Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 477 - Methods in Clinical Trials
Type: Contributed
Date/Time: Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
Sponsor: Biometrics Section
Abstract #313035
Title: Sample Size Calculation for Two-Arm Trials for the Non-Proportional Hazards Scenario Using Two Different Weibull Distributions
Author(s): Milind A Phadnis*
Companies: University of Kansas Medical Center
Keywords: censored data; non-proportional hazards; Relative Time; sample size; time-to-event; Weibull

Traditional methods for sample size calculations for two-arm randomized control trials (RCT) with a time-to-event endpoint use the assumption of proportional hazards (PH) or the assumption of exponentially distributed survival times. When the PH assumption is not satisfied, the ‘go-to’ method available in software is to assume a piecewise linear survival curve (Lakatos approach) for both study arms with judiciously chosen cut-points. Motivated by recent advances in literature that encourage using the Weibull distribution, we propose a new method for sample size calculation in the presence of non-proportional hazards by using the concept of Relative Time (RT). This is done by assuming that the survival times for the two study arms come from two different Weibull distributions with different scale and shape parameters. A test statistic based on asymptotic normality is developed under the RT framework. The advantage of this method is that it allows designing an RCT in scenarios where the treatment benefit relative to the control is not constant over the trial period. Simulations are conducted to assess the performance of the proposed method and a real-life application is discussed.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2020 program