I wonder about whether NNT should even be reported if the findings aren’t statistically significant. Even if the CIs are reported and savvy readers can see that they don’t cross the line of meaningful clinical difference, less thoughtful readers can (and in my experience, often do) just see the NNT and assume that for at least some patients (1 out of X), one of the treatments actually is superior. Thoughts?
Teresa Benson, MA, LP
Clinical Lead, Evidence-Based Medicine
McKesson Health Solutions
Prior Lake, MN 55372