motebrian/Forecasting-HIV-new-infections-and-ART-coverage-in-Kenya-vs.-peer-countries

# Forecasting HIV Treatment Demand in Kenya (2025–2030) ### How many additional people will need to start ART by 2030 to close the treatment gap? --- ## Project Overview This project uses epidemiological data and time-series modelling to quantify Kenya's remaining antiretroviral treatment (ART) gap and forecast how many additional people living with HIV (PLHIV) will need to start treatment by 2030 to meet the UNAIDS 95-95-95 targets. The analysis combines exploratory data analysis, interrupted time-series regression, and forecasting to tell a coherent policy story — from Kenya's 1990–2024 epidemic trajectory through to actionable 2030 projections. **Key finding:** Kenya's ART coverage reached ~87% by 2024, leaving an estimated 7–8 percentage point gap to the UNAIDS 95% target. At the current rate of growth (~1.9 percentage points per year post-2015), Kenya is projected to approach but not reach the 95% target by 2030 without additional programmatic effort. --- ## Research Questions 1. How has the trajectory of new HIV infections in Kenya compared to regional peers (Uganda, Tanzania, Ethiopia) since 1990? 2. What was the measurable impact of the 2015 WHO "Treat All" policy on ART coverage in Kenya? 3. How many additional people will need to start ART by 2030 to close the treatment gap? --- ## Repository Structure ``` ├── HIV_New_Infections___ART_Coverage_Forecasting_.ipynb # Main analysis notebook ├── Forecasting.ipynb # Time-series modelling notebook ├── cleaned_hiv_data.csv # Cleaned UNAIDS estimates data ├── HIV Estimates Data.xlsx # Source data (UNAIDS) └── README.md ``` ### Notebook 1: Analysis (`HIV_New_Infections___ART_Coverage_Forecasting_.ipynb`) Covers the full analytical pipeline: - Data ingestion and cleaning (UNAIDS Excel format with mixed-type estimates) - Exploratory analysis of new infections, PLHIV, prevalence, and ART coverage acr …