A Scalable Machine Learning Strategy for Chronic Kidney Disease Screening Across U.S. Healthcare Systems

Abstract

Chronic Kidney Disease (CKD) is a progressive and frequently underdiagnosed condition that poses a significant public health burden within the U.S. healthcare system. Late-stage detection often results in increased mortality and higher treatment costs, highlighting the need for effective early screening solutions. This study investigates the application of machine learning techniques for early-stage CKD prediction and classification using a reliable clinical dataset obtained from the UCI Machine Learning Repository. Four supervised learning algorithms Support Vector Machine (SVM), AdaBoost (AB), Linear Discriminant Analysis (LDA), and Gradient Boosting (GB) are implemented and evaluated to identify the most accurate predictive model. Model performance is assessed using multiple evaluation metrics, including accuracy, precision, recall, and F1-score. Experimental results demonstrate that the Gradient Boosting classifier achieves the highest predictive accuracy of 99.80%, outperforming the other approaches. The findings suggest that optimized machine learning models can support early CKD detection and enhance clinical decision-making in U.S. healthcare settings.