Table 1.

Impact of Variation in Observed Versus Estimated Control Arm Median on 1) Absolute Difference That Can Be Detected With Designed Study Power and 2) the Power to Detect Target Absolute Improvement

Estimated Control Arm Median, Months	Target Hazard Ratio (HR)	Target Absolute Improvement, Months	Designed Study Power to Detect Target HR	Observed Control Arm Median (Ratio to Original Estimate)	Absolute Difference That Can Be Detected With Designed Study Power, Months	Power to Detect Target Absolute Improvement
15	0.833	3	80	15 (1)	3.0	80
15	0.833	3	80	16.875 (1.125)	3.375	71
15	0.833	3	80	18.75 (1.25)	3.75	63
15	0.833	3	90	15 (1)	3.0	90
15	0.833	3	90	16.875 (1.125)	3.375	83
15	0.833	3	90	18.75 (1.25)	3.75	75
30	0.833	6	80	30 (1)	6.0	80
30	0.833	6	80	33.75 (1.125)	6.75	71
30	0.833	6	80	37.5 (1.25)	7.5	63
30	0.833	6	90	30 (1)	6.0	90
30	0.833	6	90	33.75 (1.125)	6.75	83
30	0.833	6	90	37.5 (1.25)	7.5	75

Abbreviation: HR, hazard ratio.