ABSTRACT
The aim of this study was to determine the accuracy of a wet-mount bacterial morphology scoring (BMS) system and Nugent's Gram stain analysis for the diagnosis of bacterial vaginosis, using Amsel's criteria as the gold standard. The three diagnostic criteria were assessed independently The BMS diagnosis was based on a scoring system which weighed the number of small bacterial morphotypes regarded as typical of bacterial vaginosis against lactobacillary morphotypes in phase-contrast microscopy of wet-mount preparations. Three groups of non-pregnant women attending either because of vaginal discharge, other genitourinary symptoms, or for a routine check-up, and a group of pregnant women attending for antenatal care were studied. The diagnostic accuracy was measured by sensitivity, specificity, positive and negative predictive values, and likelihood ratio. The accuracy of the BMS diagnosis was substantially high in all of the examined groups (LR 15.4-20.3). The accuracy of the Gram stain diagnosis was lower (LR 7.6-10.9). In the total material, the accuracy of the BMS diagnosis was higher than that of the Nugent's Gram staining. Intra- and inter-observer reproducibility of all three criteria applied was high. We propose greater routine use of the new BMS diagnosis for point-of-care testing in family practice as well as in research and in microbiology laboratories.