The mitotic index (MI) is the most commonly used quantitative measure
in anatomic pathology but has been widely criticized for supposed ''ir
reproducibility.'' In this article, the authors show that mitotic figu
re (MF) counts may be described by a Poisson distribution, give confid
ence intervals for the MI, and show that the supposed irreproducibilit
y is largely a consequence of the counting techniques. They also show
how to obtain the MI to a predetermined level of precision and present
an analysis of why mitotic figure counting works, when properly used,
despite the inherently low precision of the estimates that are usuall
y obtained. This is a US government work. There are no restrictions on
its use.