The concept of training traditional birth attendants to improve maternal and perinatal health in developing countries began over 100 years ago and was promoted by the World Health Organization, United Nations, and donor agencies during the 1970s-1990s as a strategy to reduce maternal and neonatal mortality. Since then, there have been repeated assessments to ascertain whether the strategy works. The latest Cochrane review based on four studies, including one from Malawi, concluded: “The potential of TBA [traditional birth attendant] training to decrease newborn death is promising, when combined with improved health services. The number of studies however, is insufficient to provide the necessary evidence for TBA training effectiveness.”1

Closer examination of the review shows that this conclusion is not fully supported by the results of the studies, and that the reverse might indeed be the case. Small sample size and the impossibility of pooling together the results of all four studies due to various weaknesses were important problems. But particularly telling were statements by the reviewers and researchers in …