Improving Survey Data on Births and Neonatal Deaths in Low-Income Countries

Associate Name: Stephane Helleringer

Funding Source/Period of the Grant: NICHD R21 4/11/16-3/31/18


The risk of dying under the age of 5 has declined significantly worldwide. This trend is primarily due to reductions in diseases and conditions that affect children aged 1 to 59 months old (e.g., measles, diarrheal diseases, pneumonia, Malaria). As a result, an increasing proportion of remaining children deaths occurs during the first 28 days of life. New health programs funded by federal institutions (e.g., NIH, CDC, USAID) and international organizations, now invest large amounts of resources in preventing such neonatal deaths in low- income countries. But the impact of such investments remains unclear because robust measures of neonatal mortality rates are difficult to obtain, particularly in countries of sub-Saharan Africa and south Asia where the risk of neonatal deaths is the highest. In these countries, vital registration and health information systems are too inadequate to yield robust counts of a) the total number of births occurring in a population, and b) the number of deaths that occurred during the first 28 days of life. Instead, data on neonatal mortality are collected retrospectively during population-based surveys such as the Demographic and Health Surveys (DHS). Women aged 15-49 years are asked to recall their birth history and report whether each of their children is still alive or has died prior to the survey. Such data may misrepresent the extent of neonatal mortality due to 1) under- reporting of births, particularly those that resultedin a neonatal death, 2) misreporting of the age at death of a child (e.g., neonatal deaths reported as occurring during the post-neonatal period) and 3) misclassifications of stillbirths as neonatal deaths and vice-versa. In this project, we will test whether a new survey instrument helps improve the quality of data on neonatal deaths during surveys. This instrument - the mobile birth history calendar (mBHC) - uses simple tools that have been widely and successfully used in other areas of survey research, but have not been used to improve neonatal mortality data. These include, for example, recall cues, anchoring vignettes and event history calendars. To further improve the effectiveness of these tools, we will integrate them into a flexible mobile dat collection platform (ODK) increasingly used for data collection in population-based surveys in low-income countries. We will then conduct a randomized controlled trial of the new mBHC in Guinea-Bissau, which will determine whether this new instrument improves mortality data relative to the standard instrument currently in use. If successful, the proposed mBHC will constitute a new approach to eliciting survey data on neonatal mortality that can be tested on a larger scale and possibly incorporated in national surveys of adult mortality (e.g., DHS). The source code for the mBHC will be made publicly-available so that other data collection initiatives can easily integrate similar tools into their study instruments. (From NIH Report)