Cost-Effective Designs for Trials with Discrete-Time Survival Endpoints

Abstract

In studies on event occurrence, the timing of events may be measured continuously using thin precise units or discretely using time periods. The design of trials with continuous-time survival endpoints has been studied for years, but very little is known about the design of trials with discrete-time survival endpoints. The optimal designs for trials where observations are recorded at discrete points in time is calculated using the generalized linear model and Weibull distribution. Applying a cost function, the optimal number of subjects and time periods are found in such a way that a sufficient power level is achieved at a minimal cost or the power level is maximized for a fixed budget. Taking the budget for a trial and the cost ratio between recruiting a new subject and obtaining a measurement per subject into account, it is observed that the cost ratio and the shape of the survival function have the greatest influence on the optimal design.