Hi everyone,
I have a question about conducting conditional relative survival analysis in Stata.
I want to analyse 5-year conditional relative survival (counted from date of diagnosis) of patients, conditional on surviving at least 2 years.
I understand that this requires adjusting the 'origin' and 'enter' options in the stset command.
My code is as follows:
stset survyears, failure(vit_stat==1) origin(time 2) enter(time 2) id(id) (with survyears defined as the number of days between date of diagnosis and death or last moment of follow-up, divided by 365.25)
I hope this code is correct?
After this code, I use the strs command to calculate relative survival.
Now what I am in doubt about, is whether I should adjust the variables 'age' and 'year' in my dataset as well (add 2 to both), which are used in the strs command to merge with the population mortality data.
I thought, as I'm trying to calculate survival conditional on having already survived 2 years, that these variables perhaps need to be adjusted as well in order for Stata to use the correct mortality data.
For instance, a patient who was 50 years old at diagnosis and who is 52 after 2 years -> does adjustment of 'origin' and 'enter' automatically take into account that the probability of surviving from the second year onwards is now different because this patient is two years older en now lives in a different calender year? I think it doesn't but I am not sure?
Thank you in advance for the response
I have a question about conducting conditional relative survival analysis in Stata.
I want to analyse 5-year conditional relative survival (counted from date of diagnosis) of patients, conditional on surviving at least 2 years.
I understand that this requires adjusting the 'origin' and 'enter' options in the stset command.
My code is as follows:
stset survyears, failure(vit_stat==1) origin(time 2) enter(time 2) id(id) (with survyears defined as the number of days between date of diagnosis and death or last moment of follow-up, divided by 365.25)
I hope this code is correct?
After this code, I use the strs command to calculate relative survival.
Now what I am in doubt about, is whether I should adjust the variables 'age' and 'year' in my dataset as well (add 2 to both), which are used in the strs command to merge with the population mortality data.
I thought, as I'm trying to calculate survival conditional on having already survived 2 years, that these variables perhaps need to be adjusted as well in order for Stata to use the correct mortality data.
For instance, a patient who was 50 years old at diagnosis and who is 52 after 2 years -> does adjustment of 'origin' and 'enter' automatically take into account that the probability of surviving from the second year onwards is now different because this patient is two years older en now lives in a different calender year? I think it doesn't but I am not sure?
Thank you in advance for the response
