Hi all

I have been asked to use life table analysis to calculate life expectancy for cancer patients from birth.
I am comparing 3 different groups - 1. diagnosed 1980-1990, 2. diagnosed 1990-2000 3. diagnosed 2000-2010.

I have follow up data until 2013 for each group.

I feel that I could calculate life expectancy for group 1 as I have between 23 and 33 years of follow up but do not know if I can do the same for group 3 (or maybe even group 2) due to the lack of follow up and high rates of censoring.

How much follow up would be needed to get reasonable life expectancy estimates?
How much censoring is too much?

Thank you in anticipation!