Skip to main content

Table 5 Empirical power according to scenarios in setting 2 (non-exponential case) using both measures (gAHR/nHR)

From: Using the geometric average hazard ratio in sample size calculation for time-to-event data with composite endpoints

   Power descriptive
  s Min Q1 Med Q3 Max
Treatment effect       
Any HRj=0.6 1800/1800 0.782/0.208 0.794/0.596 0.797/0.770 0.801/0.884 0.813/0.997
HR1=HR2=0.9 288/336 0.789/0.724 0.797/0.762 0.800/0.774 0.803/0.786 0.810/0.804
Other cases 1080/1080 0.787/0.457 0.796/0.692 0.799/0.772 0.802/0.835 0.812/0.957
Observed proportions       
Any \(p_{j}^{(0)}=0.05\) 1440/1488 0.782/0.208 0.794/0.645 0.797/0.777 0.801/0.870 0.813/0.997
\(p_{1}^{(0)}=p_{2}^{(0)} \ge 0.3\) 648/648 0.788/0.420 0.797/0.717 0.800/0.768 0.802/0.790 0.813/0.949
Other cases 1080/1080 0.783/0.250 0.796/0.609 0.798/0.766 0.801/0.878 0.811/0.991
Correlation       
Weak (ρ=0.1) 1056/1072 0.783/0.252 0.795/0.693 0.799/0.789 0.801/0.859 0.811/0.995
Mild (ρ=0.3) 1056/1072 0.783/0.226 0.795/0.671 0.798/0.770 0.801/0.841 0.813/0.996
Moderate (ρ=0.5) 1056/1072 0.782/0.208 0.795/0.645 0.798/0.753 0.801/0.828 0.812/0.997
Laws of the components       
β1=β2=0.5 396/402 0.786/0.223 0.796/0.676 0.798/0.771 0.801/0.844 0.812/0.995
β1=β2=2 396/402 0.783/0.219 0.796/0.680 0.799/0.770 0.801/0.842 0.811/0.995
β1β2 2376/2412 0.782/0.208 0.795/0.677 0.798/0.772 0.801/0.840 0.813/0.997
Global 3168/3216 0.782/0.208 0.795/0.677 0.798/0.771 0.801/0.841 0.813/0.997
  1. First column (s) is the number of scenarios. Min: minimum; Q1: first quartile; Med: Median; Q3: third quartile; Max: maximum