Group sequential procedures: Calendar versus information time

Publisher Website
Google Scholar
Abstract
This paper describes the relationship between information time and calendar time in the use of group sequential procedures to monitor interim analyses of clinical trials. Information time is the proportion of maximum subjects or events already observed. Interim analyses often occur at specified calendar times while the definition of classic group sequential procedures occurs in terms of information time. We extend the general group sequential approach of Lan and DeMets1 to include this situation as well.