Survival Analysis: Person Time error

[Kola Adeniran]

Hello everyone, I have just declared my data to be survival data using the stset command and I observed that the "Total analysis time at risk and under observation" is given as 2.2759e+09.

I have looked at all the dates variables in my data and they are all float type and in %td format. Not sure why I am getting the above. Will appreciate any help as I am quite new to Stata.

[Paul Dickman]

You need to provide more information; ideally a few rows of data and the code you used. We need to see the stset command.

One possibility is that you are using dates and haven't appropriately specified origin and entry. If your time units are in days then 21,914 days is around 60 years. If you haven't specified an entry date then Stata will assume entry at the time origin (1JAN1960) which is consistent with last exit in 2020. That's just pure speculation; we need to see more info to help.

[Kola Adeniran]

You need to provide more information; ideally a few rows of data and the code you used. We need to see the stset command.

One possibility is that you are using dates and haven't appropriately specified origin and entry. If your time units are in days then 21,914 days is around 60 years. If you haven't specified an entry date then Stata will assume entry at the time origin (1JAN1960) which is consistent with last exit in 2020. That's just pure speculation; we need to see more info to help.

Your comment was helpful. I have gone through the survival analysis manual and specified both origin and exit and the total analysis time is now meaningful.