Using Cure Models: A Five-year Survival Analysis of Breast Cancer Patients in Mazandaran Province

 The mortality rate has dropped as a result of recent considerable advancements in cancer therapies. Instead of the more common survival models, cure models are employed to analyze certain disorders. The current work uses Weibull models, which are adaptable cure models, to assess data from breast cancer patients.

 The data set of the current descriptive study are from patients with breast cancer, who were referred to Imam Khomeini’s hospital in Sari County during 2014-2017 and were followed in 2021. Individual characteristics of 221 patients were recorded overall using the census method. The data were analyzed using SPSS 23 and R 3.6.3 packages with the significance level set at 0.05.

The results showed that 99 (45%) of patient’s deaths occurred after treatment. The rate of cure after one, three, and five years was found to be 0.97, 0.84, and 0.75 respectively. The results of the fitting model showed that Weibull’s Non-Mixture Cure model (AIC=659) has a better fit compared to the other cure models. According to this model, factors affecting a patient's recovery include their age at diagnosis, smoking, metastasis, chemotherapy, and need for emergency hospitalization. Also, the cure rate was estimated to be 68%.

Weibull's Non-Mixture Cure model, modeling with Poisson distribution, and studying the effects of factors influencing the occurrence of death allow for a better analysis of the cancer trend and the provision of more precise data for researchers.