Censoring distributions as a measure of follow‐up in survival analysis

The mean or median of the follow-up times as a measure of the quality or completeness of the follow-up in a survival analysis can sometimes be misleading--a more severe disease will have shorter follow-up times because of earlier deaths. We suggest here that estimates of the censoring distribution provide a more useful measure of the follow-up. This paper gives estimation procedures for both grouped-time (cohort) data and continuous data, and provides an application.