Health centers network analysis with Gephi and ForceAtlas2 approach: Case of Burkina Faso

Abstract

Burkina Faso, like many developing countries, faces significant challenges in public health, particularly regarding healthcare access and infrastructure distribution. Healthcare centers are unevenly distributed across regions, resulting in disparities in access to care. This study aims to analyze the structure and efficiency of the healthcare network in Burkina Faso using graph theory, leveraging Gephi and the ForceAtlas2 algorithm for visualization. We constructed a graph representing 80 healthcare centers as nodes and the distances between them as weighted edges. By applying network theory metrics such as degree, modularity, centrality, and density, we identified the strengths and weaknesses of the healthcare network.

The analysis reveals that the healthcare network has an average degree of 0.985, indicating that most healthcare centers are connected to fewer than one other center on average. The network’s density is 0.015, showing that it is highly sparse. Modularity analysis identified eight distinct communities, with a modularity score of 0.556, reflecting a moderately well-defined community structure. The average path length is 1.45, indicating that most centers are relatively close to each other, but regional disparities remain, especially in isolated areas.

These findings suggest that improving connectivity in underserved regions could significantly enhance access to healthcare. Concrete policy actions such as deploying mobile clinics in peripheral zones, establishing intermediate logistics hubs in strategic locations, or enhancing routes toward high centrality but low degree centers can be derived from the network structure. Although this study uses standard graph theoretical tools, its contribution lies in the pragmatic application to a low-resource health system. The proposed framework is easily adaptable and reproducible for other low and middle-income countries (LMICs). Furthermore, this work provides a basis for future integration with dynamic or health outcome data, enabling more comprehensive simulations for infrastructure planning and emergency response.