A Hierarchical Model for Estimating HIV Epidemics

TL;DR

This paper introduces a hierarchical modeling approach for estimating HIV epidemics at sub-national levels, improving accuracy in data-scarce areas by leveraging similarities across regions.

Contribution

It presents a novel hierarchical model that enhances HIV epidemic estimates by sharing information across regions, addressing data variability and scarcity issues.

Findings

Hierarchical model outperforms independent models in predictive accuracy.

The approach efficiently utilizes existing data without refitting for each region.

Demonstrated effectiveness across multiple Sub-Saharan African countries.

Abstract

As the global HIV pandemic enters its fourth decade, increasing numbers of surveillance sites have been established which allows countries to look into the epidemics at a finer scale, e.g. at sub-national level. However, the epidemic models have been applied independently to the sub-national areas within countries. An important technical barrier is that the availability and quality of the data vary widely from area to area, and many areas lack data for deriving stable and reliable results. To improve the accuracy of the results in areas with little data, we propose a hierarchical model that utilizes information efficiently by assuming similar characteristics of the epidemics across areas within one country. The joint distribution of the parameters in the hierarchical model can be approximated directly from the results of independent fits without needing to the refit the data. As a result, the hierarchical model has better predictive ability than the independent model as shown in examples of multiple countries in Sub-Saharan Africa.