Sensitivity and specificity of the Papanicolaou-stained cervical smear in the diagnosis of Chlamydia trachomatis infection.

Two hundred young women had simultaneously prepared cultures for Chlamydia trachomatis and cervical smears; they also completed a questionnaire. Twelve of the chlamydial cultures were positive. There was poor correlation between the culture results and the cytologic morphology or symptoms. On initial blind reading, only 10% of the smears cytologically interpreted as positive were actually positive by culture. Under the most favorable (non-blind) interpretation, only 23% of the smears cytologically interpreted as positive for chlamydial infection were also culture positive. Because of the high incidence of false positives, we conclude that routine cytologic examination of Papanicolaou-stained smears is not an acceptable method for the diagnosis of chlamydial infections of the cervix. Immunoperoxidase staining of duplicate smears did not appear to be a successful replacement for culture.