PAREXEL - The Differences Between Traditional and Adaptive Design Trials
It's important to recognize that there are differences between adaptive trails and traditional trials – and thus some different requirements. I like to think about adaptive trials as traditional trials on steroids. These trials utilize a more revolutionary than evolutionary approach compared to traditional trials. Likewise, they also utilize a more iterative as opposed to the traditional linear approach. But the advantages of an adaptive trial come at a cost – the operational infrastructure required to execute an adaptive trial is more complex. These complexities increase when the design calls for unblinding the data. In this case, you need to take special precautions to guard against the inadvertent release of data to unauthorized individuals which could potential invalidate the results of your trial. A key requirement is to modify study elements without undermining the validity and integrity of the study. Validity refers to the requirement to avoid any form of statistical bias in the study and integrity refers to the need to minimize operational bias. So, in addition to very specialized statistical experience, these trials also require an experienced cross-functional and integrated operational approach. Very importantly, these trials also require a technology infrastructure or environment that enables real-time decision making and implementation of these decisions.