic amines (including trimethylamine, putrescine, and cadaverine) ( 24 ). Complications caused by dysbiosis arise from the dysbiosis itself, rather than from any of the specific complaints that the woman presents with ( 25 ). The Amsel criteria ( 26 ), which require the identification of at least three of four findings—homogeneous vaginal discharge, pH greater than 4.5, a positive amine test using 10% KOH, and the presence of clue cells in a vaginal smear—and the Nugent Score ( 27 ) have traditionally been used for diagnostic purposes. The Nugent scoring system is based on the identification of bacterial morphotypes in Gram-stained vaginal smears. Each morphotype is assigned a specific point value, and the total score is calculated accordingly. A score ranging from 7 to 10 is indicative of BV and is widely regarded as the gold standard for its diagnosis ( Figure 1 ). Nugent score suggestive of normal (A); intermediate microbiota (B); and bacterial vaginosis (C) (Gram; 1,000×). Although the Pap smear is not primarily designed to diagnose BV, in many countries, particularly those with limited resources, it remains the sole method for diagnosing vaginitis and vaginosis, in addition to its use in cervical cancer screening. The accuracy of identifying BV in both conventional and liquid-based Pap smears varies when compared to that of the Nugent score. The cytological characteristics typically show a lack of inflammatory cells, but the microbiota appears as a dust cloud with clue cells present ( Figures 2,3 ) (28,29). Pap smear for a case of bacterial vaginosis (Pap; 400×). Liquid-based cytology for a case of bacterial vaginosis (SurePath; 1,000×). Another method for diagnosing BV is the Osom BVBlue Test (Sekisui Diagnostics LLC, Burlington, MA, USA), which has been demonstrated to have an agreement with the Nugent score of more than 98% ( 30 ). Point-of-care molecular biology techniques, such as BD Affirm VPIII ® (Becton, Dickinson and Company, Becton Drive, Franklin Lakes, New Jersey, USA) with a reported sensitivity of 83% ( 31 ), should be systematically evaluated in comparison with clinical findings. These methods detect only G . vaginalis , which alone is insufficient for a definitive diagnosis of BV. Furthermore, a negative result does not exclude the possibility of BV. Methods utilizing molecular biology techniques on multiplex platforms effectively detect specific nucleic acids from bacteria, and exhibit high sensitivity and specificity; such platforms include a BV panel that identifies individual bacteria, including Gardnerella vaginalis , Fannyhessea vaginae , Bacteroides fragilis , Mobiluncus spp., and others ( 32 ). Routine diagnosis of BV remains complex, primarily reliant on Amsel criteria and Nugent scoring in the majority of services. But, accurate detection of the involved dysbiotic microbiota could help guide personalized management. Molecular methods targeting relevant biomarkers—including nucleic-acid amplification tests (NAATs)—are being integrated into clinical workflows. NAATs demonstrate sensitivity ranging from 83% to 96% and specificity from 88% to 98% ( 33 ). Comparative analysis of these diagnostic methods provides insight into asymptomatic carriage at the population level and within-host recurrence. Findings from Amsel, Nugent, and DNA-based systems suggest that, given the heterogeneity of individual patient histories, no single test adequately addresses all aspects of BV ( Table 2 ) (30-33). Gold standard: Nugent score. BV, bacterial vaginosis. BV is significantly associated with gynecological diseases and sexually transmitted infections, including human immunodeficiency virus ( 34 ). Notably, BV increases the risk of human papillomavirus (HPV) infection as well as the development of precancerous and cancerous cervical lesions (35,36). In an extensive retrospective cohort study involving more than 6,000 women, a two-way pathway in the interrelationship between dysbiosis and HPV was identified ( 37 ). Using pyrosequencing techniques, a greater bacterial diversity and reduced number of lactobacilli was confirmed among women with genital HPV infection than among a group of women who tested negative for the virus ( 38 ). HPV appears to inhibit the basal expression of the pro-inflammatory defense peptides of the host by subverting the NF-κB (Nuclear factor kappa-light-chain-enhancer of activated B cells) and Wnt/β-catenin pathways. As a result of the immune evasion activity of the HPV, the lactobacilli decrease, enabling the virus to evade the immune system ( 39 ). BV has been associated with an increased risk of infertility and endometriosis, although the evidence supporting this link is weak (40,41). Furthermore, BV is linked to poor obstetric outcomes, including miscarriage, premature birth, and low birth weight (42-46). The mechanisms involve a microbiota dominated by Leptotrichia and or Sneathia , bacterial vaginosis-associated bacterium 1 (BVAB1), or the Megasphaera phylotype, which promote an inflammatory response characterized by increased production of cytokines, chemokines, prostaglandins, metalloproteinases, and sialidase (45,46). In BV, a polymicrobial biofilm is formed that protects dysbiosis from the immune system, complicating treatment ( 47 ). The mechanisms and bacteria involved in the formation