To the Editor.
—Dr Mansergh and colleagues1 have made an important contribution to the study of methods of reducing vertical transmission of human immunodeficiency virus (HIV) in developing countries, and we endorse their conclusion that zidovudine treatment will be an inefficient use of limited resources in the poorest settings. However, we would like to highlight the sensitivity of their findings to levels of breast-feeding.The authors describe an average sub-Saharan African setting where most mothers choose the healthier option of breast-feeding.1 The description of the model overlooks the fact that any intervention that reduces the risk of prenatal or perinatal HIV-1 transmission alone is likely to increase the number and proportion of infections that take place through breast-feeding. Even if there is an overall gain in infections prevented, a number of mother-to-child infections/transmissions prevented prenatally will merely be delayed into the postnatal period.The size of this effect
Nicoll A, Newell ML. Preventing Perinatal Transmission of HIV: The Effect of Breast-feeding. JAMA. 1996;276(19):1552. doi:10.1001/jama.1996.03540190024018