Bacterial Vaginosis
Robert E. Gutman, MD, Jeffrey F. Peipert, MD, MPH, Sherry Weitzen, PhD, and Jeffrey Blume, PhD
OBJECTIVE: To determine whether the current clinical crite- ciated with increased risk of preterm delivery,3 prema-
ria for diagnosing bacterial vaginosis can be simplified by ture rupture of membranes,4 amniotic fluid infection,5
using 2 clinical criteria rather than the standard 3 of 4 chorioamnionitis,6 postpartum endometritis,79 pelvic
criteria (Amsels criteria).
inflammatory disease,10 12 and postoperative infec-
METHODS: This was a prospective observational study of tions.13,14
269 women undergoing a vaginal examination in the Currently, the clinical criteria for diagnosing bacterial
Womens Primary Care Center, Division of Research, or
vaginosis requires the presence of 3 of the following 4
Colposcopy Clinic at Women & Infants Hospital. All 4
criteria: 1) a thin, homogeneous discharge, 2) vaginal pH
clinical criteria for diagnosing bacterial vaginosis were
collected, and Gram stain was used as the gold standard. greater than 4.5, 3) a positive whiff test or release of
Sensitivity and specificity were calculated for each individ- amine odor with the addition of base, and 4) clue cells on
ual criterion, combinations of criteria, and a colorimetric microscopic evaluation of saline wet preparation. These
pH and amine card. Receiver operating characteristic criteria are based on the original work of Amsel et al.15
curve was generated to estimate the preferred pH and The Gram stain is believed by many to be the gold
percentage of clue cells for diagnosing bacterial vaginosis. standard for diagnosing bacterial vaginosis. However,
RESULTS: The prevalence of bacterial vaginosis in our interpreting the Gram stain requires experience, and it is
study population was 38.7%. Vaginal pH was the most often difficult to get timely results for the clinical diagno-
sensitive of all the criteria, at 89%, and a positive amine sis of bacterial vaginosis. Thus, it is important to have
odor was the individual criteria with the highest specificity, simple and reliable clinical criteria that clinicians can use
at 93%. Similar specificity was seen with combinations of 2 in practice. Because many clinicians do not routinely
criteria and Amsels criteria. Receiver operating character- evaluate patients for all 4 of Amsels criteria, it would be
istic curve analysis yielded a preferred pH and percentage
helpful to know if different combinations of criteria can
of clue cells of 5.0 and 20%, respectively. However, a pH of
be used to accurately diagnose bacterial vaginosis.
4.5 or greater improves sensitivity with minimal loss of
specificity. Although many research studies have looked at the
sensitivity and specificity of Amsels criteria, few have
CONCLUSION: The clinical criteria for diagnosing bacterial
looked at these criteria individually and in various com-
vaginosis can be simplified to 2 clinical criteria without loss
of sensitivity and specificity. (Obstet Gynecol 2005;105: binations using the Gram stain as the gold standard. We
551 6. 2005 by The American College of Obstetricians conducted a MEDLINE search from 1966 to November
and Gynecologists.) 2003, using the terms bacterial vaginosis, BV, and
LEVEL OF EVIDENCE: II-2 nonspecific vaginitis. We were unable to find any
studies that evaluate all of the different combinations of
Bacterial vaginosis is the most common vaginal infec- clinical criteria for the purpose of simplifying and possi-
tion, with a prevalence of 9 37%, depending on the bly improving the accuracy of current clinical methods
population studied.1,2 Bacterial vaginosis has been asso- for diagnosing bacterial vaginosis.
We hypothesize that the current clinical criteria for
From the Division of Research, Department of Obstetrics and Gynecology, Women diagnosing bacterial vaginosis can be simplified by using
& Infants Hospital, and the Center for Statistical Sciences, Brown University, 2 clinical criteria, rather than the standard 3 of 4 criteria
Medical School, Providence, Rhode Island. (Amsels criteria), without loss of sensitivity or specificity.
Preliminary data presented at the Infectious Disease Society of Obstetrics and Furthermore, we believe that the addition of the FemExam
Gynecology (IDSOG) Annual Meeting in Quebec City, Canada, August 9, 2001. card (CooperSurgical Inc, Trumbull, CT), a colorimetric
The authors thank CooperSurgical for the donation of the FemExam cards used in pH and amine card, will have better sensitivity and speci-
this study. ficity when compared with Amsels criteria.
