I believe when they report this, they're talking about whether there has been an unplanned pregnancy over the course of a year with either typical use or "perfect" use. Someone correct me if I'm wrong.

That's not how that statistic work