Skip to main content
. Author manuscript; available in PMC: 2014 Aug 1.
Published in final edited form as: J Clin Epidemiol. 2013 Aug;66(8 0):S130–S137. doi: 10.1016/j.jclinepi.2013.02.015

Table 2.

Results of 1,000 simulated trials for each of six response rate scenarios. tmax Early and tmax Max N show the proportion of trials in which an arm is identified as the best treatment at an interim analysis and at the end of the trial, respectively. tmax is the total probability of identifying a superior treatment. tmin is the probability of identifying the least effective treatment at the end of the trial. tmax or tmin is the probability of identify either (or both). Values are bolded denoting true positives under tmax and tmin.

Probability Identify
Scenario Response Rates tmax Early tmax Max N tmax tmin tmax or tmin
Null
0.50 – 0.50 – 0.50
0.012 0.001 0.013 0.018 0.031

One Good
0.50 – 0.50 – 0.65
0.879 0.013 0.892 0.033 0.902

Two Good
0.50 – 0.65 – 0.65
0.115 0.003 0.118 0.672 0.763

One Middle One Good
0.50 – 0.575 – 0.65
0.481 0.022 0.503 0.245 0.682

All Bad
0.25 – 0.25 – 0.25
0.016 0.001 0.017 0.030 0.044

All Really Bad
0.10 – 0.10 – 0.10
0.006 0.000 0.006 0.000 0.006