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
- /
- Help to program bayesian adaptive design sample si...

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
- Permalink
- Email to a Friend
- Report Inappropriate Content

12-17-2014 07:23 PM

I'm trying to program sample size for a bayesian adaptive design, with a binary performance goal of say, AE rate less than 20%. Data will be analyzed after X subjects, and if needed after each subsequent Y subjects. There are three possible outcomes after each stage until success, futility or max sample size of N is reached. So I need to program simulations and probability required for determining study success at each analysis stage.

Note that success decision can be something like: "Termination of enrollment for success if the posterior probability using observed data was sufficient to meet the performance goal (99% probability or greater for 50 and 75 subject interim checks, 97.5% thereafter)"

Continued accrual due to reasonable predictive probability (10% or greater) that performance goal can be met with more subjects.

Futility if predictive probability the performance goal could be met with enrollment of more subjects was less than 10%

Thank you for any insight....

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

Posted in reply to msecic

12-17-2014 08:50 PM

I've done this once...it's not easy or straightforward.

You basically have to determine the math and then program it, but no procs in SAS to accomplish it