of the biofilm are still being studied, but initial research findings suggest that G . vaginalis and Prevotella bivia play significant roles in promoting initial colonization, whereas F . vaginae is thought to contribute in a secondary capacity ( 48 ). The process begins with a virulent strain of G . vaginalis , leading to a reduction in Lactobacillus and the formation of the biofilms. This initiates proteolysis and promotes the growth of P . bivia , which degrades the mucin layer in the epithelium. As a result, the adhesion of bacteria associated with BV increases, facilitating the formation of a polymicrobial biofilm ( 11 ). Given the complexities associated with BV, a crucial question is whether screening is necessary, especially considering that at least half of the cases are asymptomatic. Currently, the United States (US) Preventive Services Task Force does not recommend screening for BV, including in pregnant women ( 49 ). No consensus has yet been reached, but some advocate for screening during the first trimester of pregnancy ( 50 ). In addition to alleviating symptoms, treatment aims to lower the risk of being infected with sexually transmitted diseases and developing complications related to BV. The standard treatment recommended by the US Centers for Disease Control and Prevention is ( 51 ): Metronidazole 500 mg orally 2 times per day for 7 days, or Metronidazole gel 0.75% one full applicator (5 g) intravaginally, once a day for 5 days, or Clindamycin cream 2% one full applicator (5 g) intravaginally at bedtime for 7 days. The following alternative schemes can also be considered: Clindamycin 300 mg orally 2 times per day for 7 days, or Clindamycin ovules 100 mg intravaginally once at bedtime for 3 days, or Secnidazole 2 g oral granules in a single dose, or Tinidazole 2 g orally once daily for 2 days or Tinidazole 1 g orally once daily for 5 days. Only metronidazole and clindamycin vaginal cream are approved for use during pregnancy. A recent study suggested the possibility of treating partners ( 52 ). However, the follow-up period was short, and the criterion used was a Nugent score of 4 or greater. Therefore, the evidence to demonstrate any clear benefit is insufficient. Biofilm is a significant factor in treatment failures because it increases the tolerance of pathogens to antimicrobial agents. This occurs due to the slow diffusion of these drugs through the biofilm, difficulty in reaching the deeper layers, reduced metabolic activity of the cells, and presence of resistant anaerobic bacteria within the biofilm ( 53 ). Research indicates that the simultaneous presence of F . vaginae and Gardnerella spp. in the biofilm may be linked to a higher rate of treatment failure in BV ( 54 ). The use of probiotics, both vaginally and orally, has been suggested as an adjunct to treatment, but studies validating their effectiveness remain lacking (53,55). Probiotics are a safe, potentially beneficial adjunctive therapy for BV, especially for preventing recurrence, but more research is needed for precise guidelines ( 55 ). Biofilm breakers (as boric acid) are increasingly used as adjunctive treatments for BV to address the protective polymicrobial biofilm, which contributes to high recurrence rates (up to 80% within a year) following standard antibiotic therapy. Another adjuvant approach to the treatment of BV involves pH modulation, which may be achieved with vitamin C or lactic acid ( 56 ). Consequently, the following recommendations have been proposed for the management of persistent and recurrent cases ( 56 ): Oral metronidazole 1 g daily for 14 days, followed by vaginal boric acid 600 mg daily for 3 weeks, then vaginal metronidazole 500 mg twice weekly for 4 to 6 weeks, or Vaginal metronidazole 750 mg daily for 4 to 6 months, or Vaginal clindamycin cream 2% daily for 7 days, followed by vaginal boric acid 600 mg daily for 3 weeks. Boric acid is contraindicated during pregnancy, whereas oral metronidazole is considered safe for use in pregnant women. During breastfeeding, the potential consequences of oral antibiotic use remain uncertain; therefore, vaginal treatment is generally considered preferable (32,51). The frequency of therapeutic failure and relapse is high ( 57 ); therefore, we believe that conducting an evaluation approximately 30 days after completing treatment is essential, even though no specific guidelines recommend this. Assessment 30 days after treatment completion, using the Nugent score or point-of-care tests, enables identification of immediate treatment failure and differentiation from recurrence. When available, biomolecular methods applied at 6 months can detect recurrent cases, which become increasingly common thereafter ( 11 ).

BV is a prevalent condition that, despite its widespread occurrence, remains poorly understood in terms of its causes. The management of BV should be adapted to the specific circumstances of each setting. Diagnostic methods range from the use of Amsel criteria and Nugent scoring to advanced bacterial profiling with pyrosequencing techniques. Obtaining an accurate diagnosis in each clinical context is critical to ensure appropriate treatment, which may include antibiotics, biofilm disruptors, pH modulators, and potentially probiotics.

The article’s supplementary files as