Improving Estimates of Childhood Mortality: Comparison of Pregnancy History and Birth History for Assessment of Childhood Mortality
Childhood mortality rates are important indicators of health status of low- and middle-income countries. However, they are complex to measure. Direct estimation of these indicators relies on retrospective full birth histories through information on all live births and child deaths experienced by women age 15-49 years. While birth history approach has been widely used to estimate childhood mortality in LMICs, another potentially superior approach of data collection is the full pregnancy history. This latter approach has the added advantage of allowing estimation of additional indicators such as pregnancy loss, still births and perinatal mortality rate. In addition pregnancy history is likely to better capture neonatal deaths than birth history. However scientific publications comparing these two approaches are lacking. We propose to follow-up a sub-sample of women interviewed during a household mortality survey conducted by the Institute for International Programs in two districts in Malawi that included a full birth history. We will implement a full pregnancy history and compare both approaches in terms of completeness of recording of births and deaths, accuracy in placement of events in time, and accuracy in early neonatal, neonatal and under-five mortality rates. In addition, we will compute still birth and perinatal mortality rates and compare them with rates from other sources. The current project will provide a basis for improving the measurement of neonatal and under-five mortality in Malawi, as well as new estimates of stillbirths and rates of perinatal and early neonatal mortality.