the card, the sensitivity was 89% and specificity was 61%. A receiver operating characteristic curve was gener-
When both of the criteria were present, the sensitivity ated for pH and clue cells on wet preparation. We
was 40% and the specificity was 95%. visually estimated the point of inflection on the receiver
Table 2. Sensitivity, Specificity, and 95% Confidence Intervals of the Clinical Criteria for Diagnosing Bacterial Vaginosis
Sensitivity 95% CI Specificity 95% CI AUC 95% CI
Thin homogeneous discharge 0.79 0.690.87 0.54 0.460.62 0.77 0.690.84
pH 4.5 0.89 0.820.95 0.74 0.660.80 0.82 0.760.86
Positive amine odor 0.67 0.570.76 0.93 0.880.97 0.80 0.750.85
Clue cells present ( 20% per hpf) 0.74 0.650.82 0.86 0.800.91 0.80 0.750.85
pH 4.5 and thin homogeneous discharge 0.69 0.590.79 0.86 0.800.91 0.78 0.720.83
pH 4.5 and amine odor 0.64 0.540.73 0.95 0.910.98 0.80 0.740.84
pH 4.5 and clue cells 0.69 0.590.78 0.92 0.860.95 0.80 0.750.85
Clue cells and amine odor 0.63 0.530.73 0.95 0.900.97 0.79 0.740.84
Clue cells and thin homogeneous discharge 0.61 0.500.71 0.91 0.860.95 0.76 0.700.81
Amine odor and thin homogeneous 0.58 0.470.68 0.94 0.890.97 0.76 0.700.81
discharge
Amsels criteria ( 3 of 4 criteria) 0.69 0.590.78 0.93 0.870.96 0.81 0.760.86
CI, confidence interval; AUC, area under the curve; hpf, high-power field.
Table 3. Sensitivity, Specificity, and 95% Confidence Intervals of the FemExam Card for Diagnosing Bacterial Vaginosis
Sensitivity 95% CI Specificity 95% CI AUC 95% CI
pH positive 0.88 0.800.93 0.64 0.560.72 0.76 0.700.81
Amine positive 0.41 0.320.51 0.91 0.850.95 0.66 0.600.72
pH or amine positive 0.89 0.810.94 0.61 0.530.68 0.75 0.690.79
pH and amine positive 0.40 0.310.50 0.95 0.900.97 0.67 0.610.73
CI, confidence interval; AUC, area under the curve.
VOL. 105, NO. 3, MARCH 2005 Gutman et al Clinical Diagnosis of Bacterial Vaginosis 553
black and Hispanic women, who appear to be at higher
risk of having bacterial vaginosis than white women.17
We confirmed the findings seen in another large cross-
sectional study, which showed that patients with bacte-
rial vaginosis are more likely to have vaginal symptoms,
specifically vaginal discharge and foul odor.17,18 A study
by Larsson et al19 revealed similar risk factors for sexu-
ally transmitted diseases and bacterial vaginosis, includ-
ing lower age of first intercourse and higher number of
lifetime sexual partners. Yen et al17 displayed a lower
rate of bacterial vaginosis among the nonsexually ex-
perienced and a higher rate among those with multiple
sexual partners in the past 3 months. However, in our
study and in Yens, a history of a sexually transmitted
disease was not associated with an increased prevalence
Fig. 1. Receiver operating characteristic curve for pH. of bacterial vaginosis.
Gutman. Clinical Diagnosis of Bacterial Vaginosis. Obstet Gynecol 2005. We confirmed the findings of Eschenbach et al18 that
vaginal pH had the highest sensitivity, but the lowest
operating characteristic curve. A pH of 5.0 or greater specificity, of all clinical diagnostic criteria. The low
maximized sensitivity (83%) and specificity (82%). How- specificity is not surprising considering that elevated pH
ever, a pH of 4.5 or more improves sensitivity to over may be altered by semen, cervical mucus, and blood. An
89% while still maintaining a relatively high specificity of elevated pH is more common in postmenopausal wom-
74% (Fig. 1). Sensitivity and specificity of clue cells appears en20 and can also be seen with trichomoniasis infections.
to be maximized with a cutoff of more than 20% clue cells The high sensitivity should serve as an asset in screening
on saline wet preparation (Fig. 2; sensitivity 74%, specific- algorithms. Our findings for amine odor also support
ity, 86%). previous studies that found it to have the highest speci-
ficity and lowest sensitivity of the individual criteria.18,21
The combination of any 2 criteria decreased sensitivity
DISCUSSION and increased specificity. However, the sensitivity and
The prevalence of bacterial vaginosis was higher in our specificity of each pair of criteria is similar to those
study (39%) than in most studies,1,2,17 and thus repre- observed for Amsels criteria. This implies that there is
sents a high-risk population. Although the ethnic diver- no clear advantage to using Amsels criteria ( 3 of the 4
sity of this population is not representative of most U.S. criteria). This supports our hypothesis that the current
communities, it provides us with a better sampling of clinical criteria for diagnosing bacterial vaginosis (Am-
sels criteria) can be simplified by using 2 clinical criteria,
without significant loss of sensitivity or specificity.
