Bayesian Age Category Reconciliation for Age- and Cause-specific Under-five Mortality Estimates

TL;DR

This paper introduces a Bayesian method to unify and improve age-specific under-five mortality estimates from diverse data sources with varying age classifications, enhancing public health data accuracy.

Contribution

It develops a Bayesian approach that calibrates differently aggregated age data to produce coherent, accurate cause-specific mortality estimates, addressing a key challenge in health data analysis.

Findings

The method effectively imputes incomplete age classifications.

It improves the accuracy of cause-specific mortality estimates.

Numerical studies reveal conditions for estimator improvements.

Abstract

Age-disaggregated health data is crucial for effective public health planning and monitoring. Monitoring under-five mortality, for example, requires highly detailed age data since the distribution of potential causes of death varies substantially within the first few years of life. Comparative researchers often have to rely on multiple data sources yet, these sources often have ages aggregated at different levels, making it difficult to combine the data into a single, coherent picture. To address this challenge in the context of under-five cause-specific mortality, we propose a Bayesian approach, that calibrates data with different age structures to produce unified and accurate estimates of the standardized age group distributions. We consider age-disaggregated death counts as fully-classified multinomial data and show that by incorporating partially-classified aggregated data, we can construct an improved Bayes estimator of the multinomial parameters under the Kullback-Leibler (KL) loss. We illustrate the method using both synthetic and real data, demonstrating that the proposed method achieves adequate performance in imputing incomplete classification. Finally, we present the results of numerical studies examining the conditions necessary for obtaining improved estimators. These studies provide insights and interpretations that can be used to aid future research and inform guidance for practitioners on appropriate levels of age disaggregation, with the aim of improving the accuracy and reliability of under-five cause-specific mortality estimates.