Fig. 3From: Evaluation of two-stage designs of Phase 2 single-arm trials in glioblastoma: a systematic reviewLoss of power according to the underestimation rate of a null hypothesis (p0) when a single-arm single-stage design is used for a binary endpoint Under 80% of power and 5% of one-sided Type 1 error rate. The x-axis is underestimation rate of a null hypothesis and y-axis is loss of power from the power of 80%Back to article page