COMPOSITE MODELS FOR MATERNAL AND CHILD SURVIVAL: THE CASE OF GHANA IN WEST AFRICA

Abstract

Since the adoption of the Millennium Development Goal in 2000, much interest has been given to the study of maternal and child health. However, developing reliable, informative and non static estimates still remains a challenge. Furthermore, despite several linkages among them, maternal and child survival continue to be studied independently. In this study, we apply survival analysis techniques to develop measures of maternal and child survival and develop models for them. We further develop protective tables and composite models for the joint maternal and child survivals. The Results show that while maternal and foetal survival are well described by a Weibull model, Neonatal, Infant and Child survivals are each well described by log logistic models. During pregnancy, the composites maternal and child survival, maternal survival and child mortality as well as maternal and child mortality are well described by an exponential, a logarithmic and a power model respectively. Applying these results to the Ghana maternal health 2007 data yields a protective table that shows that about 70 percent of conceptions are expected to be successful at delivery. Out of the remaining 30 percent, approximately 14 percent are expected to result in mothers without their babies and 16 percent resulting in both mother and child dead. One month after delivery, approximately 0.5 percent of successful deliveries are expected to result in mother and child dying, 7 percent of children expected to be motherless and 4 percent of mothers, childless: The remaining 88.5 percent will be mothers who survive with their babies.