Pseudo-observations in survival analysis

In recent years, pseudo-observations have gained much interest as they address one of the main issues with survival data, i.e. incomplete observations due to right-censoring, and they allow for the analysis of such data by standard statistical methods. In this work, a method based on pseudo-values will be illustrated for direct regression modelling of the survival function, the restricted mean survival time, and the cumulative incidence function in a competing risks situation. The method based on pseudo-observations will be compared to the traditional Cox proportional hazard and Fine-Gray models.