preview

Evaluation Of A Phase II Selection Design

Good Essays
Open Document

The main objective of a Phase II selection design is to correctly select the most promising treatment amongst multiple competing experimental regimens, when it truly exists. How- ever, as it serves as an important pathway to sieve out the best treatment, which could then progress to a simple, de nitive randomized controlled Phase III study, it is important that the selected treatment has sucient evidence of clinically desirable response rate.
In this chapter, we discussed BAR and ELIM which incorporate two main modi cations to conventional selection designs, by sequentially eliminating arms due to futility, inferiority, or a combination of both, before nal selection of the most promising arm with the highest response rate. Firstly, we …show more content…

Using a conventional SWE design, even in the most ideal setting (which is a typical SWE 's assumption and when it performs best in terms of PCS) where all arms are active with one that is 15% better than the rest (0:64; 0:64; 0:64; 0:79), the PCS for the best arm is only
0.69. However, if the best arm is 20% better than the rest (0.64,0.64,0.64,0.84), the PCS is higher at 83%. Understandably if we have a larger sample size, the PCS for all methods will be substantially improved. The example in this paper demonstrated that it is feasible to conduct a trial with limited patients and still be able to obtain useful information in planning a subsequent larger trial, using selection designs.
Based on our simulation results, we demonstrated that both adaptive selection designs which incorporate futility assessment and superiority selection jointly at interims, o er greater eciency (higher PCS for genuinely superior and active arms and terminating early if all arms are futile) as well as ethical bene ts (allocating fewer patients to poor performing arms and more patients to better performing arms) than a single stage design with no interim adaptations. Performance of a design that allows for futility assessment but not superiority selection at the interims is intermediate between the two.
Generally, the performance of BAR and ELIM are fairly comparable. BAR performs better overall in terms of allocating more patients to the superior arm and fewer patients to the

Get Access
Related
Get Access