Associations between intravaginal practices and HIV acquisition, adjusting for different measures of disrupted vaginal flora.

HRs from two-stage random effects meta-analysis. Intravaginal practices measured at baseline; reference category is no vaginal practice or use of water only.aNumbers of observations differ from Figures 1–4 because they exclude those with no BV measurement, mostly from study 7 in which BV was measured only in a subset at baseline in the Zimbabwe site; model for insertion of products also excludes study 2, which did not ask about this practice.bAdjusted for age, marital status, and reported number of sex partners in last 3 mo, as recorded at cohort entry.cBV as binary variable defined as Nugent score of 7–10, Ison-Hay grade III, or the presence of three or more Amsel criteria.