Vaginosis Bakterialis

01Pathophysiology

Bacterial vaginosis (BV) is fundamentally a disruption of the normal vaginal microbiome. Under healthy conditions, the vaginal flora is dominated by hydrogen peroxide-producing Lactobacillus species (primarily L. crispatus and L. jensenii). These organisms maintain vaginal pH between 3.8 and 4.5, creating an acidic environment that suppresses the growth of pathogenic and opportunistic anaerobes.

In BV, there is a dramatic shift from Lactobacillus dominance to a polymicrobial anaerobic overgrowth. Gardnerella vaginalis serves as the initial colonizer, forming a dense biofilm on the vaginal epithelium. This biofilm provides a scaffold for other anaerobes such as Prevotella, Mobiluncus, and Atopobium vaginae to co-colonize and proliferate. The loss of lactobacilli removes the protective acid barrier, allowing vaginal pH to rise above 4.5.

The anaerobic bacteria produce amino acid decarboxylases, enzymes that break down peptides and proteins into volatile amines, specifically trimethylamine and putrescine. These amines are responsible for the characteristic fishy odor. The odor becomes more pronounced when the discharge is exposed to alkaline conditions, which is why it intensifies after unprotected intercourse (semen is alkaline, pH around 7.2) and during menses (blood is alkaline relative to vaginal pH). This is the biochemical basis of the positive whiff test when potassium hydroxide (KOH) is added to the discharge.

Clue cells, the hallmark microscopic finding, are vaginal squamous epithelial cells whose surfaces are so densely coated with adherent bacteria (predominantly Gardnerella and anaerobes) that their borders become obscured and stippled. The cell edges appear granular rather than sharply defined. This finding directly reflects the biofilm-forming nature of the causative organisms.

Critically, BV is not a true infection in the inflammatory sense. There is no significant host inflammatory response, which is why the vaginal mucosa shows no erythema, no edema, and no pain. The white blood cell count on wet mount is characteristically low. This distinguishes BV from true vaginitis caused by Trichomonas or Candida, where inflammation is prominent.

BV during pregnancy carries clinical significance because the ascending anaerobic organisms can cause chorioamnionitis, preterm premature rupture of membranes, and preterm labor. It is also associated with increased susceptibility to sexually transmitted infections, including HIV, gonorrhea, chlamydia, and herpes simplex virus, due to the loss of the protective Lactobacillus barrier and alteration of local immune defenses.

02Classification and Clinical Manifestation

Diagnostic Classification (Nugent Scoring System)

Clinical Presentation

03Diagnostic Workup

Diagnostic Summary Table

The best initial diagnostic approach in clinical practice is the Amsel criteria, which can be performed at the bedside during a pelvic examination. A diagnosis of BV requires at least 3 of the following 4 criteria:

1. Thin, homogeneous, grayish-white discharge that smoothly coats the vaginal walls

2. Vaginal pH greater than 4.5, measured by touching pH paper to the vaginal sidewall (not the cervix, which has a naturally higher pH and would give a false reading)

3. Positive whiff test: a fishy odor released when 10% potassium hydroxide (KOH) is added to a sample of the discharge, caused by volatilization of amines

4. Clue cells on saline wet mount, comprising at least 20% of the epithelial cells observed

The most accurate (gold standard) test is the Nugent score, a standardized Gram stain scoring system that quantifies Lactobacillus, Gardnerella, and Mobiluncus morphotypes on a 0 to 10 scale. A score of 7 or above confirms BV. This test is primarily used in research settings and reference laboratories because it requires trained microscopists. For exam purposes, recognize Nugent scoring as the confirmatory test when the question stem asks for the "most accurate" method.

An important exam pearl: cultures for Gardnerella vaginalis are NOT useful for diagnosing BV. Gardnerella can be found as a commensal in up to 50 to 60% of healthy women without BV. A positive culture does not distinguish colonization from disease and should never be ordered.

Molecular/NAAT-based point-of-care tests (e.g., BD MAX BV Panel, Aptima BV) are increasingly available and detect nucleic acids of BV-associated organisms. These are highly sensitive and are used in settings where microscopy is not available.

