Table 2.
Metrics for Evaluable Study Endpoints.
| Pobs larger that Pexp (n = 72) | Under-powered Endpoints (n = 19) | Well-powered Endpoints (± 2.5%) (n = 3) | |
|---|---|---|---|
| Expected power (%) | 87.3 ± 7.75 | 86.4 ± 6.6 | 96 ± 4.3 |
| Observed power (%) | 98.2 ± 3.5 | 65.6 ± 19.7 | 96.5 ± 5 |
| ΔPO-E n (%) | |||
| < − 2.5% | 19 (20%) | ||
| − 2.5 to + 2.5% | 3 (3%) | ||
| 2.51–5% | 14 (15%) | ||
| 5.1–10% | 39 (42%) | ||
| 10.1–20% | 16 (17%) | ||
| > 20% | 3 (3%) | ||
| Expected HR (n = 91)* | 0.667 ± 0.07 (n = 69) | 0.69 ± 0.08 | 0.66 ± 0.06 |
| Observed HR | 0.54 ± 0.13 | 0.76 ± 0.6 | 0.71 ± 0.09 |
| Alpha | 0.053 ± 0.06 | 0.038 ± 0.019 | 0.019 ± 0.026 |
| Sample size – experimental arm | 362.8 ± 225 | 414.7 ± 497.5 | 475 ± 145 |
| Median outcome time in control arm (months) | 14.8 ± 24.6 | 30.9 ± 79.4 | 10 ± 2.8 |
*3 studies the expected HR was not available and therefore observed power could not be calculated.