Feed

Breast cancer is the most common cancer among women and represents 24.5% of all cancer cases worldwide. As there are few studies on the subject in Angola, it is particularly interesting to characterize the survival of women diagnosed with breast cancer in that country. The collected data set contains all diagnosed patients followed up at the Angolan Cancer Control Institute (IACC) between 2013 and 2022. However, due to the impact of COVID-19, the data set under analysis consists of the prepandemic period (2013 to 2019). The aim is to study the population diagnosed with cancer in the country, not only descriptively but also to check which factors influence the survival time of patients from diagnosis to death from breast cancer. To this end, various methods are applied to see which best describes the time until the event under study occurs. Namely, this study examines survival analysis models, beginning with the well-known Cox proportional hazards model. It then transitions to parametric models based on exponential and Weibull distributions and finally explores generalizations of these models that offer greater flexibility. To achieve this flexibility in the survival distribution, we adopt the proportional hazards framework proposed by Younes and Lachin but applying a new approach by Royston and Parmar. The results show that the median survival time from breast cancer in the observed women was 560 days. The variables that were revealed to be significant risk factors were age, stage of the disease, and body mass index. According to Akaike's criterion, the flexible proportional hazards model with one knot proved to be the most appropriate model for these data.