OBJECTIVE: An adverse drug reaction (ADR) is a harmful, unwanted patient response to a drug that happens at doses normally used for prevention, diagnosis, or treatment. Despite limited and inconsistent primary evidence on the cost of ADRs in Africa, this systematic review aims to synthesize the magnitude of ADR-related costs and assess the reporting quality of existing studies. METHOD: The authors followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses checklist. Three databases (PubMed, Scopus, and ScienceDirect) and a hand search were employed. All empirical African cost-of-illness studies on ADRs were included regardless of publication year, disease type, or population, while letters to the editor, case reports, conference proceedings, and abstracts were excluded. Two independent authors conducted screening, data extraction, and quality assessment. A consensus-based cost-of-illness checklist was used to assess reporting quality. The primary outcome was the cost of adverse drug reactions, reported in 2024 international dollars (I) ; the secondary outcome was the reporting quality of studies. RESULTS: A total of 324 studies were identified, of which eight were included in the final analysis. Nearly all studies reported only direct medical costs, excluding direct non-medical and indirect costs. ADR-related hospitalization costs per patient ranged from I54. 75 in Nigeria to I7438. 47 in South Africa. Costs were mainly driven by hospitalization, medications, and investigations, and influenced by patient, disease, and health system-related factors. While all studies described their population, objectives, time horizon, and data collection methods, many failed to report important methodological details. CONCLUSION: Adverse drug reactions impose a substantial economic burden through direct medical costs influenced by multiple factors; however, many studies lacked key reported details. Future research should adopt recommended methodologies and comprehensive cost reporting to improve the reliability of cost estimates.
Kebede et al. (Tue,) studied this question.