Résumé
Background. The identification of risk factors for in utero and intrapartum transmission of human immunodeficiency virus type 1 (HIV-1) is crucial to the design and understanding of preventive interventions.
Methods. The randomized Perinatal HIV Prevention Trial-1 enrolled 1437 pregnant women and their non breast-fed infants, to compare the efficacy of various durations of zidovudine prophylaxis. Using univariate and multivariate logistic regression analyses, we studied the role that factors known or occurring at various times during gestation or delivery play in in utero and intrapartum transmission.
Results. Variables independently associated with in utero transmission were HIV-1 load >35,000 copies/mL (adjusted odds ratio [AOR], 4.2) and delayed initiation of maternal zidovudine prophylaxis until >31.4 weeks gestation (AOR, 3.0). Variables associated with intrapartum transmission were HIV-1 load >10,000 copies/mL (AOR, 3.8 for 10,000-35,000 copies/mL and 7.1 for >35,000 copies/mL), induction of labor (AOR, 2.6), and premature labor with tocolysis (AOR, 15.1).
Conclusions. With the exception of very high HIV-1 load, risk factors for in utero transmission were different from those for intrapartum transmission. Optimal prophylactic interventions must address each of the major risk factors, with appropriate timing.