SPSS-FORUM.DE

[Neuro]

I just struggle over a survival analysis and am wondering, whether I may ask you for a short advise?
I wish to fill the following lines for death at 1, 5 and 10 y as in this example, consisting of n, % 95% CI (), p.
In my data set, there are some patients, reaching more than 10 years and some patients not even reaching 1 year because they die, or the follow-up was shorter than 1 year (=some censored data).
a) My question is, how can I get the % 95% CI () and p at these time points? I can think of the following strategy: Do I simply have to calculate different models with resulting different log rank tests and set the longest follow-up times to 365 days, 1850 days resp. 3650 days and set the patients with death after the cut-offs as event=0?
b) How can I get the % 95% CI ()? By taking the SE and calculating with the + or - 1.96x formula?
c) How can I get the 95% CI () of HR, with the same forumula like in b?
d) Is it correct, that I may get the % from the labeled Kaplan Mayer curve itself?

n / % (95% CI) / n / % (95% CI) / p-value for death at 1, 5 and 10 years.

[strukturmarionette]

Hi,

zunächst besser:
http://www.statistik-forum.de/
- und dann möglichst in deutsch

Gruß
S.