Last Updated: 12/09/2022

Zambia Catchment Modeling


The main objective of this sub-project is to develop robust geostatistical framework for estimating health facility catchment area populations. Integrate features from individual facilities, and allow catchments to have geographic overlap. Validate using field survey data.

Principal Institution

PATH, United States

Principal Investigators / Focal Persons

Justin Millar

Rationale and Abstract

Facility-level population denominators required for estimating incidence rates are difficult to maintain within surveillance systems. Reported headcount data are infrequently updated and may not respond to changes within a specific facility or neighboring facilities (e.g., commodity stockouts, opening/closing of nearby facilities). Simple geospatial methods, such as the nearest facility by distance or travel time, may not be representative of treatment-seeking behavior and may result in skewed rate estimates. 

Gravity models offer a natural method for estimating spatial accessibility patterns and incidence rates from aggregate facility data. Benefits of this method include accounting for differences between individual facilities, incorporating information on health system organization and referral structures, and creating geographically-overlapping catchment areas. While gravity models have been used in some specific facility-level research studies, there is a lack of tools and resources for standardized implementation. 

In this project, we develop a gravity model for estimating health facility catchment populations in Zambia, using data facility-level data from DHIS2 and open source data on population density and travel time estimation. These estimates are used to determine catchment level incidence rates from cases reported into DHIS2. 

Study Design

This study will use open source data from World Pop and Malaria Atlas Project, as well as health facility data from Zambia DHIS2 system.


Related Resources

Related Projects