This study aims to apply machine learning techniques to predict COVID-19 progression amongst patients in Botswana through risk analysis on survival and mortality rates. When the pandemic was at its peak in Botswana, the healthcare system demonstrated a lack of optimum capacity for response and resource management across the country. The study is motivated by similar studies recently undertaken around the world which have demonstrated through findings the feasibility and value of early predictions to derive insights on disease progression for the purposes of capacity planning in order to reduce the strain on overwhelmed healthcare infrastructures.

This study appreciates the existence of significant variations and context in medical and socio-demographic factors and trends in disease burden that are normally applied for the modelling purposes in different countries. Thus the study will then contribute to Botswana’s own unique case from the local data collected from medical health facilities across the country. The predictive analysis will reveal impact of local circumstances and provide effective guidance for public healthcare decisions and future policy.

The study will significantly contribute to innovative research in Botswana where the recommended models could be adapted and re-modeled to apply and prepare for future disease outcomes even in SADC