Comparing Different Methods of Handling Overrun in Group Sequential Trials

• Activity Number: 194
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Timothy Skalland*+ and Sarah Emerson
• Companies: Department of Statistics, Oregon State University and Oregon State University
• Keywords: Clinical Trials ; Overrun ; Interim Analysis ; Group Sequential Methods

Abstract

A common occurrence in sequential clinical trials is the accrual of data after a stopping region has been reached. The data acquired after a stopping boundary has been passed is called overrun and several methods have been examined for dealing with overrun. Four methods were explored by Sooriyarachchi et al (2003) using an analysis-time ordering of the outcome space, which orders observations based on the observed statistic (ie: sample mean) and the time of the analysis. Another approach is to order the outcome space by the sample mean (maximum likelihood ordering), which orders observations solely on the observed statistic. In this presentation, I use simulations to compare the deletion and combining p-values methods using these two orderings of the outcome space.