|
Abstract:
|
Interim analysis includes any statistical analysis comparing treatment groups with respect to efficacy and/or safety at any time prior to formal completion of a clinical trial. It is desirable because it allows researchers to terminate or change a trial when evidence has emerged before the completion of the trial. However, inappropriate interim analysis may introduce bias into the trial therefore it becomes misleading and it reduces the credibility of trial results. In this roundtable discussion we would like to discuss statistical design and review issues involved in interim analysis, such as the completeness and integrity of the interim analysis plan, precise performance of analysis for interim data, sample size re-estimation, and unexpected circumstance handling and so on. Several failure examples will be reviewed. Hopefully this discussion will help researchers maximize the benefits of application of interim analysis in clinical trials.
|
