Analyzing relative survival of groups that outlive the population: the French olympians study
vendredi 25 mars 2016, 9h30 - 10h30
When comparing the survival experience of a group of patients to the general population, we usually assume that their hazard due to the disease is added to the hazard they have as members of the general population. However, this may not always be a realistic assumption, on the contrary, we may deal with subgroups that survive better than the general population. The motivation for this presentation is the study of survival of French olympic athletes, which is based on a database that contains all athletes who competed from the year 1912 onwards and includes their survival status, cause of death information and several other variables of interest.
We will focus on the years lost or saved compared to the general population and the decomposition of this measure with respect to the causes of death. We also detail the usage of pseudo-values in fitting a regression model to this outcome.