Background: Highly active antiretroviral therapy (HAART) drastically reduces mother-to-child transmission of HIV, but where breastfeeding is the only safe infant feeding option, HAART for the prevention of mother-to-child transmission needs to be evaluated in relation to both HIV transmission and infant mortality.
Design and methods: One hundred and two > or = 18-year old women on HAART in rural Uganda who delivered one or more live infants between March 1, 2003 and January 1, 2007 were enrolled in a prospective study to assess HIV transmission and infant survival. All pregnant women were counseled to exclusively breastfeed for 3-6 months according to national guidelines at the time. Infants were followed-up for > or = 7 months and were offered HIV polymerase chain reaction testing quarterly from 6 weeks of age until > or = 6 weeks after complete weaning.
Results: Of 118 infants born during follow-up, 109 (92%) were breastfed. Median durations of exclusive and total breastfeeding were 4 months (interquartile range 3-6) and 5 months (interquartile range 3-7), respectively. None of the infants tested HIV polymerase chain reaction positive over follow-up but 16 infants died without a definitive HIV status at a median age of 2.6 months. In total, 23 (19%) infants died during follow-up at a median age of 3.7 months; 15 (65%) of whom with severe diarrhea and/or vomiting in the week preceding their death. In multivariate analysis, there was a 6-fold greater risk of death among infants breastfed for less than 6 months independent of maternal CD4 count closest to delivery, maternal marital status or maternal death (adjusted hazard ratio = 6.19; 95% confidence interval 1.41-27.0, P = 0.015).
Conclusions: In resource-constrained settings, HIV-infected pregnant women should be assessed for HAART eligibility and treated as needed without delay, and should be encouraged to breastfeed their infants for at least 6 months.