Methods
The Preventing Vaginal Infections trial was a randomized trial of periodic presumptive treatment to reduce vaginal infections over 12 months in US and Kenyan women. Eligible women were aged 18-45, HIV-negative, and not pregnant or breastfeeding. Vaginal fluid specimens were collected using Dacron swabs at enrollment and every 2 months. Bacterial concentrations were measured using quantitative PCR targeting the 16S rRNA gene for key bacterial species that are markers of BV or health including BVAB1, BVAB2, M. indolicus, A. vaginae, Megasphaera species, Leptotrichia/Sneathia, G. vaginalis, L. crispatus, L. jensenii, and L. iners. Self-reported contraceptive use (injectable, oral, implant, copper IUD, no highly effective method) were assessed as independent exposures using linear mixed models. Analyses were repeated among the subset of women who used the same method throughout follow-up.
Methods
The Preventing Vaginal Infections trial was a randomized trial of periodic presumptive treatment to reduce vaginal infections over 12 months in US and Kenyan women. Eligible women were aged 18-45, HIV-negative, and not pregnant or breastfeeding. Vaginal fluid specimens were collected using Dacron swabs at enrollment and every 2 months. Bacterial concentrations were measured using quantitative PCR targeting the 16S rRNA gene for key bacterial species that are markers of BV or health including BVAB1, BVAB2, M. indolicus, A. vaginae, Megasphaera species, Leptotrichia/Sneathia, G. vaginalis, L. crispatus, L. jensenii, and L. iners. Self-reported contraceptive use (injectable, oral, implant, copper IUD, no highly effective method) were assessed as independent exposures using linear mixed models. Analyses were repeated among the subset of women who used the same method throughout follow-up.
Results
Among 116 women enrolled in the placebo arm, 107 women were eligible for this analysis. The proportion of visits with BV differed by contraceptive method (BV+ at 24% of injectable visits versus 39% of non-method visits; p=0.01). After adjusting for age, site, sex without a condom, and new sex partners, IUD users had an average -1.2 (95% CI -1.9, -0.4) log load of L. jensenii compared to women using no highly effective method; likewise, injectable users had an average -0.7 (95% CI -1.3, 0) log load of L. iners compared to women using no method. L. crispatus, Leptotrichia/Sneathia and G. vaginalis quantities were similar across all groups. The magnitude of the difference in bacterial loads was larger when restricted to consistent contraceptive users. However, there were no significant associations in this smaller population.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
