Introduction: Chronic kidney disease (CKD) is an irreversible kidney function that can lead to the end stage of the renal disease. The aim of this study is to determine the relationship between blood Creatinine levels and survival of patients with kidney disease using joint longitudinal and survival model.

Methods: In this descriptive-longitudinal study, demographic and disease information of 112 patients with chronic renal failure that referred to a Medical-educational center in Gonbad-e Kavus for hemodialysis purposes between 2012 and 2014. Longitudinal creatinine levels and survival (i.e. the time interval between the beginning of the study until death or the end of the patients’ study) are considered as response variables. Data is analyzed by Joint model for assessing effective factors on survival and the result is compared with survival and longitudinal models. The models are fitted by R 3-1-3 and the significant level is considered as 0.05.

Results: In joint model, factors such as marital status, initial amount of hemoglobin, hemoglobin, blood urea and phosphorus had statistically significant relationship with survival patients’ survival. In addition, the risk of death and sex, age, marital status, diabetes, duration of dialysis, initial amount of creatinine had statistically significant relationship with creatinine level. Moreover, results demonstrate a correlation between the survival and longitudinal creatinine models. In other words, a unit decrease in creatinine level leads to 0.34 increase in risk of death.

Conclusions: This study shows that the standard error of coefficients in the joint model is less than longitudinal and survival models. The use of joint model is recommended for binary correlated responses in the medicine and public health fields.