External validation of logistic-based risk in time-to-event setting

This entire exercise is based on an invalid analysis that will result in a major calibration problem. This project should be abandoned. Note that one censoring is ignored, what is being predicted is a function of the enrollment date of a patient and is not one year survival.