ABSTRACT
The term interim analysis refers to statistical analysis conducted during the course of a clinical trial for the purpose of monitoring efficacy and safety. While interim analyses serve a variety of purposes, a primary goal of interim analyses is to determine if the trial should stop before its planned termination time if the superiority of the intervention under study is clearly established, if the ultimate demonstration of a relevant intervention difference has become unlikely (futility), or if unacceptable adverse effects are apparent. Also, as a result of interim analyses, the intervention may be modified or elements of the experimental design, such as subject eligibility criteria, changed. Since all interventions have the potential for causing harm, there is an ethical obligation to the study participants that a trial not continue beyond the point at which the potential risks to the study participants outweigh the potential benefits. On the other hand, if one intervention is conclusively demonstrated to be substantially superior to another intervention, there is an ethical obligation not to continue to expose subjects to an inferior therapy. Finally, if it is clear that the study is unlikely to provide definitive answers to the study question, continued participation in the trial may expose subjects to potential risk with minimal scientific justification. The statistical issues related to early stopping because of unexpected adverse side effects are more complex and sound decision making generally requires a combination of careful statistical analysis and informed clinical judgment. It is difficult, if not impossible, to prespecify stopping criteria that address all possible safety issues that might arise. Consequently, formal stopping rules for safety monitoring are typically specified for a small number of safety outcomes such as all-cause mortality or adverse events of particular concern such as liver abnormalities that may be suspected of being associated with the treatment.
