Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Vaginal signs and symptoms perform poorly in diagnosing vaginal candidiasis, bacterial vaginosis, and vaginal trichomoniasis


ACP J Club. 2004 Sep-Oct;141:47. doi:10.7326/ACPJC-2004-141-2-047

Related Content in the Archives
• Letter: Review: Vaginal signs and symptoms perform poorly in diagnosing vaginal candidiasis

Clinical Impact Ratings

Emergency Med: 5 stars

GIM/FP/GP: 6 stars

Infectious Disease: 5 stars

Source Citation

Anderson MR, Klink K, Cohrssen A. Evaluation of vaginal complaints. JAMA. 2004;291:1368-79. [PubMed ID: 15026404]



In patients with vaginitis, how do individual symptoms, physical examination signs, and laboratory tests perform in diagnosing vaginal candidiasis (VC), bacterial vaginosis (BV), and vaginal trichomoniasis (VT)?


Data sources: Studies were identified by searching MEDLINE (1966 to April 2003), hand-searching the most recent American College of Obstetricians and Gynecologists Technical Bulletin, and scanning bibliographies of relevant studies.

Study selection and assessment: Studies were selected if they included symptomatic patients in primary care or sexually transmitted disease clinics, compared a diagnostic test with a recognized gold standard, calculated sensitivity and specificity, and discussed tests that would provide diagnostic information during the office visit. Gold standard tests for VC, BV, and VT included a positive culture or identification of yeast by microscopy, the Amsel criteria (3 of a thin, homogeneous vaginal discharge, clue cells, positive whiff test, and vaginal pH level > 4.5), and a positive culture, respectively. Studies were assessed for methodological quality using a 3-level scale (level 1 = highest quality).

Outcomes: Sensitivity, specificity, and likelihood ratios.

Main results

18 articles met the selection criteria. The quality of the studies ranged from level 2 to 3. The test characteristics of individual symptoms and physical examination signs are in the Table. Laboratory tests performed better than signs and symptoms in diagnosing VC, BV, and VT (Table).


In patients with vaginitis, individual symptoms and physical examination signs do not perform well in diagnosing vaginal candidiasis, bacterial vaginosis, and vaginal trichomoniasis. Laboratory tests perform better for diagnosing these conditions.

Source of funding: No external funding.

For correspondence: Dr. M.R. Anderson, Albert Einstein College of Medicine, Bronx, NY, USA. E-mail

Table. Test characteristics of symptoms, signs, and laboratory tests for diagnosing vaginal candidiasis (VC), bacterial vaginosis (BV), and vaginal trichomoniasis (VT)*

Tests Diagnosis Sensitivity Specificity +LR −LR
Individual symptoms
Cheesy discharge VC 65% 73% 2.4 0.48
Odor BV 97% 40% 1.6 0.07
Itching VC 27% 92% 3.3 0.79
Redness VC 28% 86% 2.0 0.84
External dysuria VC 33% 85% 2.2 0.79
Previous yeast infection VC 35% 90% 3.3 0.72
Physical examination signs
White, curdy discharge VC 16% 97% 6.1 0.86
Yellow discharge BV 60% 85% 4.1 0.46
Yellow discharge VT 89% 93% 14 0.12
Moderate discharge BV 62% 75% 2.5 0.5
Profuse discharge BV 4% 99% 3.0 0.98
High cheese odor BV 78% 75% 3.2 0.30
Redness or edema VT 18% 97% 6.4 0.85
Any inflammation VC 46% 78% 2.1 0.69
Normal pH level (< 4.9) VC 71% 90% 7.2 0.32
Gram stain VC 65% 100% 31 0.36
Bacilli with corkscrew motility BV 65% 100% 44 0.36
Positive whiff test VT 67% 65% 1.9 0.51
Wet mount VT 67% 100% 100 0.34

*Diagnostic terms defined in Glossary.


Common, non–life-threatening problems are often the least-studied areas of medicine, despite the overall burden of disease that they cause. At last, we are beginning to accumulate evidence to guide diagnosis and management in these areas. The review by Anderson and colleagues presents a summary of the available evidence on the accuracy of the symptoms, signs, and office laboratory tests for diagnosing vaginal complaints.

With the exception of white curdy discharge, which helps to rule in the diagnosis of VC; yellow discharge, which can indicate the presence of either BV or VT; and redness or edema, which can also indicate VT, these symptoms and signs are only moderately helpful in determining if a patient has any of these vaginal conditions. The absence of symptoms and signs is even less helpful in ruling out disease, with only the absence of odor ruling out BV, and absence of yellow discharge on physical examinatino ruling out VT. Office laboratory tests perform better because they rule in the disease if they are positive, but they do not rule out the disease if they are negative. Furthermore, many clinicians do not have the skills, time, or equipment to use these tests.

Several questions remain after reading this review: Could there be combinations of symptoms and signs that might better rule in or rule out the diagnosis? How does this information combine with more definitive testing, such as culture of a high vaginal swab? What is the treatment threshold for these conditions? Might it be better to treat on clinical suspicion, and test only those patients in whom a course of treatment fails? Primary care gynecology requires more research to guide clinical decision making in this area.

Jenny Doust, BMBS
University of Queensland
Brisbane, Queensland, Australia