dushimep6/Prediabetes-Risk-Predictor-Rwanda

Bilingual self-assessment decision support tool for prediabetes risk prediction among Rwandan adults # Prediabetes Risk Predictor — Rwanda A bilingual self-assessment decision support tool for prediabetes risk prediction among Rwandan adults, developed using the 2021–2022 Rwanda WHO STEPS survey data. ## About This tool uses a Logistic Regression model (AUC=0.9885) trained on 5,010 Rwandan adults to predict individual prediabetes risk. It classifies users into: - 🟢 **Low Risk** — probability < 20% - 🟡 **Moderate Risk** — probability 20–58.75% - 🔴 **High Risk** — probability ≥ 58.75% ## Features - Available in **English** and **Kinyarwanda** - Tab 1: Basic assessment (no equipment needed) - Tab 2: Optional clinical measurements (waist circumference, cholesterol) - Personalised lifestyle recommendations per risk zone ## Study **Title:** Development and Evaluation of Machine Learning Models for Prediabetes Risk Prediction Among Rwandan Adults and Translation into a Bilingual Self-Assessment Decision Support Tool **Author:** Pasteur DUSHIMIMANA **Institution:** University of Rwanda, College of Medicine and Health Sciences **Supervisor:** Assoc. Prof. Gerard RUSHINGABIGWI ## Data Rwanda WHO STEPwise Approach to Surveillance (STEPS) Survey 2021–2022 IRB Approval: CMHS/IRB/349/2026 ## Usage Enter your health information in Tab 1. Optionally add measurements in Tab 2 for a more precise estimate. Click **Calculate Risk** to see your result and personalised recommendations.