Power Calculation for Interaction Odds Ratio Between Binary Exposure and Independent Binary Covariate



Parameters


To Calculate
Type I Error Rate
Power (ignored when calculating power)
Total Number of Subjects (ignored when calculating sample size)
Proportion of Treatment in the Population
Proportion of Group 2 in the Population
Reference Event Rate (Proportion of Case)
Ratio of Odds Ratio (Expected Effect Size)


Range of Effect Size for Plotting


Minimal Ratio of Odds Ratio
Maximum Ratio of Odds Ratio


Reference: Eugene Demidenko, Sample size and optimal design for logistic regression with binary interaction. Statistics In Medicine. 2008: 27:36-46