Table
3: Simulation
results when true hazard ratio = 0.9.
|
Control median |
Sample size |
Percent censored |
Bias |
ASE |
ESD |
Proportion of HR |
||
|
< 0.5 |
< 0.4 |
< 0.3 |
||||||
|
6 |
800 |
7.8 |
0.0019 |
0.0738 |
0.0750 |
0.0000 |
0.0000 |
0.0000 |
|
600 |
7.8 |
-0.0012 |
0.0852 |
0.0851 |
0.0000 |
0.0000 |
0.0000 |
|
|
400 |
7.8 |
-0.0029 |
0.1045 |
0.1057 |
0.0000 |
0.0000 |
0.0000 |
|
|
200 |
7.8 |
0.0017 |
0.1481 |
0.1506 |
0.0000 |
0.0000 |
0.0000 |
|
|
100 |
7.8 |
-0.0014 |
0.2105 |
0.2124 |
0.0040 |
0.0000 |
0.0000 |
|
|
50 |
8.1 |
-0.0113 |
0.3008 |
0.2987 |
0.0290 |
0.0050 |
0.0008 |
|
|
12 |
800 |
27.4 |
-0.0002 |
0.0831 |
0.0818 |
0.0000 |
0.0000 |
0.0000 |
|
600 |
27.4 |
-0.0023 |
0.0960 |
0.0967 |
0.0000 |
0.0000 |
0.0000 |
|
|
400 |
27.4 |
0.0007 |
0.1176 |
0.1181 |
0.0000 |
0.0000 |
0.0000 |
|
|
200 |
27.5 |
-0.0060 |
0.1668 |
0.1638 |
0.0002 |
0.0000 |
0.0000 |
|
|
100 |
27.5 |
0.0013 |
0.2369 |
0.2431 |
0.0096 |
0.0006 |
0.0000 |
|
|
50 |
27.7 |
0.0062 |
0.3381 |
0.3437 |
0.0432 |
0.0102 |
0.0006 |
|
|
24 |
800 |
52.1 |
-0.0007 |
0.1024 |
0.1029 |
0.0000 |
0.0000 |
0.0000 |
|
600 |
52.0 |
0.0010 |
0.1182 |
0.1165 |
0.0000 |
0.0000 |
0.0000 |
|
|
400 |
52.2 |
0.0006 |
0.1451 |
0.1456 |
0.0000 |
0.0000 |
0.0000 |
|
|
200 |
52.1 |
0.0022 |
0.2056 |
0.2063 |
0.0018 |
0.0002 |
0.0000 |
|
|
100 |
52.1 |
-0.0028 |
0.2927 |
0.2927 |
0.0238 |
0.0028 |
0.0000 |
|
|
50 |
52.3 |
-0.0103 |
0.4210 |
0.4344 |
0.0890 |
0.0346 |
0.0080 |
|
Median
HR rounded to two decimal places was 0.90 in all cases except for one case when
it was 0.91 (Control median = 24, Sample size = 200) and one case when it was
0.89 (Control median = 24, Sample size = 50).