---

 Background and Aim: According to the importance of cancerous diseases and the increasing attention to data about cancers and also the need for more accurate and appropriate analyzes about the survival of cancerous patients, their recurrency, and the heterogeneity of the risk factors, appropriate approaches are needed to be developed. In recent years, the mixture model has been considered as a suitable model for fitting the heterogeneous survival data. In this study, Weibull mixture model was used to assess the bone marrow transplantation survival data and the effect of cancer recurrency on the mortality rate.
Material and Method: In this study, 495 individuals with leukemia who referred to Taleghani hospital in Tehran and received bone marrow transplantation were examined. Individuals had received bone marrow transplantation during 2007 to 2015. They were followed up to 2016. Weibull simple and mixture models were used to fit the data. Finally, different models were compared.
Results:  The mean survival lifetime (standard deviation) of individuals who had experienced the event was 563 days (57.4) and  999 (45.4) for censored individuals. The AIC index showed that the Weibull mixture model fitted the data better than the simple model. It could explain the heterogeneity in both recurrent and non-recurrent groups better than the Weibull simple model.
Conclusion: This study showed that the use of mixture models can perform better than simple parametric models to analyze heterogeneous survival data such as cancer data.