COX PROPORTIONAL HAZARD AND EXPONENTIAL SURVIVAL ANALYSIS IN PATIENTS WITH END-STAGE CHRONIC KIDNEY FAILURE AT BOJONEGORO

Abstract

End-stage chronic renal failure is a condition that requires long-term treatment such as haemodialysis and poses a serious threat to patient survival. However, the survival time of each patient varies, depending on various clinical and demographic factors. Identifying variables that have a significant effect on survival time is important to help medical personnel prioritise patient care. Cox proportional hazard and exponential regression are statistical methods used to identify independent variables that affect the dependent variable, survival time. In this study, Cox proportional hazard and exponential regression survival analysis were modelled on end-stage chronic renal failure patients who were hospitalised in January-April 2024 at RSUD Dr. R. Sosodoro Djatikoesoemo Bojonegoro. This study aims to identify independent variables that have a significant influence on the survival rate of patients with end-stage chronic renal failure and the best model between Cox proportional hazard and exponential models. The Cox Proportional Hazard method is a semi-parametric method that analyses the influence of variables without having to know the specific shape of the failure time distribution. Meanwhile, the exponential model is a parametric model that assumes that the hazard function is constant over time. In this study, 10 variables were used to see their influence on the risk of occurrence. The results of Cox proportional hazard and exponential regression analysis obtained independent variables that have a significant effect, the variables of main complaint (X3), urea (X6), and diastolic blood pressure (X8) on the survival time of patients with chronic renal failure. The hazard ratio value on significant variables, the variable that can increase the risk of death, is urea. Every additional 1 mg/dL urea value will increase the risk of death of chronic renal failure patients by 0.9%. The exponential model of 383.4526 is the best model based on the AIC value.