Clinical utility of measuring white blood cells on vaginal wet mount and endocervical gram stain for the prediction of chlamydial and gonococcal infections.

BACKGROUND: White blood cells on endocervical Gram stain and vaginal wet mount are frequently used to predict chlamydial and gonococcal infections. Previous studies provide conflicting evidence for the clinical utility of these tests. GOAL: To evaluate the clinical utility of measuring white blood cells on vaginal wet mount and endocervical Gram stain for the prediction of chlamydial infection and gonorrhea. STUDY DESIGN: Women undergoing pelvic examinations at 10 county health department family planning and sexually transmitted disease clinics were tested for chlamydial infection by ligase chain reaction assay (n = 4550) and for gonorrhea by culture (n = 4402). Vaginal wet mount and endocervical Gram stains were performed in county laboratories at the time of examination. RESULTS: The prevalences of chlamydial infection and gonorrhea were 8.8% and 3.2%, respectively. For detection of chlamydial or gonococcal infection, the likelihood ratio was 2.85 (95% CI, 2.10-3.87) for > 30 white blood cells on vaginal wet mount and 2.91 (95% CI, 2.07-4.09) for > 30 white blood cells on endocervical Gram stain. Similar results were seen for individual diagnoses either of chlamydial infection or of gonorrhea. CONCLUSION: Vaginal wet mount and endocervical Gram stain white blood cells are useful for the presumptive diagnosis of chlamydial infection or gonorrhea only in settings with a relatively high prevalence of infection or when other predictors can increase the likelihood of infection.

Selective screening for chlamydial infection: which criteria to use?

BACKGROUND: Screening sexually active women for Chlamydia trachomatis is necessary to detect asymptomatic infections. Selective screening is a common strategy because universal screening is too costly in many settings. In order to guide local programs in the choice of selective screening criteria, we examined the performance of previously proposed screening criteria for C. trachomatis. METHODS: A clinic-based, cross-sectional study was conducted in public family planning and sexually transmitted disease (STD) clinics in ten counties in North Carolina. Women (n = 4471 in family planning and n = 2201 in STD clinics) undergoing pelvic examination were enrolled consecutively. Nine sets of screening criteria, including age alone, were compared using sensitivity, specificity, number of tests required and receiver-operator characteristic (ROC) analysis. All women underwent testing with ligase chain reaction assay of cervical specimens to identify C trachomatis infection. RESULTS: The prevalence of C. trachomatis was 7.8% and 11.0% in family planning and STD clinics, respectively. The sensitivities of published criteria ranged from 0.50 to 0.97. Specificities ranged from 0.05 to 0.66. In family planning clinics, the best performing criteria would detect 84% of infections while screening 51% of women. In STD clinics, the same criteria would detect 83% of infections but require testing 67% of women. Testing women aged < or =22 would detect 77% of infections in family planning and 74% of infections in STD clinics, while testing 51% and 48% of the women, respectively. CONCLUSIONS: When site-specific criteria cannot be developed, age alone is an acceptable strategy for selective screening for chlamydial infection.