Modeling Time to Develop End stage for Chronic Kidney Disease: A comparison of Parametric Survival Models

Abstract

Background: End-stage renal disease (ESRD) is the last stage of chronic kidney disease which is a world-wide public health problem and it is associated with adverse outcomes of kidney failure, cardiovascular disease and premature death. Nonadherence to hemodialysis continues to impact on the care of ESRD patients in the resource limited countries especially in Ethiopia. Objective: The main objective of this study was to model the time to develop end-stage for Chronic Kidney Disease(CKD) using various parametric survival models and to compare their relative performance. Methods: The retrospective study was conducted from November 1,2019 to January 30,2022. In this study,Akaie information criteria(AIC),Bayesian information criteria(BIC) and log-likelihood were used to compare the best model among parametric survival models(exponential,wei bull,loglogistic and Log-normal using End-stage renal disease (ESRD) data set). Result: A sample of 300 CKD patients who take treatment was taken from JUMC recorded from November 2019 to January 2022. Of this 180 (60%) were male, 120(40%) were female, 210(65%) were residence from rural areas. Out of 300 CKD patients 239 (79%) were develop ESRD and 61 (20%) were censored. From the history of other additional diseases among the selected CKD patients about 219(73%), 30 (10%), 249 (83%), 224 (74.6%) ,234(78%)and 166(55%) have a history of DMI,HIV, DMII,HTN,Anemia and CHD respectively. The median survival time of ESRD for CKD patients was 9 months. The majority of the participants with ESRD were male [150 (83.33%)] and [89(74.16%)] were female. The important predictors of time to ESRD at 0.05 level of significance were Diabetic type II (DMII)[φ=0.808(95% CI:0.661-0.988)] and Hypertension(HTN)[φ=1.198(95% CI:1.004-1.431)]. Conclusion: Parametric survival model with weibull distribution was found appropriate to our data set and could be taken as the best fitted model for time to ESRD data as compared to other accelerated failure time models. The result of multivariable Weibull model showed that Hypertension(HTN) and Diabetic mellitus type 2 (DMII) were predictors that have significant effect on the development of ESRD. Recommendation: From the results ,Diabetes Mellitus type 2 and Hypertension are Key Risk Factors For Kidney Disease. So, we recommends that health experts give attention to those CKD patients in order to reduce the time to develop ESRD from CKD in relation to those risk factors.