Abstract:
|
Recent advances in the topic of group sequential design with a time-to-event endpoint have gone beyond the available traditional methods. One such method is based on the Proportional Time assumption and can be used for sample size calculation when the assumption of proportional hazards or exponentially distributed lifetimes is not satisfied. Time-based measurements are easier to interpret, and past studies that report results as an improvement to longevity versus reduction in risk can be directly applied to bolster the current study design. This novel method can be applied for all distributions that belong to the generalized gamma family. The SAS macro incorporates features unique to time-to-event data, including loss to follow-up, censoring, differing accrual times and patterns, binding or non-binding futility rules with or without skips, and flexible alpha and beta spending functions. This macro includes error-checking for any parameters defined by the user, as well as suggestions for correcting erroneous input. Two real-life applications are demonstrated using this macro.
|