robynhathaway/RPDA-PHC-Project

Mixed-effects regression model to investigate the associations between digital indicators and primary healthcare utilisation in South Africa. # RPDA-PHC-Project This repository contains the processed dataset and reproducible code used to analyse the associations between district-level primary healthcare (PHC) utilisation and digital readiness in South Africa. The data and code underpin the mixed-effects modelling and Poisson-based imputation procedures described in the accompanying research report. --- ## Repository Structure **Data Files** - `data/RPDA_Merged_Dataset.csv` – Consolidated dataset (2022–2024; 156 observations) - `data/PHC_Project_Dataset.csv` – Final project dataset used for analysis (includes imputed values) **Jupyter Notebooks** - `notebooks/PHC_Project_Imputation.ipynb` – Poisson regression imputation model and diagnostics - `notebooks/PHC_Project_EDA.ipynb` – Exploratory data analysis - `notebooks/PHC_Project_Model.ipynb` – Main regression model **Documentation** - `README.md` – Repository description and access details --- ## Dataset Description **File:** `data/PHC_Project_Dataset.csv` **Contents:** 156 records (52 districts across 9 provinces × 3 years) **Purpose:** Used for district-level analysis of PHC workload, digital access, and system capacity. **Provenance:** Constructed from Statistics South Africa and Health Systems Trust sources (see Appendix A in the report). All variables are fully de-identified and correspond to the variable definitions listed in *Appendix B* of the research report. ---