Machine Learning-Based Prediction of Chronic Kidney Disease Progression Using Clinical and Biochemical Indicators

Abstract

Chronic kidney disease (CKD) remains a major global health concern, with increasing prevalence and burden from 2023 to 2025. Accurate prediction of CKD progression is essential for early intervention and optimal patient management. In this study, we developed a machine learning–based framework to predict CKD progression using clinical and biochemical indicators derived from multi-center datasets collected between 2023 and 2025. A total of 5,200 patient records were analyzed, including demographic information, blood pressure, estimated glomerular filtration rate (eGFR), serum creatinine, urinary albumin-to-creatinine ratio (ACR), fasting glucose, and lipid profile. Five machine learning algorithms—logistic regression, random forest, support vector machine, gradient boosting, and deep neural networks—were compared in terms of predictive accuracy, sensitivity, and specificity. The gradient boosting model achieved the best performance, with an AUC of 0.91, sensitivity of 87%, and specificity of 85%, outperforming traditional statistical approaches. Feature importance analysis revealed that baseline eGFR, serum creatinine, systolic blood pressure, and urinary ACR were the most significant predictors of CKD progression.