Table I.
Robustness simulation study results to compare Bayesian doubly optimal group sequential (BDOGS) with O’Brien–Fleming (OF) procedure when the assumption of proportional hazards is not met in a clinical trial with five analyses. The true median TTF is 12 months for A.
| True hazard |
Method | False pos. |
Power | Sample size, δ =0 (δ=3)
|
|||||
|---|---|---|---|---|---|---|---|---|---|
| Mean | 2.5 per cent | 25 per cent | 50 per cent | 75 per cent | 97.5 per cent | ||||
| Proportional hazards assumption met | |||||||||
| Exp | BDOGS | 0.05 | 0.80 | 625 (651) | 380 (395) | 536 (669) | 668 (716) | 716 (716) | 716 (716) |
| OF | 0.05 | 0.80 | 618 (658) | 512 (449) | 536 (573) | 655 (698) | 687 (716) | 716 (716) | |
| Robustness study-proportional hazards assumption not met | |||||||||
| WI | BDOGS | 0.04 | 0.90 | 371 (389) | 347 (366) | 363 (381) | 371 (389) | 380 (398) | 398 (415) |
| OF | 0.04 | 0.99 | 585 (503) | 478 (375) | 500 (408) | 613 (514) | 638 (532) | 716 (666) | |
| PI | BDOGS | 0.01 | 0.84 | 356 (374) | 332 (353) | 348 (366) | 356 (375) | 364 (384) | 379 (402) |
| OF | 0.05 | 0.99 | 572 (409) | 464 (354) | 485 (370) | 596 (381) | 623 (482) | 716 (519) | |
| VS | BDOGS | 0.05 | 0.95 | 369 (378) | 335 (350) | 350 (367) | 359 (377) | 369 (386) | 492 (408) |
| OF | 0.05 | 0.99 | 578 (461) | 463 (354) | 487 (381) | 599 (490) | 626 (508) | 716 (641) | |
| LN2 | BDOGS | 0.05 | 0.40 | 481 (543) | 381 (397) | 403 (422) | 418 (477) | 573 (716) | 716 (716) |
| OF | 0.05 | 0.38 | 655 (682) | 550 (568) | 580 (624) | 716 (716) | 716 (716) | 716 (716) | |
| WD | BDOGS | 0.04 | 0.27 | 406 (458) | 373 (388) | 391 (408) | 400 (421) | 412 (447) | 553 (716) |
| OF | 0.05 | 0.24 | 638 (675) | 530 (546) | 559 (678) | 686 (716) | 716 (716) | 716 (716) | |
| PD | BDOGS | 0.05 | 0.26 | 473 (433) | 398 (384) | 420 (404) | 433 (415) | 538 (430) | 680 (587) |
| OF | 0.05 | 0.36 | 637 (674) | 545 (539) | 580 (608) | 680 (716) | 680 (716) | 680 (716) | |