AI-ASSISTED PREDICTION AND PREVENTION OF CARDIOMETABOLIC DISEASES THROUGH COMMUNITY HEALTH SURVEILLANCE IN PAKISTAN

Abstract

Cardiometabolic diseases (CMDs) represent a rapidly growing public health challenge in low- and middle-income countries, including Pakistan, where late diagnosis and weak preventive surveillance systems contribute to increasing morbidity and mortality. This study aimed to develop an artificial intelligence (AI)-assisted predictive and preventive framework for cardiometabolic diseases through community health surveillance in Pakistan. A quantitative cross-sectional design was employed, integrating community-level health data collected through structured questionnaires and clinical assessments. Machine learning algorithms, including Logistic Regression, Random Forest, and XGBoost, were applied to predict CMD risk and identify key contributing factors. The findings revealed that metabolic indicators such as blood glucose, body mass index, and blood pressure were the strongest predictors of CMD risk, followed by behavioral factors including physical inactivity and smoking. Among the tested models, XGBoost demonstrated the highest predictive performance with superior accuracy and ROC-AUC values compared to traditional statistical methods. The study further highlighted the effectiveness of integrating community health surveillance systems with AI-based analytics for early detection and risk stratification. The results confirm that AI-enabled community surveillance can significantly enhance preventive healthcare strategies and support early intervention in resource-constrained settings. The study provides a scalable framework for integrating artificial intelligence into primary healthcare systems to reduce the burden of cardiometabolic diseases in Pakistan.