Etiology, Diagnosis, and Management of Vaginitis

Jane Mashburn, CNM, MN, FACNM


J Midwifery Womens Health. 2006;51(6):423-430. 

In This Article

Bacterial Vaginosis

BV is the most common cause of vaginal discharge,[5] and one of the most prevalent lower genital tract infections in women of reproductive age.[6] The prevalence is reported to be between 25% and 36% in women attending sexually transmitted infection clinics.[7] A substantial number of women are asymptomatic. Among women with BV who report symptoms (an estimated 10%–66%), vaginal malodor is the most suggestive symptom.[8,9] This condition is not associated with vaginal inflammation (such as finding excess leukocytes in the discharge or vaginal wall erythema), thus the term "vaginosis" is used instead of "vaginitis."[10] In fact, a finding of more than one leukocyte per epithelial cell on a microscopic evaluation of vaginal discharge should lead the provider to look for a diagnosis other than BV.[4]

BV has been associated with many complications, including second trimester miscarriage, pelvic inflammatory disease, preterm birth, preterm premature rupture of the membranes, chorioamnionitis, postpartum endometritis, postoperative infection after gynecologic surgery, and easier acquisition of HIV.[6,11,12,13,14]

The normal vagina of a woman of reproductive age is colonized with lactobacilli. These lactobacilli produce bacteriocins, hydrogen peroxide, and lactic acid, all of which are substances that lower the vaginal pH. The low pH creates a hostile environment for bacteria other than lactobacilli. If the number of lactobacilli are decreased, the resulting increase in pH favors an overgrowth of anaerobic and facultative bacteria, which can predispose to the development of BV.[11] The predominant organisms that cause BV are Gardnerella vaginalis, Mycoplasma hominus, and Ureaplasma urealyticum. Other anaerobes, such as Prevotella, Mobiluncus, Bacteroides, and Peptostreptococcus, have also been identified as flora associated with BV.[1,11]

Risk factors for BV have been identified and include douching and race, among others ( Table 1 ).[5,11,15,16] Zhang et al.[17] found in increased rate of BV in women who douched one or more times per week when compared to those who douched less frequently or not at all. Schwebke[18] studied douching and the prevalence of BV among predominately black adolescents who douched regularly. BV was diagnosed in 44.8% of the study population. He found that douching in the previous week was positively associated with BV and that there was a strong association between douching after menses and the prevalence of BV. The reason for these associations is not known. It is possible that douching decreases lactobacilli, thus facilitating the growth of the bacteria known to cause BV, but findings have been inconsistent. It is also possible that douching during the time of unstable flora, such as the time after menses, may lead to the acquisition of BV.[17] Even though several studies have shown that frequent douching is associated with BV, it is unknown if the douching causes the BV or if the women douche as self treatment for the symptoms.[16]

Race is also implicated as a risk factor for BV.[10,11,12] Klebanoff et al.[5] evaluated more than 1000 women diagnosed with BV and found it more common in young women who are African American and those who smoke. Poor general health status was associated with an increased prevalence of BV. Others have reported that African American women have a greater than 2-fold higher rate of having BV than white women.[13]

Many women who meet the laboratory criteria for a diagnosis of BV are asymptomatic. Although a malodorous vaginal discharge described as "fishy" suggests the presence of BV, it is not reliable enough to use as the only criterion for diagnosis. Some women with BV may report abnormal discharge, but this symptom is also unreliable.[5] Klebanoff et al.[5] studied women between the ages of 15 and 44 who attended 12 specific health departments for a routine annual health assessment. All participants underwent a pelvic exam that included tests for pH, wet mount, Gram stain, gonorrhea, and chlamydia. In addition, the participants were interviewed about vaginal symptoms. Reports of symptoms were compared between those who had a positive diagnosis of BV and those who did not. Their results showed that 82% of women without BV reported never noticing any vaginal odor compared to 75% of women with confirmed BV who reported noticing no odor. The authors concluded that although more women with BV report vaginal odor than do women without BV, the difference is minimal, and reliance on this symptom is not clinically valuable.

