Fig. 3
From: A gated group sequential design for seamless Phase II/III trial with subpopulation selection
The stopping time (panels A-I) and the corresponding power (panels J-L) among 10,000 trials for three simulation settings. Only simulation results of three weight sets (P1: (\(\sqrt{0.5},\sqrt{0.5}\)) for PFS hypotheses and (\(\sqrt{0.7},\sqrt{0.3}\)) for OS hypotheses; P2: (\(\sqrt{0.7},\sqrt{0.3}\)) for both PFS and OS hypotheses; P3: (\(\sqrt{0.8},\sqrt{0.2}\)) for both PFS and OS hypotheses) with the highest power are shown in the figure. Panels A/D/G/J: setting 1; Panels B/E/H/K: setting 2; Panels C/F/I/L: setting 3; Panels A-C: stopping time for P1; Panels D-F: stopping time for P2; Panels G-I: stopping time for P3. The power for each setting under different weight sets is summarized in (panels J-L)