A finding that is notably absent in BV and helps distinguish it microscopically: white blood cells are NOT elevated on wet mount. If the wet mount shows abundant neutrophils, consider Trichomonas vaginalis or cervicitis from gonorrhea/chlamydia rather than BV.

04Management and Treatment

Detailed Management

Acute treatment centers on metronidazole 500 mg orally twice daily for 7 days, which remains the most studied and most frequently tested regimen. Patients must be counseled to avoid alcohol during treatment and for 48 hours after completion due to a disulfiram-like reaction (nausea, vomiting, flushing, headache). This is a high-yield exam point: the question may describe a patient on metronidazole who develops flushing and vomiting after drinking alcohol.

Tinidazole is a second-generation nitroimidazole with a longer half-life and improved GI tolerability compared to metronidazole. It can be given as a shorter course (2 g daily for 2 days) but is more expensive.

Intravaginal clindamycin cream is an effective alternative but carries an important caveat: it is oil-based and can weaken latex condoms and diaphragms for up to 5 days after use. This must be disclosed to the patient.

Treatment of sexual partners is NOT recommended. Multiple randomized controlled trials have shown that treating male sexual partners does not reduce recurrence rates in women. BV is not classified as a sexually transmitted infection, though sexual activity is a risk factor. This is a commonly tested point: the question may offer partner treatment as an answer choice, and it is incorrect.

Asymptomatic BV generally does not require treatment in non-pregnant women. However, treatment is recommended before gynecologic surgical procedures (hysterectomy, endometrial biopsy, IUD insertion, abortion) to reduce the risk of postoperative infectious complications.

In pregnancy, treatment of symptomatic BV is recommended to alleviate symptoms and reduce obstetric complications. Screening and treatment of asymptomatic BV in pregnant women is currently recommended only for those with a prior history of preterm birth. Oral metronidazole is preferred over intravaginal preparations in pregnancy because systemic treatment more effectively reaches the upper genital tract where ascending infection causes complications.

Recurrent BV (three or more episodes per year) is common, with recurrence rates approaching 50% within 12 months of treatment. After completing a standard treatment course, suppressive therapy with metronidazole vaginal gel twice weekly for 4 to 6 months has been shown to reduce recurrence.

05Differential Diagnosis and Distractors

06Traps and High-Yield Pearls

The most common way students lose points on BV questions is by confusing BV with Trichomonas vaginalis. Both conditions elevate vaginal pH above 4.5 and can produce a positive whiff test. The critical differentiator the test-writer embeds in the stem is the presence or absence of inflammation: BV is a non-inflammatory condition (no erythema, no pruritus, no WBCs on wet mount), while Trichomonas triggers a brisk inflammatory response. If the vignette describes pruritus, erythema, or a frothy greenish discharge, it is not BV regardless of the pH.

A second trap involves ordering a vaginal culture for Gardnerella. The test-writer may include this as a tempting answer choice, but it is never correct because Gardnerella is a normal commensal in many women. The correct diagnostic approach is Amsel criteria (bedside) or Nugent scoring (laboratory).

Students also err by recommending treatment of the male sexual partner. Unlike Trichomonas or Chlamydia, there is no evidence that partner treatment improves outcomes in BV. If the question asks for the next step after diagnosing BV, the answer is treating the patient alone.

Another tested concept is the metronidazole-alcohol interaction. A vignette describing nausea, vomiting, and facial flushing in a woman recently started on metronidazole who attended a social event is testing your knowledge of the disulfiram-like reaction.

Finally, recognize that BV in pregnancy is linked to preterm labor and premature rupture of membranes. If a pregnant woman presents with preterm contractions and a workup reveals BV, the question is testing whether you know to treat with oral metronidazole (safe in pregnancy, including the first trimester) and that intravaginal therapy is less preferred in pregnancy because it does not adequately treat the upper genital tract.

The core competency being tested across BV questions is pattern recognition of a non-inflammatory vaginal condition (no itch, no redness, no pain) with a characteristic fishy odor and clue cells, and the ability to distinguish this presentation from inflammatory causes of vaginitis.