Malaria burden and ITN intervention effectiveness analysis across sub-saharan Africa 2000-2024
# Malaria Burden & Intervention Effectiveness Analysis
A full-pipeline epidemiological analysis of malaria burden and intervention coverage across sub-Saharan Africa (2000–2024), combining SQL data engineering, statistical regression, machine learning, and time series forecasting.
## The Question
Does insecticide-treated bed net (ITN) coverage significantly reduce malaria incidence and mortality across sub-Saharan Africa? Which demographic, economic, and geographic factors moderate this relationship?
## Key Findings
- Malaria incidence declined **40%** and mortality declined **65%** across sub-Saharan Africa between 2000 and 2024 — but progress has slowed markedly since 2015
- **Economic capacity (GDP per capita)** is the strongest and most consistent predictor of lower malaria burden across every method tested — regression, machine learning, and descriptive analysis alike
- **IRS (indoor residual spraying)** shows a disproportionately strong protective effect against **mortality specifically** — confirmed independently by both OLS regression and Random Forest
- **ITN coverage shows a persistent positive association with burden** even after statistical adjustment — this reflects deliberate targeting bias (highest-burden countries received the most nets), not evidence against ITN effectiveness. Country-level data cannot resolve this; individual-level data would be required
- **Central Africa carries the highest current burden and shows zero projected improvement through 2030** under current trends — the clearest policy-relevant finding in this analysis
Full methodology and limitations are documented throughout the notebooks below.
## Data Sources
| Source | Dataset | Coverage |
|--------|---------|----------|
| WHO Global Health Observatory | Malaria incidence & mortality | 2000–2024, ~46 countries |
| WHO Global Health Observatory | ITN, IRS, ACT intervention coverage | 2015–2024 |
| World Bank Open Data | GDP per capita, urban population % | 2000–2024 |
## …