Back to search

University of Southampton

Innovating with routinely collected health data to estimate population denominators for health facilities of sub-Saharan Africa: case studies from Ghana’s Eastern region

Abstract

dc:description.abstract

Health facilities may be an appropriate spatial unit for health system management in sub-Saharan Africa (SSA). A lack of suitable demographic data has complicated the production of facility-level catchment areas and population denominators to facilitate these activities. Facilities generate routinely collected health data (RCHD) as a product of providing services to the population. Although these data have been under-utilised in SSA, recent developments, including widespread uptake of District Health Information Software (DHIS2), may now expand their scope as a source of health intelligence. This ‘four-paper’ thesis explores innovative uses of RCHD in facility-level catchment/denominator estimation, focussing on childhood diarrhoeal disease in Ghana’s Eastern region.<br/>The first paper presents a scoping review appraising existing approaches to facility-level catchment/denominator estimation from SSA. Most were implemented in a single/small number of facilities only, signifying a need for approaches that can generate intelligence to support decision-making for local-/national-level facility networks alike. Studies increasingly estimated catchments via geostatistical/travel time-based approaches and denominators by intersecting catchment boundaries with gridded population surfaces. Few considered inter-facility competition or ‘aspatial’ determinants of health-seeking, and none used the aggregated service utilisation data collected within DHIS2.<br/>The second paper evaluates a data-driven approach to denominator/intervention coverage estimation. Initially developed for first-/second-level administrative units, it was adapted here for facility-level estimation. Shared utilisation patterns are assumed for a benchmark intervention with near-complete coverage (denominator) and target intervention (numerator) used by the same population. The variability and proportion of implausible (>100%) vaccination coverage estimates increased through lower levels. All estimates of diarrhoea service utilisation fell short of survey-derived equivalents. Facility-level denominators, especially, may be compromised by suboptimal data quality, health system reconfiguration, non-viable benchmark interventions and inconsistent health-seeking patterns.<br/>The third/fourth papers, together, use a Huff-modified two-step floating catchment area model (H2SFCA) to estimate catchments and spatially explicit denominators. This approach defined the location-specific probability of facility utilisation as dependent upon spatial (travel time) and aspatial (‘attractiveness’) factors. Under the hypothesis that facilities capable of providing more services gain attractiveness to health-seekers, the third paper derives a facility-level child health ‘service provision’ index from RCHD. The fourth paper integrates the index and a cost impedance model within the H2SFCA to delineate catchments, then computes denominators by intersection with fine-scale modelled estimates of childhood diarrhoeal prevalence. In general, facilities receiving the highest scores gained coverage when delineating catchment boundaries using the H2SFCA rather than a travel time threshold. The findings suggest that this method of adjusting for inter-facility competition aligns with the real-world geography of health-seeking in SSA.<br/>Intelligence describing the health needs of populations in small areas is needed to realise the Sustainable Development Goals’ mantra of ensuring that no one is left behind. This thesis fits with the growing interest in using RCHD to generate intelligence for local progress monitoring and evidence-based decision-making in SSA: its outputs could be used to guide targeted resource allocation and the design of context-appropriate intervention programmes. This work also provides use cases that may stimulate further investment in health data infrastructure and contribute to RCHD being established as a valued input to routine decision-making in Ghana. It may also have more far-reaching application in other DHIS2-adopter countries. The thesis advances the literature: its case studies were designed to address methodological needs identified by the scoping review and further our understanding of how RCHD from SSA might be leveraged for facility-level analyses that can be scaled to national coverage.

Degree

thesis:*
Name dc:type.qualificationname
Ph.D.
Level dc:type.qualificationlevel
doctoral
Grantor dc:publisher.institution
University of Southampton
Year dc:date.issued
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Johnson, Matthew James
Advisors dc:contributor.advisor
  • Utazi, Edson
  • Mcgrath, Nuala
  • Wright, Jim

Rights

Language dc:language
English, eng

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
University of Southampton
Base URL
eprints.soton.ac.uk/cgi/oai2
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Johnson, Matthew James. Innovating with routinely collected health data to estimate population denominators for health facilities of sub-Saharan Africa: case studies from Ghana’s Eastern region. doctoral thesis, University of Southampton, 2026. https://doi.org/10.5258/SOTON/PG/T149