A data-driven framework for predicting and mitigating risk of cardiovascular diseases and diabetes in Sri Lanka

Abstract

Non-communicable diseases, particularly cardiovascular diseases and diabetes have emerged as the primary causes of mortality and morbidity in Sri Lanka. This research presents a comprehensive data-driven framework aimed at predicting and mitigating disease risk using advanced statistical and machine learning methodologies. The study integrates nationally representative Self-Reported Health Survey 2021 data and environmental data, to comprehensively model disease risk across socio-demographic, behavioral, biological, and environmental determinants. This research addresses two primary objectives: To develop a robust predictive frame- work using statistical and machine learning models that accurately estimates in- dividual and population-level risk for cardiovascular diseases and diabetes and to evaluate model performance and generalizability through cross-validation techniques, ultimately proposing an iterative algorithmic workflow for effective risk prediction and management. The study addresses a critical gap in public health infrastructure, specifically the lack of integrated predictive analytics systems capable of identifying high-risk populations and forecasting disease prevalence. While some countries possess tools with limited capabilities, Sri Lanka currently lacks such systems entirely. This study addresses a key gap in existing research by introducing a comprehensive analytical framework that integrates advanced machine learning techniques to assess health risks. Moving beyond traditional single-domain analyses, it incorporates multidimensional data spanning demographic, socioeconomic, behavioral, biological, and environmental factors to support proactive and targeted health interventions in Sri Lanka. The primary contribution of this work is the development of an evidence-based, decision-support framework for health authorities that enables effective risk stratification and targeted prevention efforts. Through these contributions, the research directly advances the objectives of Sustainable Development Goal 3, which emphasizes ensuring healthy lives and promoting well-being across the population.

Date of Thesis submission	2025
Original language	English
Awarding Institution	Western Sydney University
Sponsors	Australian Government Research Training Program Scholarship
Supervisor	Liwan Liyanage (Supervisor), Lakshman Dissanayake (Supervisor), Karen Soldatic (Supervisor) & Michael O'Connor (Supervisor)