Hi,
New here and hope to get help on a problem. I've searched statalist (and the internet) with no success so far.
I'm not able to upload a data example as I'm working on the data on a remote desktop with protected view.
I'm doing a study on non-participation and loss to follow-up in a cohort and I need to estimate the prevalence ratio (PR) of multiple characteristics and at three time points. PR is defined as prevalence at t1/prevalence at baseline, prevalence at t2/prevalence at t1, and prevalence at t3/prevalence at t1, for the three time points respectively.
I'm hoping there is an immediate command that allows me to input two proportions and give me the result in terms of a ratio between the two and a corresponding CI.
I have three variables regarding participation (particpant_t1, particpant_t2, particpant_t3) with two categories coded 0 "non-participant" 1 "participant.
As long as the characteristic I'm looking at is binary, it's pretty straight forward. Say the characteristiv was gender (0 "Woman" 1 "Man"), I would use ci gender participant_t1 to estimate the PR of male participants to non-participants. However, when I estimate the PR for the follow-ups I can't use participant_t2 and participant_t3 as they use non-particpants as reference and I need to use partcipants at t1 as reference. For that particular issue, I've used cii and just input the frequencies manually. It's tideous as there are many variables, but it works.
The main problem is when I need to estimate the PR for a characteristic with multiple categories e.g., education (4 categories). I need to estimate the PR for each of the categories sepeartely at the three time points and I'm not sure how to go about it.
Hoping that there's help to find in this forum.
Thanks in advance!
Kind Regards
Sofie
New here and hope to get help on a problem. I've searched statalist (and the internet) with no success so far.
I'm not able to upload a data example as I'm working on the data on a remote desktop with protected view.
I'm doing a study on non-participation and loss to follow-up in a cohort and I need to estimate the prevalence ratio (PR) of multiple characteristics and at three time points. PR is defined as prevalence at t1/prevalence at baseline, prevalence at t2/prevalence at t1, and prevalence at t3/prevalence at t1, for the three time points respectively.
I'm hoping there is an immediate command that allows me to input two proportions and give me the result in terms of a ratio between the two and a corresponding CI.
I have three variables regarding participation (particpant_t1, particpant_t2, particpant_t3) with two categories coded 0 "non-participant" 1 "participant.
As long as the characteristic I'm looking at is binary, it's pretty straight forward. Say the characteristiv was gender (0 "Woman" 1 "Man"), I would use ci gender participant_t1 to estimate the PR of male participants to non-participants. However, when I estimate the PR for the follow-ups I can't use participant_t2 and participant_t3 as they use non-particpants as reference and I need to use partcipants at t1 as reference. For that particular issue, I've used cii and just input the frequencies manually. It's tideous as there are many variables, but it works.
The main problem is when I need to estimate the PR for a characteristic with multiple categories e.g., education (4 categories). I need to estimate the PR for each of the categories sepeartely at the three time points and I'm not sure how to go about it.
Hoping that there's help to find in this forum.
Thanks in advance!
Kind Regards
Sofie
