I need to calculate HR and 95% CI for median survival rate in two different groups. i.e.,
Group 1: median survival=x and CI:(a-b)
Group 1: median survival=y and CI:(c-d)
I appreciate your help
What program are you using?
In R, the survfit function from the survival package will give median survival and corresponding 95% CI.
Then hazard ratios can be computed from a Cox proportional hazards model using cph function.
Hi Joe,
I use Stata, can you please help me with Stata commands for the same analysis?
Unfortunately I'm not that well versed in Stata. The hazard ratio should be simple by doing a Cox model - there are a lot of guides for doing that in Stata. Someone more familiar with Stata will need to pitch in for the median survival confidence interval. Median survival should output with Kaplan Meier estimates.
If you have the data you can run a Cox proportional hazards regression to get the hazard ratio.
If all you have are the median survival and CI for each group, then Stata won't help you get the HR. I think the HR is proportional to the ratio of the times, or maybe the log of the times - it's been a very long time since I had to work that out by hand.
- Badges
- Science topic