Background: It can be used different models such as nonparametric, semi-parametric and parametric to estimation of survival. Depending on the nature of the data, some methods may be more appropriate. Using of appropriate model result more accuracy estimation of survival and it's affected. so the aim of this study estimation of survival and determination of its related risk factor. Materials and Methods: In this study, 169 patients with colorectal cancer who had treated in Hamadan MRI Center from March 2005 to Jun 2013 were evaluated. Survival time of patients is considered from diagnosis time to death or end of study. The effect of Demographic and clinical risk factors on survival is assessed by parametric model. follow up of patient Data were analyzed using Spss16 and R statistical software.

Results: The median survival was 4.3±40 months. Univariate analysis showed that the effect of staging of tumor, tumor grade, histology type of tumor, metastasis and relapse on survival were significant but in multivariate model histology type of tumor and metastasis were significant.

Conclusion: using of appropriate statistical model for detection of affected risk factor on patient with colorectal cancer can be prevented progression of disease and increased of survival of them.