The Silent Epidemic: Why 60% of Africans with Diabetes Don’t Know They Have It

## The Stark Reality of Sub-Saharan Africa’s Diabetes Burden

According to the International Diabetes Federation (IDF), **over 24 million adults** in Africa live with diabetes, and a staggering **54% to 62% of them are completely undiagnosed**.

Most patients in Zimbabwe only find out they have diabetes after presenting at a hospital with severe late-stage complications: irreversible vision loss (retinopathy), non-healing foot ulcers requiring amputation, or sudden renal failure.

---

Why Traditional Screening Fails in Low-Resource Settings

• 1. **Geographical Distance to Clinics:** Rural and peri-urban patients must travel hours and pay transportation costs simply to access a blood test.
• 2. **Laboratory Reagent Stockouts:** Centralized clinics frequently experience reagent and strip shortages.
• 3. **Out-of-Pocket Cost:** Routine preventative checkups are often deprioritized when families are managing daily living expenses.

---

How Diabetos Machine Learning Democratizes Early Detection

Diabetos was engineered from the ground up to eliminate diagnostic friction. Using an advanced **XGBoost gradient-boosted decision tree architecture**, our model evaluates non-invasive clinical risk biomarkers:

• Age, BMI, and waist-to-height ratio
• Family history of metabolic disease
• Hypertension and cardiovascular history
• Physical activity levels and dietary grain composition
• Sub-clinical symptoms (nocturia, unquenchable thirst, slow wound healing)

With a validated **98.4% diagnostic sensitivity**, the Diabetos screening engine flags individuals in early prediabetic stages without requiring upfront blood tests—empowering patients and community health workers to take immediate action through mobile apps and USSD.