Early clinical trial termination: Simulation-based design of a robust stopping rule using difference in interventional effect on mortality

Stopping rules are established for early clinical trial termination based on a statistically apparent treatment efficacy as part of a randomised controlled trial (RCT) interim analysis. Well-designed stopping rules are required to prevent participants from ongoing use of a treatment found inferior or even harmful, or potentially to stop a trial early if sufficient evidence of a superior outcome has been obtained. In this study, a stopping rule based on observing an inferior mortality rate of a two-arm RCT is developed. The stopping rule is developed using a Monte Carlo simulation approach, informed by previously collected data. This simulation based method can be adapted for any two-arm randomised controlled trial that monitors binary outcomes. Practical case examples of how the stopping rule can be implemented are presented and the stopping rule is also validated against retrospective data. This simulation based approach in designing a RCT stopping rule can be adjusted to the participating cohort based on the desired study characteristics, and fulfils ethical needs for a statistically justifiable stopping rule.