VALIDITY OF MATERNAL AND NEONATAL INDICATORS OF LOW-BIRTH-WEIGHT

In Burkina Faso birth-weight is not taken for the majority of newborns, since scales and skills to read them are scarce and professional health services reach only a minority of the population. Therefore the risk information conveyed by low birth weight (LBW) cannot be used for protecting those at highest risk of dying. Neonatal arm and chest circumference have been suggested as appropriate proxy measures for use by traditional birth attendants or the mothers themselves. In the first part of this paper we examine the validity of these circumference measures as proxies of birth-weight (BW) in a West African population. Neonatal arm-circumference (AC) was found to combine high sensitivity (89%) and specificity (87%) at the optimal cut-off point of 9.4 cm. The authors stress the need for establishing the optimal cut-off points for each country, where the indicator is to be used. In contrast, chest circumference had clearly less favorable validity characteristics and cannot be recommended for use in our West African population. Using neonatal AC as a proxy for birth weight has the potential to bring down infant mortality rate by concentrating scarce health care resources more efficiently on those babies with a low AC. In the second part of the paper we examine whether maternal arm circumference is a valid predictor of birth-weight, as has been claimed in the literature. Such a finding would be of paramount importance for the endeavor of bringing down LBW rates. However, our data do not support this suggestion: maternal AC has very low validity indices as indicator of birthweight. Also the classic anthropometric indicators of maternal nutrition -weight, weight-for-height and skinfold thickness- were found to be invalid predictors of BW.