Estimation of Health and Demographic Indicators with Incomplete Geographic Information

TL;DR

This paper introduces a spatial hierarchical model to accurately estimate health and demographic indicators using incomplete or imprecise geographic data from household surveys, improving subnational estimates in low and middle income countries.

Contribution

It develops a fast computational algorithm for spatial modeling that accounts for location inaccuracies, enhancing the analysis of survey data with uncertain geographic information.

Findings

Model outperforms naive methods in simulations

Provides more accurate subnational estimates

Efficient algorithm reduces computational burden

Abstract

In low and middle income countries, household surveys are a valuable source of information for a range of health and demographic indicators. Increasingly, subnational estimates are required for targeting interventions and evaluating progress towards targets. In the majority of cases, stratified cluster sampling is used, with clusters corresponding to enumeration areas. The reported geographical information varies. A common procedure, to preserve confidentiality, is to give a jittered location with the true centroid of the cluster is displaced under a known algorithm. An alternative situation, which was used for older surveys in particular, is to report the geographical region within the cluster lies. In this paper, we describe a spatial hierarchical model in which we account for inaccuracies in the cluster locations. The computational algorithm we develop is fast and avoids the heavy computation of a pure MCMC approach. We illustrate by simulation the benefits of the model, over naive alternatives.