RESUMO
OBJECTIVE: To assess the variability in Nugent score and leukocyte count measured in vaginal secretions collected from 3 vaginal sites. METHODS: Fifty pregnant women at less than 20 weeks of gestation were consecutively recruited at the time of their first prenatal visit. Three vaginal smears were collected from each woman, 1 from the posterior fornix, 1 from the mid-lateral wall, and 1 from the introitus. Smears were Gram stained and evaluated for bacterial vaginosis using Nugent's criteria. Each smear was classified as positive for bacterial vaginosis if the Nugent score was 7 or greater, intermediate if the score was between 4 and 6, and negative if the Nugent score was 3 or less. A mean leukocyte value was obtained for each slide by evaluating 5 fields under oil immersion. Repeated-measures analysis of variance was used to compare mean Nugent scores and leukocyte counts across sites. RESULTS: The sample consisted of mostly African-American, young, low-income women in their first trimester of pregnancy. Forty-seven percent were bacterial vaginosis-positive (Nugent score 7 or greater). Mean Nugent scores (+/- standard deviation) across the vaginal sample collection sites were similar: posterior fornix = 4.2 (4.4); mid-lateral wall = 4.2 (4.5); introitus = 4.2 (4.6). In contrast, the mean leukocyte count varied significantly across vaginal sample collection sites: 2.5 (4.2); mid-lateral wall = 2.8 (4.4); introitus = 6.5 (8.6) (F = 11.26 (1, 47.5); P = .002). CONCLUSION: Leukocyte counts vary according to the site from which vaginal secretions were obtained, with the highest leukocyte count in samples obtained from the introitus. Bacterial vaginosis diagnosis by Nugent score does not vary by site of sample collection.