ABSTRACT
In classical survival analysis, one considers the time from an origin to a single event of interest. However in many applications, and in particular in medical applications, different types of event may occur and the occurrence of one type of event may preclude or modify the risk of occurrence of the other ones. Many examples occur in oncology, where the long follow-up of the patients and the disease history make them particularly prone to the occurrence of different event types. For example, the question of long-term adverse effects is often encountered, for example, in case of early-stage breast cancer or prostate cancer, and often subjected to competing risks. Evaluating the risk-benefit ratio of treatment is then a complicated task as patients with better prognostic appear thus at higher risk of experiencing long-term side effects and secondary malignancy. Competing risks data are very frequent in oncology but also occur in a wide variety of other situations.
