文摘
Sargent and Goldberg proposed a randomized phase II flexible screening design (SG design) which took multiple characteristics of candidate regimens into consideration in selecting a regimen for further phase III testing. In this paper, we extend the SG design by including provisions for an interim analysis and/or a comparison to a historical control. By including a comparison with a historical control, a modified SG design not only identifies a more promising treatment but also assures that the regimen has a clinically meaningful level of efficacy as compared to a historical control. By including an interim analysis, a modified SG design could reduce the number of patients exposed to inferior treatment regimens. When compared to the original SG design, the modified designs increase the sample size moderately, but expand the utility of the flexible screening design substantially.