Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.794819
Title: Adaptable spatial agent-based facility location for healthcare coverage
Author: Olowofoyeku, Olukemi
ISNI:       0000 0004 8501 1330
Awarding Body: De Montfort University
Current Institution: De Montfort University
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Lack of access to healthcare is responsible for the world's poverty, mortality and morbidity. Public healthcare facilities (HCFs) are expected to be located such that they can be reached within reasonable distances of the patients' locations, while at the same time providing complete service coverage. However, complete service coverage is generally hampered by resource availability. Therefore, the Maximal Covering Location Problem (MCLP), seeks to locate HCFs such that as much population as possible is covered within a desired service distance. A consideration to the population not covered introduces a distance constraint that is greater than the desired service distance, beyond which no population should be. Existing approaches to the MCLP exogenously set the number of HCFs and the distance parameters, with further assumption of equal access to HCFs, infinite or equal capacity of HCFs and data availability. These models tackle the real-world system as static and do not address its intrinsic complexity that is characterised by unstable and diverse geographic, demographic and socio-economic factors that influence the spatial distribution of population and HCFs, resource management, the number of HCFs and proximity to HCFs. Static analysis incurs more expenditure in the analytical and decision-making process for every additional complexity and heterogeneity. This thesis is focused on addressing these limitations and simplifying the computationally intensive problems. A novel adaptable and flexible simulation-based meta-heuristic approach is employed to determine suitable locations for public HCFs by integrating Geographic Information Systems (GIS) with Agent-Based Models (ABM). Intelligent, adaptable and autonomous spatial and non-spatial agents are utilized to interact with each other and the geographic environment, while taking independent decisions governed by spatial rules, such as •containment, •adjacency, •proximity and •connectivity. Three concepts are introduced: assess the coverage of existing HCFs using travel-time along the road network and determine the different average values of the service distance; endogenously determine the number and suitable locations of HCFs by integrating capacity and locational suitability constraints for maximizing coverage within the prevailing service distance; endogenously determine the distance constraint as the maximum distance between the population not covered within the desired service distance and its closest facility. The models' validations on existing algorithms produce comparable and better results. With confirmed transferability, the thesis is applied to Lagos State, Nigeria in a disaggregated analysis that reflects spatial heterogeneity, to provide improved service coverage for healthcare. The assessment of the existing health service coverage and spatial distribution reveals disparate accessibility and insufficiency of the HCFs whose locations do not factor in the spatial distribution of the population. Through the application of the simulation-based approach, a cost-effective complete health service coverage is achieved with new HCFs. The spatial pattern and autocorrelation analysis reveal the influence of population distribution and geographic phenomenon on HCF location. The relationship of selected HCFs with other spatial features indicates agents' compliant with spatial association. This approach proves to be a better alternative in resource constrained systems. The adaptability and flexibility meet the global health coverage agenda, the desires of the decision maker and the population, in the support for public health service coverage. In addition, a general theory of the system for a better-informed decision and analytical knowledge is obtained.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.794819  DOI: Not available
Share: