It's amazing to me that medical journals print results based on 20 control and 20 test subjects.