How is Bacterial Vaginosis (BV) diagnosed?

Diagnosis of Bacterial Vaginosis

Bacterial vaginosis should be diagnosed using the Amsel clinical criteria, requiring 3 of 4 findings: homogeneous white discharge coating vaginal walls, vaginal pH >4.5, positive whiff test (fishy odor with KOH), and presence of clue cells on microscopy. 1

Primary Diagnostic Approach: Amsel Clinical Criteria

The CDC-endorsed Amsel criteria remain the gold standard for BV diagnosis in clinical practice and require meeting three of the following four criteria 2, 1:

• Homogeneous, white, noninflammatory discharge that smoothly coats the vaginal walls (not clumpy or thick) 2, 1
• Vaginal pH greater than 4.5 measured using narrow-range pH paper 2, 1
• Positive whiff test: fishy amine odor detected when 10% KOH is added to vaginal discharge 2, 1
• Clue cells present on microscopic examination of wet mount (vaginal epithelial cells with adherent bacteria obscuring cell borders) 2, 1

This clinical approach has high specificity (>92%) and reasonable sensitivity (61%) when compared to microbiota analysis 3. The Amsel criteria can be performed at point-of-care without specialized laboratory equipment, making it practical for most clinical settings 1.

Alternative Diagnostic Method: Gram Stain (Nugent Score)

Gram stain evaluation is an acceptable alternative laboratory method that determines the relative concentration of bacterial morphotypes characteristic of BV 2, 1. This method:

• Assesses the ratio of Lactobacillus morphotypes (large gram-positive rods) to Gardnerella morphotypes (small gram-variable rods) and other anaerobic bacteria 4
• Has comparable sensitivity (64%) to Amsel criteria when compared to microbiota analysis 3
• Shows excellent inter-evaluator agreement (≥90%) 4
• Is more objective than clinical criteria but requires laboratory expertise 1

A Gram stain showing mixed flora with decreased or absent Lactobacillus morphotypes (0 to 2+) is consistent with BV 4.

What NOT to Use for Diagnosis

Culture for Gardnerella vaginalis should NOT be used diagnostically because it lacks specificity—G. vaginalis can be isolated from vaginal cultures in half of normal women without BV 2. This is a common pitfall to avoid in clinical practice.

Point-of-Care Testing Options

For settings without microscopy capabilities, several alternatives exist 1, 5:

• pH testing alone can be performed on discharge collected from the introitus without speculum examination 5
• Rapid sialidase enzyme tests (like BVBlue) have high specificity (98-100%) and can provide presumptive diagnosis 6
• Molecular NAAT panels are increasingly available and may detect BV with higher sensitivity (81%) than traditional methods 3

The AmpliSens qPCR assay showed the highest agreement with microbiota analysis (81% sensitivity, >92% specificity) in comparative studies, though this represents emerging technology not yet widely adopted 3.

Critical Clinical Considerations

Up to 50% of women meeting diagnostic criteria for BV are asymptomatic 2, 1. This has important implications:

• Symptomatic women always require treatment regardless of how diagnosis was made 1, 5
• Asymptomatic screening is recommended before invasive gynecologic procedures (surgical abortion, hysterectomy, IUD placement, endometrial biopsy) because treatment substantially reduces post-procedure pelvic inflammatory disease 1, 5
• High-risk pregnant women (those with prior preterm delivery) should be evaluated for BV even if asymptomatic 2

Common Diagnostic Pitfalls to Avoid

• Do not rely on vaginal discharge appearance alone—half of BV cases are asymptomatic 1
• Do not culture for G. vaginalis—it is present in 50% of women without BV 2
• Do not overlook pH testing—it is simple, inexpensive, and can be done without speculum examination 5
• Do not assume normal pH rules out BV—some cases may have borderline pH values, so use all four Amsel criteria 2

Underlying Pathophysiology Relevant to Diagnosis

BV represents replacement of normal H₂O₂-producing Lactobacillus species with high concentrations of anaerobic bacteria (Prevotella, Mobiluncus), G. vaginalis, and Mycoplasma hominis 2, 1. This shift explains the diagnostic findings:

• Loss of Lactobacillus → elevated pH (>4.5) 1
• Anaerobic metabolism → production of amines → positive whiff test 2
• Bacterial adherence to epithelial cells → clue cells on microscopy 2

Understanding this pathophysiology helps clinicians recognize that BV is a polymicrobial syndrome, not a single infection, which is why culture for individual organisms is not diagnostically useful 2.