Temporal models for demographic and global health outcomes in multiple populations: Introducing a new framework to review and standardize documentation of model assumptions and facilitate model comparison

TL;DR

This paper introduces TMMPs, a standardized framework for documenting and comparing diverse temporal models used in estimating demographic and global health indicators across multiple populations.

Contribution

It proposes a new class of models, TMMPs, that standardizes assumptions documentation and enables comparison of different modeling approaches in global health estimation.

Findings

TMMPs can represent existing models for various health indicators.

The framework facilitates systematic comparison of model assumptions.

It highlights the importance of distinguishing latent trends from observed data.

Abstract

There is growing interest in producing estimates of demographic and global health indicators in populations with limited data. Statistical models are needed to combine data from multiple data sources into estimates and projections with uncertainty. Diverse modeling approaches have been applied to this problem, making comparisons between models difficult. We propose a model class, Temporal Models for Multiple Populations (TMMPs), to facilitate documentation of model assumptions in a standardized way and comparison across models. The class distinguishes between latent trends and the observed data, which may be noisy or exhibit systematic biases. We provide general formulations of the process model, which describes the latent trend of the indicator of interest. We show how existing models for a variety of indicators can be written as TMMPs and how the TMMP-based description can be used to compare and contrast model assumptions. We end with a discussion of outstanding questions and future directions.