Censorship in survival-time (time-to-event, failure-time) studies refers to incomplete data.
Examples of incomplete data are:
- individual still alive (no event) at end of study
- individual lost to follow up or left study before the end
- event not recorded properly
...in these situations the lack of completeness of data exists at the end of the study, a situation more strictly known as right censorship. Censorship in StatsDirect is used to mean right censorship.
Coding for censorship in StatsDirect is:
- 0 for censored data (no event observed)
- ≥1 for uncensored data (e.g. 1 for an individual event/death at a time point, 2 for 2 events etc. - note that 2 events can be expressed either in a single row with 2 in the censorship column or in two rows with 1 in the censorship column)
If there are problems with your data at the start of a study (e.g. start time not recorded properly) then the data are left censored. Left censored data are best avoided or dealt with by specialist Statisticians.