Raufiat/west-africa-health-expenditure-analysis

Out-of-Pocket health expenditure analysis and ML risk prediction across 15 West African nations using World Bank data # Health Expenditure as a Poverty Driver: OOP Costs & UHC Gaps in West Africa ## Overview This project analyzes Out-of-Pocket (OOP) health expenditure across 15 West African nations over a 22-year period (2000–2022), examining how direct medical costs impact Universal Health Coverage (UHC) and drive household poverty. ## Key Findings - **Nigeria** recorded the highest average OOP expenditure at **71.47%** — citizens pay the majority of their medical costs directly - **The Gambia** recorded the lowest at **19.23%** — functioning tax systems and NGO presence make a measurable difference - **8 out of 15 countries** exceed the 50% high-risk threshold - Strong negative correlation (**r = -0.553**) between OOP expenditure and UHC coverage - ML model achieved **94.2% accuracy** predicting high-risk countries (AUC: 0.929) ## Tools & Technologies - **Python** — pandas, numpy, matplotlib, seaborn, scikit-learn - **Machine Learning** — Random Forest, Logistic Regression, Decision Tree - **Data Source** — World Bank World Development Indicators (WDI) ## Project Structure - Data cleaning and reshaping (melt, merge) - Exploratory Data Analysis (EDA) - Statistical correlation analysis - Machine Learning risk classification - Policy recommendations ## Policy Recommendations 1. Target funding to 8 high-risk countries exceeding 50% OOP threshold 2. Support National Health Insurance Programs in high-burden nations 3. Establish emergency micro-insurance safety nets for vulnerable households ## Author **Abdullateef Raufiat** - Kaggle: kaggle.com - LinkedIn: linkedin.com