ABSTRACT
BACKGROUND: Infectious vaginitis is a common gynecologic disease that is primarily caused by three pathogens (Trichomonas vaginalis, Gardnerella vaginalis, and Candida species). The aim of this study was to confirm the effects of other infectious vaginitis-related test results on the interpretation of Gram stain and Papanicolaou (Pap) smear test results for disease diagnosis. METHODS: A total of 300 vaginal samples were collected from women presenting symptoms of vaginitis. The presence of the three previously mentioned pathogens was evaluated using both a Gram stain and Pap smear test, and interpreted twice by 4 different observers. The first interpretation was performed without any information, and a second interpretation was performed with knowledge of results of an Affirm VPIII test that was used to diagnose infectious vaginitis. The results from the two interpretations were compared and the sensitivity and specificity of both tests were evaluated. RESULTS: For the Gram stain samples, the detection rates of G. vaginalis were increased in the second interpretation by 6.2%, while the detection rates of Candida spp. were decreased by 0.3%. For the Pap smear test samples, the detection rates of G. vaginalis were increased in the second interpretation by 7.0%, and the detection rates of Candida spp. were increased by 2.0%. The sensitivity of both tests was increased in the second interpretation by 5.5% to 66.7%. There was no difference in the specificity between the two interpretations. CONCLUSIONS: We demonstrated that there is significant inter-observer variation when using Gram stain and Pap smear test results to diagnose infectious vaginitis. The detection rates and sensitivity of both tests changed when the results from an additional test were incorporated into the interpretation. Additional studies are needed to develop objective criteria and a standardized interpretation system for the evaluation of results from these diagnostic tests.