Summary of evidence-based guideline update: Prevention of stroke in nonvalvular atrial fibrillation: Report of the Guideline Development Subcommittee of the American Academy of Neurology / Author Response

摘要

I read the most recent American Academy of Neurology guidelines regarding the prevention of stroke in nonvalvular atrial fibrillation. In the meta-regression of studies using continuous monitoring techniques, the authors reported a "p value of less than 0.0000." This seems incorrect because a p value cannot be smaller than zero or negative. The value of a "p value" is that it is used to either reject or accept the null hypothesis, which indeed depends upon the a value set forth by the researchers. It is likely that the reported p value does not represent a real zero, but it is small enough to reject the null hypothesis. I could not find thename of the software used to perform the meta-regression and that is important, because the software used either cannot distinguish between the very small p value and 0 or was not designed to provide a specific p value when it is too small. Because a p less than 0.000000001 will not indicate anything more than a p less than 0.001, a "p less than 0.001" will suffice.