Accuracy of the Papanicolaou smear in the diagnosis of asymptomatic infection with Trichomonas vaginalis.

OBJECTIVE

To assess the accuracy of Papanicolaou smears in reporting cytologic evidence of Trichomonas vaginalis in asymptomatic women attending a resident clinic.

METHODS

In phase I of this study, we prospectively enrolled 100 asymptomatic gynecologic patients to be screened for vaginal trichomoniasis using wet preparation, vaginal culture, and Papanicolaou smear. During phase II, asymptomatic patients (40 gynecologic and 20 obstetric) whose screening Papanicolaou smears showed cytologic evidence of trichomoniasis returned for wet preparation, culture, and repeat Papanicolaou smear. Patients were considered infected with T vaginalis if either the wet preparation or culture was positive, and uninfected if both tests were negative. The cytopathologist was not informed of the patient's enrollment in this study or of the results of culture or wet preparation.

RESULTS

The prevalence of asymptomatic trichomonas infection in gynecologic patients enrolled in phase I was 6%. In asymptomatic gynecologic patients enrolled in phase II, repeat Papanicolaou smear had a sensitivity and specificity of 86 and 83%, respectively, when diagnosing infection. Thirty percent of these patients would have been treated unnecessarily for trichomoniasis based upon screening Papanicolaou smear. In obstetric patients, the sensitivity of repeat Papanicolaou smear was 94% and specificity was 100%. Had therapy been initiated based on screening cytology, 20% of obstetric patients would have received unindicated therapy. The differences in sensitivity and specificity between the groups were statistically significant (P < .05).

CONCLUSION

When a screening Papanicolaou smear reports cytologic evidence of T vaginalis infection in the asymptomatic patient, a confirmatory test should be performed before initiating therapy.