Skip to main content

Table 4 Score results \(S(\lambda )\) for the recalculation rules in the APSAC simulation. The values in the table correspond to Fig. 3. The first stage sample size is set to \(n_{1}=90\), the maximum sample size is set to \(n_{max}=270\) with global one-sided significance level \(\alpha = 0.025\), efficacy stopping boundaries were chosen according to Pocock [15] and binding futility stopping bound represented by \(\alpha _{0}=0.5\). OCP: Observed conditional power approach; restrOCP: Restricted observed conditional power approach; Promising: Promising zone approach; OptFunc: Optimization function approach; classicGS: group sequential approach

From: Extension of a conditional performance score for sample size recalculation rules to the setting of binary endpoints

\(\varvec{\lambda }\)

OCP

restrOCP

Promising

OptFunc

classicGS

0

0.524

0.684

0.640

0.630

0.737

0.05

0.491

0.624

0.598

0.589

0.714

0.1

0.457

0.566

0.557

0.549

0.687

0.15

0.432

0.514

0.523

0.517

0.666

0.2

0.411

0.467

0.491

0.488

0.645

0.25

0.646

0.366

0.572

0.593

0.644

0.3

0.577

0.508

0.617

0.628

0.749

0.35

0.544

0.495

0.576

0.591

0.706

0.4

0.527

0.470

0.552

0.570

0.681

0.45

0.542

0.482

0.559

0.579

0.680

0.5

0.536

0.480

0.552

0.572

0.669

0.55

0.539

0.487

0.553

0.575

0.667

0.6

0.553

0.509

0.568

0.586

0.671