Finding of Vaginitis on Liquid-Based Papanicolaou Test Compared to Wet Mount and DNA Probe

Abstract

Previous studies have shown that the Pap test is not a reliable screening tool for vaginal infections. There are few data on the diagnostic accuracy of vaginal infections found incidentally on Pap specimens. This prospective study compared the diagnostic accuracy of incidental findings of Trichomonas vaginalis, bacterial vaginosis, and/or yeast/Candida infections on the liquid-based Pap test with the DNA-hybridization assay and the wet mount/KOH prep/whiff test. The participants were 609 women who underwent a Pap test as part of their regular obstetrical examination and had the other two tests during the same visit. Vaginal infections were present in 145 (23.8%) of the 609 women who underwent Pap tests, in 161 (26.4%) of those receiving the wet mount/KOH prep/whiff test, and in 365 (59.9%) analyzed with the DNA probe. Compared to the DNA probe, the specificity and positive predictive value of the Pap test were 99.6% and 93.1% for Trichomonas; 99.3% and 97.1% for bacterial vaginosis; and 96.3% and 61.5% for yeast/Candida. Compared to the wet mount/KOH prep/whiff test, the specificity and positive predictive value of Pap test were: 98.1% and 45.0% for Trichomonas; 91.5% and 41.4% for bacterial vaginosis; and 92.6% and 17.3% for yeast/Candida. The wet mount/KOH prep/whiff test demonstrated low sensitivities for the diagnosis of Trichomonas, bacterial vaginosis, and yeast/Candida: 31.0%, 23.4%, and 33.3%, respectively. The investigators conclude from these findings that incidental diagnosis of Trichomonas or bacterial vaginosis on the liquid-based Pap test is accurate, with a false-positive rate of <1%, and requires no additional testing.