Application of Parametric Models to a Survival Analysis of Hemodialysis Patients

BACKGROUND: Hemodialysis is the most common renal replacement therapy in patients with end stage renal disease (ESRD). OBJECTIVES: The present study compared the performance of various parametric models in a survival analysis of hemodialysis patients. METHODS: This study consisted of 270 hemodialysis patients who were referred to Imam Khomeini and Fatima Zahra hospitals between November 2007 and November 2012. The Akaike information criterion (AIC) and residuals review were used to compare the performance of the parametric models. The computations were done using STATA Software, with significance accepted at a level of 0.05. RESULTS: The results of a multivariate analysis of the variables in the parametric models showed that the mean serum albumin and the clinic attended were the most important predictors in the survival of the hemodialysis patients (P < 0.05). Among the parametric models tested, the results indicated that the performance of the Weibull model was the highest. CONCLUSIONS: Parametric models may provide complementary data for clinicians and researchers about how risks vary over time. The Weibull model seemed to show the best fit among the parametric models of the survival of hemodialysis patients.