Evaluation of vaginal complaints.

CONTEXT

Vaginal symptoms are one of the most common reasons for gynecological consultation. Clinicians have traditionally diagnosed vaginal candidiasis, bacterial vaginosis, and vaginal trichomoniasis using some combination of physical examination, pH, the wet mount, and the whiff test.

OBJECTIVES

To evaluate the role of the clinical examination and determine the positive and negative likelihood ratios (LRs) for the diagnosis of vaginal candidiasis, bacterial vaginosis, and vaginal trichomoniasis.

DATA SOURCES

Using a structured literature review, we abstracted information on sensitivity and specificity for symptoms, signs, and office laboratory procedures. We chose published (1966 to April 2003) articles that appeared in the MEDLINE database and were indexed under the combined search terms of diagnosis with vaginitis, vaginal discharge, candidiasis, bacterial vaginosis, and trichomoniasis.

STUDY SELECTION

Included studies of symptomatic premenopausal women seen in primary care settings. Tests were evaluated only if they would provide diagnostic information during the office visit and were compared with an acceptable criterion standard.

DATA EXTRACTION

All 3 authors extracted the data and computed sensitivity and specificity from each article independently. The absence of standard definitions for symptoms and signs made it impossible to combine results across studies.

DATA SYNTHESIS

Symptoms alone do not allow clinicians to distinguish confidently between the causes of vaginitis. However, a patient's lack of itching makes candidiasis less likely (range of LRs, 0.18 [95% confidence interval [CI], 0.05-0.70] to 0.79 [95% CI, 0.72-0.87]) and lack of perceived odor makes bacterial vaginosis unlikely (LR, 0.07 [95% CI, 0.01-0.51]). Similarly, physical examination signs are limited in their diagnostic power. The presence of inflammatory signs is associated with candidiasis (range of LRs, 2.1 [95% CI, 1.5-2.8] to 8.4 [95% CI, 2.3-31]). Presence of a "high cheese" odor on examination is predictive of bacterial vaginosis (LR, 3.2 [95% CI, 2.1-4.7]) while lack of odor is associated with candidiasis (LR, 2.9 [95% CI, 2.4-5.0]). Office laboratory tests, particularly microscopy of vaginal discharge, are the most useful way of diagnosing these 3 conditions.

CONCLUSIONS

The cause of vaginal complaints may be easily diagnosed when typical findings appear in microscopy. However, the poor performance of individual symptoms, signs, and office laboratory tests often makes it problematic to identify the cause of vaginal symptoms.