Statistical Analysis of Renal Risk Factors and Prediction of Chronic Kidney Disease

An extensive world population, in particular aged people, is suffering from chronic kidney disease (CKD). Early prediction of CKD is crucial in mitigating disease complications, slowing its progression, and improving patient survival rates. This work analyzed CKD ailment-related issues using three computational approaches: (1) Several statistical methods were investigated to find the relationship between a heterogeneous risk factor and disease. In addition, different significance tests were exercised to classify the risk factors for two classes, with and without CKD. (2) A hybrid statistical approach was followed to identify the most critical risk factors significant to predict CKD. (3) Machine learning techniques were used to predict the onset of chronic kidney disease in terms of the significant risk factors. Several experiments were conducted to substantiate the efficacy of the proposed analysis and prognosis. Proposing a statistical approach that outperforms existing methods to identify the minimum number of significant risk factors and predict CKD using those factors without compromising maximum prediction accuracy strengthens the contribution of the research. Indeed, it incorporates a low-cost approach in the field of affordable healthcare.