N of cRAI in the past three months being reported more frequently in the placebo arm (P = 0.01). Of the 1,383 participants who tested seronegative for HSV-2 at baseline, 36 (2.6 ) did not contribute person-time to incidence analyses because they were retrospectively found to be HIVinfected at baseline, tested seropositive for HSV-2 at the enrollment visit subsequent to screening, or were lost to f