A Survival Framework for Estimating Child Mortality Rates using Multiple Data Types

2026-01-28

AI Summary

Many countries lack robust birth and death registration systems, hindering accurate tracking of dynamic changes in child mortality. To address this challenge, this work proposes the first unified Bayesian survival analysis framework that integrates heterogeneous data sources—such as household surveys and vital registration—within a single model to simultaneously estimate continuous age–time survival curves for neonatal, infant, and under-five mortality. Moving beyond conventional approaches that rely on stratified age-group modeling or multi-step procedures, the method employs flexible log-logistic and piecewise exponential survival functions for coherent estimation. Empirical applications in Kenya, Brazil, Estonia, and Syria demonstrate that the resulting estimates for the three child mortality indicators align closely with those from the UN Inter-agency Group for Child Mortality Estimation (IGME), while uniquely providing complete, continuous survival trajectories.

Technology Category

• Reasoning under Uncertainty: Graphical Models
• Machine Learning: Bayesian Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Child mortality is an important population health indicator. However, many countries lack high-quality vital registration to measure child mortality rates precisely and reliably over time. Research endeavors such as those by the United Nations Inter-agency Group for Child Mortality Estimation (UN IGME) and the Global Burden of Disease (GBD) study leverage statistical models and available data to estimate child survival summaries including neonatal, infant, and under-five mortality rates. UN IGME fits separate models for each age group and the GBD uses a multi-step modeling process. We propose a Bayesian survival framework to estimate temporal trends in the probability of survival as a function of age, up to the fifth birthday, with a single model. Our framework integrates all data types that are used by UN IGME: household surveys, vital registration, and other pre-processed mortality rates. We demonstrate that our framework is applicable to any country using log-logistic and piecewise-exponential survival functions, and discuss findings for four example countries with diverse data profiles: Kenya, Brazil, Estonia, and Syrian Arab Republic. Our model produces estimates of the three survival summaries that are in broad agreement with both the data and the UN IGME estimates, but in addition gives the complete survival curve.

Problem

• child mortality
• vital registration
• survival estimation
• data integration
• population health

Innovation

• Bayesian survival framework
• child mortality estimation
• integrated data modeling
• survival curve
• log-logistic model