Modern Approaches to Clinical Trials Using SAS: Classical, Adaptive, and Bayesian Methods

2.5 Special Issues

So far, we have demonstrated some statistical implementation of group sequential methods to clinical trials with different primary objectives. The validity of the trial that uses the sequential testing procedures relies heavily on strict adherence to the pre-specified stopping rules. In practice, deviations from the rules during the trial are common, and as a result, may compromise the integrity and interpretability of the final results. The rigidity of adhering to the rule presents challenges to investigators when deviations are ethically necessary and justified. Recent research on clinical trial monitoring introduces the use of conditional power and predictive power within the Bayesian paradigm. As discussed earlier, due ...