No single symptom has enough predictive power to accurately diagnose any of the common infections. Although signs and symptoms can assist in the diagnosis, the wet mount with microscopy is the best way to confirm the diagnosis.[4] Studies have shown that a diagnosis of BV based on symptoms alone is often inaccurate compared to diagnosis based on laboratory criteria.[18,19] In addition, basing diagnosis solely on symptoms will miss those women who would have laboratory-confirmed diagnosis of the infection but are asymptomatic.[7]

Amsel's criteria have been used for making the diagnosis of BV for many years. The four diagnostic criteria are: 1) a vaginal fluid pH >4.5; 2) >20% of epithelial cells are "clue" cells (cells with unclear borders, dotted with bacteria); 3) milky homogenous, adherent vaginal discharge; and 4) a positive "whiff" test, which is an amine or "fishy" odor noted after the addition of 10% potassium hydroxide. The presence of three out of four criteria are recommended by Amsel for diagnosis.[20,21] If there is no microscope available or if the skills of the practitioner are limited, the Affirm VPIII microbial identification system (Beckon Dickinson and Company, Sparks, MD) can be used. This DNA probe system can be used to identify BV, trichomoniasis, or candida.[22] This assay detects clinically significant levels of Gardnerella, trichomonads, and candida from vaginal fluid. Test results can be available in less that an hour. For offices that need to delay transport of the specimens, there is an extended transport system, Affirm VPIII Ambient Temperature Transport System (ATTS), which allows the specimen to remain stable in ambient temperature up to 72 hours after collection. Each of the Affirm VPIII kits come with sample swabs and transport collection tubes. Another test that can be used is the QuickVue Advance pH and Amines Test card (Quidel Corporation, San Diego, CA). A drop of vaginal secretion is placed on the card for rapid identification of pH. The sensitivity of this test is 94% when compared to culture.[4]

A newer, easy test for detecting BV is OSOM BVBlue (Genzyme Corporation, Cambridge, MA). The kit comes with sample swabs and reagent. The vagina is swabbed and the swab placed into the test tube with the reagent. The reagent turns blue or green if the sample is positive for BV.[23] This is a 10-minute test that detects elevated sialidase activity in the vaginal fluid. The sialidase is produced by bacterial pathogens associated with BV, including Gardnerella and Mobiluncus.[23]

A pH >4.5 is considered the most sensitive criterion (fewest false negatives) and the finding of >20% of epithelial cells as clue cells is the most specific (fewest false positives).[7] Gutman et al.[24] conducted a study to determine if two criteria would be as reliable as three of the four criteria, described by Amsel. They calculated the sensitivity, specificity, and 95% confidence intervals for each criterion separately and in combinations of two. The vaginal pH >4.5 had the highest sensitivity and lowest specificity. This low specificity may be explained by the fact that the elevated pH is often found in postmenopausal women and may be altered by cervical mucus, blood, or semen. On the other hand, a positive whiff test had the highest specificity and lowest sensitivity. Sensitivity of any two criteria ranged from 61% to 69%, which is similar to the sensitivity of 69% using Amsel's three out of four criteria. Specificity of any two criteria was between 86% and 95%, and similar to Amsel's. The authors conclude that using two criteria is as good as using three of Amsel's criteria. They concluded that if the pH is ≥4.5 and one other of Amsel's criteria is positive, the diagnosis of BV can be made.

According to Soper,[4] a normal pH (<4.5) should rule out the diagnosis of BV and the provider should then look for or rule out a fungal infection. If the pH is ≥4.5, one should suspect BV, trichomoniasis, or mucopurulent cervicitis, which is also associated with an elevated pH. If there is more than one leukocyte per epithelial cell in a microscopic evaluation of the vaginal fluid, one should consider either trichomoniasis or mucopurulent cervicitis.[4] Because mucopurulent cervicitis is often associated with gonorrhea or chlamydial trachomatis, testing for these infections should be done.

The Centers for Disease Control and Prevention (CDC) recommends testing all symptomatic women and treating those who are positive for BV. In addition, those who are to undergo surgical abortion or hysterectomy may be screened and treated. Follow-up visits or treatment of partners is not recommended. Treatment of partners has not been found to decrease the incidence of recurrent BV.[25] A review of six randomized control trials done by Kane and Pierce[26] found no benefit in treating male partners of women with BV.

