From: Adaptive design methods in clinical trials – a review
Interim sample size per group
Final sample size per group
Probability of identifying the optimal arm
25
50
80.3%
100
91.3%