Objectives: To investigate the use of lactobacillus as a simple tool for the diagnosis of bacterial vaginosis (BV) using wet mounts in pregnant women and to assess the positive predictive value of absence of lactobacillus for Human Immunodeficiency Virus Type-1 (HIV-1) infection.
Methods: 409 pregnant women were enrolled from three randomly selected clinics around Harare. The women underwent clinical examination during which a speculum-aided high vaginal swab was obtained and tested for BV using Amsel criteria, Nugent criteria and the simple lactobacillus method.
Results: The prevalence of BV was 29% by Amsel criteria, 34% by Nugent criteria and 49% by the simple lactobacillus method. The sensitivity and specificity of the simple method using Amsel as the gold standard is 83% and 65% respectively with a kappa value of 0.40. The sensitivity and specificity of the simple test using Nugent as the gold standard is 86% and 82% respectively with a kappa value of 0.68. Sensitivities and specificities of individual Amsel criteria including lactobacillus for determining bacterial vaginosis with Amsel as the gold standard were as follows; discharge 15% and 99%, whiff 96% and 85%, clue cells 96% and 75%, pH 99% and 22% and lactobacillus 83% and 65%. Using Nugent as the gold standard the respective sensitivities and specificities were as follows; discharge 08% and 99%, whiff 70% and 87%, clue cells 67% and 73%, pH 92% 22% and lactobacilli 86% and 82%. The HIV-1 prevalence in the BV study sample (n=392) was about 46%. According to Amsel criteria only 26% (OR= 0.78) of the HIV positive participants have BV while according to the lactobacilli method 59% (OR=2.14) of the HIV positive participants do not have lactobacillus as part of the normal flora of the lower female genital tract. Amsel BV positive predictive value for HIV is 42% while lactobacillus positive predictive value for HIV is 56%.
Conclusion: The sensitivity of the lactobacillus method is as good as Nugent criteria using Amsel as the gold standard. It is much simpler to perform, less expensive, easy to train and takes much shorter time to perform and therefore has a potential for a much wider use than both Amsel and Nugent criteria. The simple lactobacillus method has a better PPV for HIV-1 compared to both Amsel and Nugent. Amsel criteria can be improved by removal of subjective criteria.