Hi,
I would like to study the effect of a heritable exposure (gene mutation) on two outcomes (a and b) in a prospective cohort with survival data using cox regression in Stata.
It is known that the exposure under study (gene mutation) causes some condition (c) later in life, which if untreated eventually leads to cancer.
My data contains dates for enrolment into the study which varies depending on when one was tested for the mutation (controls were sampled on the testing date), date of diagnosis of the condition (for those who developed it), dates for the outcomes (a, b), and date for end of followup/study which is the same for all.
I am interested to run sub-analyses to check if the risk of the two incident events/outcomes (a,b) depends on the mutation in itself, or on the phenotype of the condition (c) the gene causes.
To do that, I want to run stratified analyses on those with/without the condition (c) followed up to only 1 year after diagnosis of the condition (c) as opposed to the normal situation of following them up to the end of the study.
I can run regular stratified cox analyses to analyse the entire follow-up period, but I haven’t figured out yet how I can set up the analyses to only analyse up to 1 year of follow-up after diagnosis of the condition. I shall appreciate any ideas and or pseudo-code.
nelly
I would like to study the effect of a heritable exposure (gene mutation) on two outcomes (a and b) in a prospective cohort with survival data using cox regression in Stata.
It is known that the exposure under study (gene mutation) causes some condition (c) later in life, which if untreated eventually leads to cancer.
My data contains dates for enrolment into the study which varies depending on when one was tested for the mutation (controls were sampled on the testing date), date of diagnosis of the condition (for those who developed it), dates for the outcomes (a, b), and date for end of followup/study which is the same for all.
I am interested to run sub-analyses to check if the risk of the two incident events/outcomes (a,b) depends on the mutation in itself, or on the phenotype of the condition (c) the gene causes.
To do that, I want to run stratified analyses on those with/without the condition (c) followed up to only 1 year after diagnosis of the condition (c) as opposed to the normal situation of following them up to the end of the study.
I can run regular stratified cox analyses to analyse the entire follow-up period, but I haven’t figured out yet how I can set up the analyses to only analyse up to 1 year of follow-up after diagnosis of the condition. I shall appreciate any ideas and or pseudo-code.
nelly