One study evaluating an objective colorimetric test for
amines found a sensitivity of 87% and specificity of 98%
compared with Gram-stain criteria.22 The authors con-
cluded that these tests are more sensitive than the whiff
test for detecting the presence of amines. We found the
FemExam card to have less sensitivity than the whiff
test for detecting amines and lower specificity for detect-
ing an elevated vaginal pH. The difference in pH may be
due to the different cutoff points: 4.5 for our study and
4.7 for the FemExam card. Overall, we felt the majority
of the differences could be explained by the difficulty in
reading and interpreting the results of the card. Although
it is conceptually easy to distinguish between a plus sign
and a minus sign, the results often had very faint blue
Fig. 2. Receiver operating characteristic curve for clue signals that were difficult to interpret. The findings may
cells. also be biased because the card was read after the clinical
Gutman. Clinical Diagnosis of Bacterial Vaginosis. Obstet Gynecol 2005. criteria were collected and recorded, although we would
VOL. 105, NO. 3, MARCH 2005 Gutman et al Clinical Diagnosis of Bacterial Vaginosis 555
10. Sweet RL. Role of bacterial vaginosis in pelvic inflamma- Boyer CB. Bacterial vaginosis in sexually experienced and
tory disease. Clin Infect Dis 1995;20(suppl):S2715. non-sexually experienced young women entering the mil-
11. Peipert JF, Montagno AB, Cooper AS, Sung CJ. Bacterial itary. Obstet Gynecol 2003;102:92733.
vaginosis as a risk factor for upper genital tract infection. 18. Eschenbach DA, Hillier S, Critchlow C, Stevens C, DeR-
Am J Obstet Gynecol 1997;177:1184 7. ouen T, Holmes KK. Diagnosis and clinical manifestations
12. Wiesenfeld HC, Hillier SL, Krohn MA, Amortegui AJ, of bacterial vaginosis. Am J Obstet Gynecol 1988;158:
Heine RP, Landers DV, et al. Lower genital tract infection 819 28.
and endometritis: insight into subclinical pelvic inflamma- 19. Larsson PG, Platz-Christensen JJ, Sundstrom E. Is bacte-
tory disease. Obstet Gynecol 2002;100:456 63. rial vaginosis a sexually transmitted disease? Int J STD
13. Lin L, Song J, Dimber N, Shott S, Tangora J, Aroutchea A, AIDS 1991;2:362 4.
et al. The role of bacterial vaginosis in infection after major 20. Brizzolara S, Killeen J, Severino R. Vaginal pH and para-
gynecologic surgery. Infect Dis Obstet Gynecol 1999;7: basal cells in postmenopausal women. Obstet Gynecol
169 74. 1999;94:700 3.
14. Persson E, Bergstrom M, Larsson PG, Moberg P, Platz- 21. Bump RC, Zuspan FP, Buesching WJ 3rd, Ayers LW,
Christensen JJ, Schedvins K, et al. Infections after hyster- Stephens TJ. The prevalence, six-month persistence, and
ectomy: a prospective nation-wide Swedish study. The predictive values of laboratory indicators of bacterial vagi-
Study Group on Infectious Diseases in Obstetrics and nosis (nonspecific vaginitis) in asymptomatic women. Am J
Gynecology within the Swedish Society of Obstetrics and Obstet Gynecol 1984;150:91724.
Gynecology. Acta Obstet Gynecol Scand 1996;75: 22. Sonnex C. The amine test: a simple, rapid, inexpensive
757 61. method for diagnosing bacterial vaginosis. Br J Obstet
15. Amsel R, Totten PA, Spiegel CA, Chen KC, Eschenbach Gynaecol 1995;102:160 1.
D, Holmes KK. Nonspecific vaginitis: diagnostic criteria
and microbial and epidemiologic associations. Am J Med
1983;74:14 22. Reprints are not available. Address correspondence to: Robert
E. Gutman, MD, 4940 Eastern Avenue, Room 125, Baltimore,
16. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnos-
MD 21224-2780; e-mail: rgutman1@jhmi.edu.
ing bacterial vaginosis is improving by a standardized
method of Gram stain interpretation. J Clin Microbiol
1991;29:297301. Received June 3, 2004. Received in revised form September 7, 2004.
17. Yen S, Shafer MA, Moncada J, Campbell CJ, Flinn SD, Accepted October 7, 2004.