The recommended treatment for BV is metronidazole 500 mg orally two times daily for 7 days, or metronidazole gel 0.75% intravaginally daily for 5 days, or clindamycin cream 2% intravaginally once per day for 7 days.[25] Other treatments, such as metronidazole 2 g orally in a single dose or ofloxacin 200 mg orally twice daily have been shown to have similar cure rates at 2 weeks after treatment, but have higher relapse rates after that 2-week timeframe.[1,26] It is important to remember that clindamycin cream can weaken latex condoms and diaphragms, and women should be advised to use other forms of contraception during treatment with clindamycin vaginal preparations.[25] The CDC-recommended regimens and alternate regimens can be found in Table 2 . Women who are HIV-positive are treated the same as women who are HIV-negative. When comparing the cost of either of the CDC-recommended treatments, it should be noted that the cost of oral metronidazole is substantially less that that of the other two regimens.[27]

In women who are allergic to metronidazole, treatment with clindamycin is recommended. For those who have gastrointestinal symptoms when taking oral metronidazole, vaginal metronidazole therapy can be used.[25]

BV has been associated with poor pregnancy outcomes. CDC recommendations include treating all symptomatic pregnant women. In 3 of 4 randomized studies, screening and treating asymptomatic women who are at high risk for preterm birth has reduced the preterm birth rate.[25] Those pregnant women who are asymptomatic and treated should be re-examined 1 month after treatment to determine if the therapy was curative.[25] The data do not support using topical treatments during pregnancy, as evidence from three studies indicated an increase in prematurity and neonatal infections in those treated with clindamycin cream.[25] The following are recommended: metronidazole 500 mg orally, twice per day for 7 days or metronidazole 250 mg orally, three times per day for 7 days, or clindamycin 300 mg orally two times per day for 7 days. Treatment can occur during any trimester of pregnancy.[25] Practitioners may be hesitant to treat with metronidazole during the first trimester based on their concern that it is teratogenic. A recent meta-analysis, including studies that had at least 10 women exposed to metronidazole during the first trimester, found no increase in birth abnormalities.[28] The CDC no longer has the stipulation that metronidazole treatment be deferred until after the first trimester.[25]

Recurrence of BV is very common and has been reported to be as high as 70% over a period of 9 months following initial diagnosis.[4] Women are often frustrated and embarrassed to come back in for follow-up, but because of the high rate of recurrence, they should be encouraged to notify the clinician at the onset of symptoms. The CDC guidelines acknowledge the problem of recurrent BV and currently do not recommend any long-term management.[25] The 2006 CDC guidelines suggest that women with multiple recurrences be managed in consultation with a specialist.[25] Others recommend confirmation of the recurrent BV diagnosis by Gram stain, which is the gold standard diagnostic technique. After the diagnosis is confirmed, one may prescribe a 10- to 14-day regimen of oral metronidazole 500 mg twice daily.[1,4,25]

Another regimen suggested is vaginal metronidazole 0.75% once daily for 10 days, followed by twice weekly application for 4 to 6 months.[16] A recent trial of treatment for persistent BV found that twice a week application of 0.75% metronidazole gel for 6 months maintained a clinical cure.[29] Even with extended therapy, relapse is common after stopping the therapy.[4] In addition, long-term use of metronidazole may lead to an increased rate of vulvovaginal candidiasis.[30] Some have suggested using nystatin and metronidazole as primary treatment for recurrent BV. In a study conducted in Peru by Sanchez et al.,[14] women with diagnosed BV were either treated with vaginal metronidazole gel (MetroGel vaginal; 3M Pharmaceuticals, Northridge, CA) 0.75% for 5 days or ovules containing metronidazole 500 mg and nystatin 100,000 U for 5 days. The recurrence rates were significantly lower in the combination medication group than the vaginal metronidazole gel group even up to 104 days after treatment.

Tinidazole has been suggested as a treatment for refractory BV[31] (an off-label use). It is a nitroimidazole derivative similar to metronidazole. It has a higher peak concentration and longer half-life than metronidazole, which would lead to more stable blood levels than occur with metronidazole. In a case report by Baylson et al.,[31] a woman with refractory BV was treated with tinidazole 500 mg twice daily for 2 weeks. She remained asymptomatic for 10 months after treatment. They also reported that a 2-gram single dose of tinidazole resolves nonrecurrent BV. Tinidazole has been shown to cause less gastrointestinal side effects than metronidazole.[31]

Recolonization of the vagina with lactobacilli has been suggested as a line of therapy for recurrent BV. Ongoing studies evaluating this therapy are in clinical trials. Because data are lacking and commercial lactobacillus preparations are not standardized, the use of these preparations is not recommended.[16,25]


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.