Compare absolute effects with Contrast function

[pavilaalo]

Would it be possible to add a time argument to contrast differences in absolute effects, such as X-time survival probability for survival models or probabilities for logistic regression?
It would useful to provide differences not only in relative terms but also absolute terms.
Thanks!

[harrelfe]

It would be a very nice addition. I think the complexity may come from getting the confidence interval. I may need help with that part.