Volume 53 - Article 26 | Pages 821–896  

A parametric survival model for child mortality using complex survey data

By Taylor Okonek, Katie Wilson, Jon Wakefield

Abstract

Background: Accurate and precise estimates of the under-5 mortality rate (U5MR) are an important summary of the health of a population. Full survival curves on the entire age range are additionally of interest to better understand the pattern of mortality in children under 5. Modern demographic methods for estimating a full mortality schedule for children have been developed for countries with good vital registration and reliable census data but perform poorly in many low- and middle-income countries (LMICs).

Objective: In LMICs, the need to utilize nationally representative surveys to estimate U5MR requires additional statistical care to mitigate potential biases in survey data, acknowledge the survey design, and handle aspects of survival data, such as censoring and truncation. We wish to develop parametric and nonparametric approaches for estimating under-5 mortality across time that appropriately utilize complex survey data.

Contribution: We propose a parametric approach that is particularly useful in scenarios where data is sparse and estimation may require stronger assumptions. The nonparametric approach we propose provides an aid to model validation. We compare a variety of parametric models to two existing methods for obtaining a full survival curve for children under the age of 5 and argue that a parametric, survey-weighted (pseudo-likelihood) approach is advantageous in LMICs. We apply our proposed approaches to survey data from four LMICs in sub-Saharan Africa. All code for fitting the models described in this paper are available in the R package pssst.

Author’s Affiliation

• Taylor Okonek - Macalester College, United States of America EMAIL
• Katie Wilson - University of Washington, United States of America EMAIL
• Jon Wakefield - University of Washington, United States of America EMAIL

Other articles by the same author/authors in Demographic Research

A probabilistic model for analyzing summary birth history data
Volume 47 - Article 11

Most recent similar articles in Demographic Research

Analysing migrant fertility using machine learning techniques: An application of random survival forest to longitudinal data from France
Volume 53 - Article 21    | Keywords: fertility, immigrants, machine learning, random survival forest, survival analysis

The use of mobile phone surveys for rapid mortality monitoring: A national study in Burkina Faso
Volume 52 - Article 16    | Keywords: age-specific mortality patterns, data quality, Demographic Health Surveys, direct estimation, health and security crises, low-and-middle-income countries, mobile phones, mortality, sample selection, surveys, under-five mortality

Losing the female survival advantage: Sex differentials in infant and child mortality in Pakistan
Volume 50 - Article 15    | Keywords: child mortality, family, gender discrimination, Pakistan, sex differentials, son preference, South Asia, survival analysis

The question of the human mortality plateau: Contrasting insights by longevity pioneers
Volume 48 - Article 11    | Keywords: age, France, Gompertz mortality, mortality, mortality plateau, older population, parametric models, supercentenarians, survival analysis, trajectories

The bootstrap approach to the multistate life table method using Stata: Does accounting for complex survey designs matter?
Volume 47 - Article 23    | Keywords: complex survey, Health and Retirement Study (HRS), health expectancies, multistate life tables, race/ethnicity, Stata