turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- Estimate Median Survival Time

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

03-23-2012 01:44 PM

I have been given a hazard ratio = 0.64 and a control median of 8 months. How can I use that information to estimate the treatment median survival time at 80% power and show 38% improvement in hazard ratios?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

03-23-2012 03:06 PM

Check out PROC POWER, and look at the syntax for TWOSAMPLESURVIVAL. I am pretty sure that you just need to plug in the values, and exponentiate the hazard rate for the reference group. There is an example 70.6 in the SAS/STAT 9.3 documentation to get started.

Steve Denham

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

03-23-2012 03:11 PM

I don't think that you can do both of those from the little information you have provided. See

http://www.statsdirect.com/help/sample_size/sample_size_for_survival_analysis.htm

for an easy to use survival analysis sample size computation.

Given the ratio and one rate, you can compute the sample size required for 80% power.

You can compute a 38% "improvement in hazard ratios" (though it is not to me well defined) to get a new HR, and then do the computation above.

I believe the SAS Power and Sample Size product can do the can computations.

Doc Muhlbaier

Duke

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

03-23-2012 03:13 PM

I guess I should have specified that we already have a sample size of 200. I hope that clears up a little bit.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

03-26-2012 07:53 AM

I guess I don't understand the question. If you have a known sample size and known power, do you wish to solve for the minimum detectable difference? You have specified what the difference is, so that seems out. More info would help here.

Thanks,

Steve Denham