ABSTRACT
The 'survival analysis' comes from the original application of the survival method to demographic data, as when one wants to calculate average life expectancy in a population. This chapter presents an example of how to perform a survival analysis, and explores the difference between risks and rates. The subjects are censored, and that could happen for any of the three reasons: the subject had not experienced the event when the study was finished, the subject was lost to follow-up from a specified date and the subject experienced an excluding event on a specified date. The distribution of time periods until some positive or negative event occurs is best studied in cohorts of subjects and analysed with survival analysis. This is especially true if the times are long, giving considerable problems with losses to follow-up during the study.
