It appears that medical authors feel the need to make black and white conclusions when their data almost never allows for such dichotomous statements. This is particularly true when comparing results to similar studies with largely overlapping CIs. Virtually all of the conclusion confusion discussed in this paper can be linked to slavish adherence to an arbitrary threshold for statistical significance. Even if the threshold is reasonable, it still cannot be used to make dichotomous conclusions.