Table 3 Sample size calculation for the primary outcome (2). The table contains specified accrual and dropout rates for 6 periods. This estimate of power was based on 1000 simulations of the study with a starting value for the random number generator of 678.
End of period, time t | 0 | 3 | 6 | 9 | 12 | 15 | 24 |
Accrual (% of total) | 100 | 0 | 0 | 0 | 0 | 0 | 0 |
Group A exponential hazard rate | 0 | 0.0744 | 0.1567 | 0.3054 | 0.0959 | 0.3662 | 0.1018 |
Group B exponential hazard rate | 0 | 0.0171 | 0.018 | 0.1959 | 0.0744 | 0.231 | 0.154 |
Group A SCREW expected % surviving time t | 100 | 80 | 50 | 20 | 15 | 5 | 2 |
Group B CAST expected % surviving time t | 100 | 95 | 90 | 50 | 40 | 20 | 5 |
Common exponential dropout rate | 0 | 0.0034 | 0.0085 | 0.0117 | 0.0107 | 0.0108 | 0.0012 |
Time t | 3 | 6 | 9 | 12 | 15 | 24 | |
Follow up rate (similar in both groups) | 0.99 | 0.95 | 0.90 | 0.88 | 0.85 | 0.75 | |
Median survival | 206.903 | 81.08 | 59.209 | 65.067 | 63.975 | 57.826 | |
Exponential parameter | 0.0034 | 0.0085 | 0.0117 | 0.0107 | 0.0108 | 0.012 | |
Test significance level | 0.05 | ||||||
Power (%) | 92 | ||||||
Number needed Group A: | 86 | ||||||
Number needed Group B: | 30 |