Background: Diagnosis of Trichomonas vaginalis (TV) infection is limited by imperfect testing methods. Newer tests, such as rapid antigen and nucleic acid amplification tests, are often compared with culture, which is not widely used but is more sensitive than wet mount. We assessed the sensitivity and specificity of 4 tests for the identification of TV using 3 statistical approaches.
Methods: Sexually active adolescent women aged 14-21 years (n=330) were recruited from a teen health center and emergency department. Vaginal swabs were tested for TV using wet mount, culture (InPouch TV; Biomed Diagnostics), rapid antigen testing (OSOM TV; Genzyme Diagnostics), and transcription-mediated amplification testing (TMA; APTIMA TV analyte specific reagents; Gen-Probe).
Results: TV was detected in 61 participants (18.5%). Compared with a composite reference standard (i.e., any TV test with positive results), the sensitivities of wet mount, culture, rapid antigen testing, and TMA were 50.8%, 75.4%, 82%, and 98.4%, respectively. Using latent class analysis, the sensitivity of wet mount (56%) was significantly lower than that of other tests, and the sensitivities of culture and rapid antigen testing were similar (83% and 90%, respectively); specificity was 100% for each of these 3 methods. TMA had a sensitivity of 98.2% and a specificity of 98%. Tests performed equally well regardless of whether the participant had bleeding or other infections. The sensitivities of the rapid antigen test and TMA were comparable (92.5% and 97.5%, respectively) in women who had vaginal symptoms.
Conclusions: Wet mount alone is insufficient for the reliable diagnosis of TV infection in women. TMA and rapid antigen tests are highly sensitive and specific, and both are superior to wet mount. Rapid antigen testing is equivalent to culture, and it compares favorably with the sensitivity of TMA for the detection of TV.