Big Data Analytics Framework for Healthcare Knowledge Management in Resource-Constrained Hospital Environments: Evidence from Sub-Saharan Africa

Abstract

Background: Healthcare institutions in sub-Saharan Africa are increasingly generating large volumes of clinical, administrative, and patient-generated data. However, the knowledge value of these data remains substantially untapped due to inadequate analytical infrastructure, workforce capacity constraints, and the absence of scalable, cost-efficient knowledge management frameworks tailored to low-resource clinical environments. Problem Statement: Existing big data analytics platforms are predominantly designed for high-resource settings with robust computational infrastructure, making them unsuitable for deployment in Nigerian hospital systems characterized by intermittent connectivity, limited server capacity, and heterogeneous EHR systems. Proposed System: This paper proposes HDAKM-Africa, a Healthcare Data Analytics and Knowledge Management framework optimized for resource-constrained hospital environments, integrating a lightweight Apache Spark-based distributed processing engine with a clinical knowledge ontology and tiered data storage architecture. Methodology: The framework was evaluated using 820 synthetic patient episodes generated to reflect clinical data patterns from three Nigerian tertiary hospitals. Key Findings: HDAKM-Africa achieved a data processing accuracy of 89.4%, knowledge retrieval precision of 86.8%, and an average query response time of 127 ms — representing a 43% reduction in processing latency compared to conventional centralized analytics systems. Applications: The framework supports clinical decision support, epidemiological knowledge discovery, hospital resource allocation, and disease surveillance.