Child mortality estimation incorporating summary birth history data

Area of Science:

• Demography
• Biostatistics
• Public Health

Background:

• Under-five child mortality is a key global health indicator, with rates concentrated in developing regions.
• Assessing progress towards Sustainable Development Goal 3.2 requires accurate child mortality data, often derived from surveys.
• Existing survey data includes Full Birth History (FBH) and Summary Birth History (SBH) with varying levels of detail.

Purpose of the Study:

• To develop a novel statistical method for analyzing child mortality trends using Summary Birth History (SBH) data.
• To accommodate known biases in SBH data and incorporate space-time smoothing for mortality rates.
• To improve the accuracy and reduce uncertainty in child mortality estimations.

Main Methods:

• A data augmentation scheme within a Bayesian framework was developed.
• Birth and death dates were introduced as auxiliary variables for SBH data.
• The method incorporates space-time smoothing and accommodates data biases.

Main Results:

• Simulations demonstrated the robustness of the approach to model misspecification.
• Incorporating SBH data reduced uncertainty compared to using only FBH data.
• The method was applied to data from Malawi, showing comparable results to the Brass method.

Conclusions:

• The developed Bayesian data augmentation method effectively analyzes child mortality using SBH data.
• This approach enhances the utility of widely available SBH data for global child survival monitoring.
• The method offers a robust alternative for estimating child mortality trends, particularly in data-scarce settings.