Biomedical Engineering Reference
In-Depth Information
TABLE 10.12: Empirical Power (%) at = 5% (two-sided): HR = 0.67,
80% Event Rate, 24 Weeks Median in C
Scenario
Interval
Right-point
Mid-point
Finkelstein's
Sun's
6
94.4
94.4
93.8
93.4
I
8
94.3
94.3
93.6
93.1
12
93.7
93.7
93.9
93.6
6
93.9
93.8
93.5
93.6
II
8
94.1
93.7
93.6
93.3
12
93.3
93.1
92.9
93.2
6
93.9
94.0
94.7
93.9
III
8
94.0
93.9
94.7
93.9
12
93.8
93.5
94.2
93.7
6
89.4
89.4
93.7
93.4
IV
8
84.4
84.4
93.6
93.7
12
74.5
74.5
93.3
93.5
6
97.3
97.3
94.7
94.2
V
8
98.5
98.5
94.8
94.5
12
99.4
99.4
94.2
94.3
Exactlogranktestpower93.9%.
portion exist, even with moderate random protocol noncompliance, slightly
inflated Type I error rates and under-powered tests are observed. Severely
inflated Type I error rates and either under-powered or over-powered tests
are observed when a systematic difference in progression evaluation occurs
between treatment arms.
Search WWH ::
Custom Search