In: Statistics and Probability
Can someone please explain how POWER works for superiority trials? the trial was found to not be superior than standard treatment.
What does "92% to detect a 22.5% relative risk reduction (RRR)" mean? Is it a good sign in terms of superiority?
In Clinical Trials, Power is equal to the probability of detection of difference of new treatment when it is not similar to the standard treatment. Basically, in Superiority Trials the Alternative geneally is the mesaurement, according to the clinical ordering, is better for new teatment. Power is the probability of detecting that particular superiority.
For example, for the instance you have given. the Null hypothesis may be that Relative Risk reduction is 0% and the alternative is Relative Risk Reduction is greater than 0%. Now, suppose the original relative risk reduction in the population will 22.5% on an average, there are 92% chance that we will be able to detect it. In a more crude way, if we are conducting 100 superiority study where all the true value of Relative Risk reduction is 22.5%, then 92 of the studies we will find the same. And yes, it is good in terms of superiority because you can see that there are 100 studies which have that property and we able to detect that in 92 of those studies. So, yes it is good.