Title: Bayesian analysis of cox proportional hazard model alongside competing risk model for breast cancer
Authors: Sandra Ramirez - Universidad Politecnica de Valencia (Spain) [presenting]
Abstract: The effect of several well-known risk factors influencing the onset of cancer on life expectancy after the disease has been detected and treated is studied. To study expected survival, a sample of 1,017 women with breast cancer was used with continuous follow-ups over a period of 30 years, classified according to a set of risk factors, and evaluating their survival probabilities using Cox proportional hazard model and competing risk model. The results for four groups were analyzed separately, patients with children and post-menopause, patients with children and pre-menopause, patients without children and post-menopause, and patients without children and pre-menopause for each type of model. Some risk factors identified for death by any cause were: older age at the time of first childbirth, older age of experiencing menopause, having a history of cancer. Some risk factors recognized for death due to breast cancer were: older age of experiencing menopause, having a history of cancer, never breastfeeding a child, older age at the time of first childbirth and older age of experiencing menopause. Furthermore, the risk of mortality from breast cancer was the dominant cause of death over the study period of breast cancer.