Table 3 Association between proportional hazard assumption results and study characteristics

From: Assessment of proportional hazard assumption in aggregate data: a systematic review on statistical methodology in clinical trials using time-to-event endpoint

 

PH assumption violated

P Fisher’s exact test

No

Yes

Treatments

  Same treatment comparison

20 (40%)

0 (0%)

 

  Different treatments comparison

30 (60%)

12 (100%)

.006

Primary endpoint

  OS

23 (46%)

2 (17%)

 

  PFS/TTP/TTF

27 (54%)

10 (83%)

.101

Superiority trial

  Positive result (superiority demostrated)

19 (41%)

7 (78%)

 

  Negative result (superiority not demostrated)

27 (59%)

2 (22%)

.069

Non-inferiority trial

  Positive result (non-inferiority demonstrated)

4 (100%)

0 (0%)

 

  Negative result (non-inferiority not demonstrated)

0 (0%)

3 (100%)

.029

  1. PH proportional hazard, OS overall survival, PFS progression-free survival, TTP time to progression, TTF time to failure
Back to article page