12 December 2018 3 8K Report

Hello,

The current common practice helps to avoid all but 1/24 (4.16%) of an adverse event.

We're trying to show that another, easier practice will not have worse outcomes.

Thus, I'm trying to find the sufficient sample size, to achieve statistical power. I tried this calculator:

http://powerandsamplesize.com/Calculators/Compare-2-Proportions/2-Sample-Non-Inferiority-or-Superiority

But I'm not sure about which values should I enter for P-B (proportions for the experiment group) or for the non-inferiority margin.

Since no previous data regarding this question was collected, should we start with a preliminary just to clarify the proportions in the new suggested practice first?

Every help would be appreciated! Thank you

Similar questions and discussions