Table 3.
Look | Information fraction | N | Cumulative alpha spent | Cumulative beta spent | Z-efficacy/harm | Z-futility | Boundary crossing probabilities under H1 | |
---|---|---|---|---|---|---|---|---|
Efficacy | Futility | |||||||
1 | 0.2 | 452 | 0.001 | 0.008 | ±3.252 | ±0.031 | 0.042 | 0.008 |
2 | 0.4 | 904 | 0.004 | 0.019 | ±2.986 | ±0.152 | 0.17 | 0.011 |
3 | 0.6 | 1355 | 0.009 | 0.036 | ±2.692 | ±0.631 | 0.281 | 0.017 |
4 | 0.8 | 1807 | 0.022 | 0.062 | ±2.374 | ±1.344 | 0.263 | 0.026 |
5 | 1.0 | 2259 | 0.05 | 0.1 | ±2.025 | ±2.025 | 0.143 | 0.038 |
Values were calculated using power = 0.90, alpha = 0.05, gamma spending function − 4 for the alpha and − 2 for the beta, expected incidence of postoperative pulmonary complications of 23% and 17.25% in the lower and higher positive end-expiratory pressure groups, respectively. Number of patients (N) is shown without correcting for dropouts. Look, interim analysis; H1, hypothesis 1 (group difference exists)