Group Sequential Monitoring of Clinical Trials in R (December 13, 2005). Category: Early stopping in clinical trials

It is very expensive to purchase software that performs group sequential monitoring of clinical trials (sometimes called interim analysis). Group sequential monitoring is looking at a trial at selected time points during the study to see if you should stop the study early. There are a couple of functions in R that will do simple calculations, and the price, of course, is free.

The HMisc library created by Frank Harrell has a function ldBands that uses the Lan-DeMets method with flexible alpha spending functions.

The seqmon library created by David A. Schoenfeld has a seqmon function that uses the algorithm by Amitage, McPherson and Rowe. An example they cite shows boundaries using the O'Brien Fleming approach.

I will try to document some of the details of these algorithms and test them against some well known examples when I have time.

Further reading:

• Stats: Controversy over stopping a study early (November 24, 2004)
• Stats: Early stopping in an animal experiment (July 1, 2004)
• Stats: Early Stopping Bibliography (July 1, 2004)
• Stats: Interim analysis

Stats >> WebLog >> Group Sequential Monitoring of Clinical Trials in R (December 13, 2005, Ethics, Early stopping)

This page was last modified on 08/21/07. Send feedback to ssimon at cmh dot edu or click on the email link at the top of the page.

You are welcome to link to this page or other pages on this web site. Individual educational uses are also okay. Please contact me for permission to use these pages in any other way. For more details, please consult my copyright notice.