Sexual Health and Genital Medicine in Clinical Practice

Chapter 4. Bacterial Vaginosis

Bacterial vaginosis (BV) is more common than "thrush" and is probably the commonest cause of abnormal vaginal discharge seen in primary care. The condition is certainly underdiagnosed and frequently misdiagnosed. BV was formerly known as "Gardnerella" and is caused by an overgrowth of predominantly anaerobic bacterial species which are commonly present in low concentrations in a healthy vagina (e.g. Gardnerella vaginalis, Prevotella spp., Peptostreptococcus, Mobiluncus, Mycoplasma hominis).

Although many clinicians regard BV as a fairly insignificant condition, this is certainly not the case for the majority of sufferers. Many women find the amount of discharge, and in some cases the associated malodor, to be particularly distressing. In addition, there is increasing evidence that BV is associated with preterm labor, late miscarriage, chorioamnionitis, postpartum endometritis and bacteremia, pelvic infection following surgery and termination of pregnancy, and, possibly, PID.


The commonest presenting symptom is excessive vaginal discharge, sometimes with a slight malodor. Some women regard a fishy vaginal odor as normal and are surprised and grateful when BV is eventually diagnosed and treated. Malodor may only be noticeable after unprotected sexual intercourse, owing to the release of amines by alkaline semen (see "Amine test" below). Vulval irritation is uncommon. As with candidiasis, many women with BV are asymptomatic.


The following are the two most important methods of diagnosis.

(1) Microscopy of vaginal secretions

BV produces a highly characteristic appearance on Gram staining. There is an absence of lactobacilli and an excess of Gram-variable or Gram-negative rods (Gardnerella, Prevotella, Peptostreptococcus; Figures 4.1 and 4.2). In some cases Gramnegative "curved rods" (Mobiluncus) may be seen. As vaginal inflammation (vaginitis) is not a feature of BV, few polymorphs are present.

(2) Amine test

This test involves the addition of two drops of 1-5% potassium hydroxide solution to a sample of the vaginal secretions, either on a slide or on a swab. The sudden release of a fishy odor represents a "positive" result. The odor results from volatilization of polyamines, in particular trimethylamine, that are thought to be produced by the anaerobic bacteria.

Compared with microscopy, the "amine test" has a sensitivity of 80-90% and a specificity of well over 90%. The test is easy, quick, and inexpensive to perform and should be part of the initial assessment of all women with vaginal discharge.

Although the amine test may be performed on air-dried swabs some hours or days later, the main advantage of the test is that it can be performed during the consultation. The odor produced is short lasting and, despite some claims to the contrary, does not linger in the room where the test is performed. Testing is probably best performed out of sight of the patient.

Figure 4.2. Gram stain of vaginal secretions showing lactobacilli

The following are other diagnostic criteria mentioned in the textbooks but less helpful than microscopy and amine testing.

(3) Vaginal pH

In BV, vaginal pH is raised from the normal value of 4.5 to above 5.0. Unfortunately, this is not specific and probably signifies simply a reduction in the number of lactobacilli. (Lactobacilli are the predominant bacterial species in the healthy vagina and maintain a protective acid environment in the vagina by, we think, producing lactic acid from vaginal glycogen.) In addition, a raised pH may be found in a woman with a normal vaginal flora if testing is performed when menstrual blood or semen is present or if cervical mucus is inadvertently sampled instead of vaginal secretions. BV, however, is very unlikely to be present if the pH is normal.

(4) Appearance of the discharge

Although the vaginal discharge in BV is classically thin, homogeneous with a creamy or milky consistency and a slight froth (Figure. 4.3), this is by no means always the case, and in most studies the appearance of vaginal fluid has been shown to be a poor diagnostic marker.

High vaginal swab culture has no place in diagnosis because the presence of Gardnerella vaginalis or anaerobes does not necessarily indicate the presence of bacterial vaginosis. Quantitative culture may be helpful but it is difficult to perform. As mentioned above, microscopy is the diagnostic test of choice.

Figure 4.3. Bacterial vaginosis - creamy, homogeneous discharge coating the vaginal wall


There are a few options available.

 Oral metronidazole is an extremely effective treatment and various regimens have been used: 2 g suspension stat dose; 400-500 mg bd for 5 days; 200 mg tds for 7 days.

 Intravaginal metronidazole gel (0.75%) daily for 5 days.

Patients should be advised to avoid alcohol whilst taking metronidazole (possibly also when used intravaginally) owing to a disulfiram effect.

 Intravaginal clindamycin cream (2%) daily for seven days is a useful alternative for patients who cannot tolerate metronidazole (worth mentioning to patient that clindamycin cream may weaken condoms).

 Oral clindamycin 300 mg bd for 7 days.

Treatment is currently reserved for women with symptoms.

A case could be made for treating asymptomatic women prior to hysterectomy, endometrial biopsy, termination of pregnancy, dilatation and curettage (D & C), and intrauterine contraceptive device (IUCD) insertion, although there are no studies supporting this for the latter two procedures. The possibility of inoculating the uterus with bacteria capable of causing endometrial infection does lend support to the suggestion that BV should be treated prior to any procedure involving instrumentation through the cervix.

Trials are ongoing to assess whether treating BV in pregnancy reduces the risk of preterm labor. Studies performed to date have given conflicting results; however, some of those failing to show a benefit have been criticized regarding the timing of diagnosis. Diagnosing and treating BV very early in pregnancy may be important. It is reasonable to say that on current evidence treatment should be considered in pregnant women with a past history of preterm labor of uncertain cause or late miscarriage, and that treatment should be given as early as possible in the pregnancy. The drugs recommended for treating BV are safe to use in pregnancy. Meta-analyses show no evidence of teratogenicity with metronidazole.


A recurrence is seen in about 20% of women after treatment irrespective of the drug used. This is often a "bacteriological" recurrence (i.e. BV is diagnosed on microscopy) rather than symptomatic recurrence. However, some women do experience frequent symptomatic recurrences which, as with recurrent candidiasis, often affect sexual relationships and cause a degree of psychological morbidity. Treating sexual partners has been shown to have no effect on reducing the recurrence rate. There is an association between BV and the IuCD and in women with particularly troublesome recurrences an alternative form of contraception should be considered. using condoms for a few months may prove beneficial for some patients.

A short course (2-3 days) of oral metronidazole or intravaginal clindamycin or metronidazole once or twice a month may also be worth considering as a prophylactic measure (the necessary studies are awaited). In addition, recent reports have suggested benefit from using an intravaginal acid gel after standard treatment with metronidazole.

If you find an error or have any questions, please email us at Thank you!