How to estimate survival difference and 95% CI of the difference at a particular time point

[lih]

How to estimate survival difference and 95% CI of the difference at a particular time? For example, estimate overall survival difference and 95% CI of difference at 5 years between group A and group B. 

[StatDave]

I assume you mean that your response variable is binary (survive or not). If so, then this can be done by fitting an appropriate logistic model and using either one of the approaches shown in this note.

[lih]

It is time to event outcome. We usually report probability of survival at a particular time (Such as progression-free survival at 3 years, x% for group A and y% for group B), but an journal reviewer requests us to report survival difference and 95% of the difference  at a time point.

Thanks!