Sample size adjustment based on promising interim results and its application in confirmatory clinical trials
*Joshua Chen, Merck and Co. Inc. 

Keywords: sample size adjustment, promising, interim analysis, clinical trials

It is recognized that confirmatory late stage clinical trials should be well designed and carefully planned. However, even designed with high confidence for success, the large confirmatory study may narrowly miss statistical significance due to unexpected reasons and the consequence of failing to meet the success criteria could be devastating. On the other hand, dramatic increase such as tripling or quadrupling an already large Phase 3 study is difficult to justify in the current environment which the biopharmaceutical industry is facing. Sample size adjustment (SSA) based on promising interim results, which has no “interference” to the well-designed study in that it does not require any change to the conventional statistical methods such as test statistics, critical values or estimation, is a particularly useful tool to mitigate the risk of failing in a well-designed study. In this talk, I will discuss SSA based on promising interim results and identify opportunities for application in well-designed confirmatory trials where dramatic sample size increase is not possible. In such applications, SSA is considered an add-on tool to the well-design studies with no interference to the design, statistical analysis or clinical interpretation. Group sequential design and SSA are intended to address two different concerns and it is not meaningful to compare efficiency between these methods. Examples will be used to illustrate the application.