Analyzing the Survival of Breast Cancer Patients Using Weibull and Poisson Beta-Weibull Non-Mixture Cure Models

  Therapies for many of diseases especially cancers have been improved significantly in the recent year, so there have been an increased number of patients who do not experience mortality. In analysis of these disease, cured models is used instead of usual survival models. Weibull model and its generalized version beta Weibull Poisson (BWP), are flexible models in cure models and are used in this study to analyze braset cancer patients data. The data of this cohort study are from patients with breast cancer, were gathered during 1997 to 2006 to the Motamed Cancer Institute in Tehran, and were followed up from 2013 to 2017. A random sample of size 270 patients was selected and individual characteristics evaluated. The data were analyzed using Stata 12 and R3.4.1 software and the significance level set at 0.05. The results showed that 43 (15.9%) of patients deaths after treatment. One, three and five year Cure of the patients was 0.99, 0.87 and 0.83 respectively. The results of this study showed that PBW Non-Mixture Cure Model (AIC=427) has better fit than weibull (AIC=593). Based on this model, variables of tumor size greater than 5,(P<0.001) and, tumor grade 3,(P<0.001) are factors affecting of patients cured; and the cure fraction was estimated to be 80%. By study the effect of factors affecting the occurrence of death, considering the unknown number of causes and using BWP models, the cancer trend can analyzed better and more accurate information can be available to researchers.