Keywords: Design, Interim Monitoring, Survival, RMST
Log-rank test and Cox proportional hazard model are the gold-standard methods for the design and analysis of randomized clinical trials with a time-to-event endpoint. The hazard ratio and median event-free time are standard summary statistics to quantify treatment benefit. However, in the era of immuno-oncology where delayed effect and long-term effect are prevalent and as cancer evolves to a chronic disease, these standard methods have some limitations and may no longer be optimal or efficient. The restricted mean survival time (RMST) is an appealing alternative method that does not rely on the proportional hazard assumption and can overcome some of the limitations of the standard methods. This presentation focuses on the design, monitoring and analysis of randomized clinical trials with a time-to-event endpoint by using the RMST as the primary analysis method. Examples are provided for illustration purpose.