Conditional Power and Enrollment Strategies in a Time-to-Event Study with a Delay in Treatment Effect

• Activity Number: 596
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Emelita de Leon-Wong*+ and Gary Greenfield
• Companies: PPDI and PPDI
• Keywords: conditional power ; time to event analysis ; non-inferiority/superiority ; sample size

Abstract

Conditional power (CP) is the conditional probability that the final study result will be statistically significant, given the data observed thus far and a specific assumption about the future treatment effect in the population. Conditional power is used for: 1) stochastic curtailment (that is, stopping the study prior to planned end for early signals of either futility or benefit) 2) increasing the sample size if the CP is 'promising' but does not reach the unconditional power specified at the design stage with the current sample size. This presentation considers the use of CP in evaluating enrollment strategies for a large time-to-event trial with uncertain timing in the onset of action for the treatment effect, and for deciding whether to stop the study early for either non-inferiority or superiority or continue the study with a possible increase in the sample size. Simulations are used to investigate the impact of the timing of the treatment effect onset and timing of interim analyses on the CP values used for stochastic curtailment and/or sample size increase.