Whenever you talk about percentage change you have to remember sample size. We are talking about 12 and 18 patients here, RR from 22% to 33% IS nominal. If it were from much larger sample size, I wouldn't call it nominal.