I have a dataset, where each person is followed from unemployment until employment, where outcome is the time untill employment measurd in months. My upper limit is 24 months. I have persons who is right-censored due to move, die, maternity leave etc, they have time until they ''fall out''.

How do i get the predicted duration for each person after running a stcox regression?

I wanna calculate the difference between the actual and predicted duration for each person, where the predicted duration is estimated by the model where i control for other things (age, sex etc)

Can somebody help me?