In the controversy over when to have mammographs, there's costs.
Ezra Klein has my thought published before I got around to it: if testing at 40 is good, then why not test at 30, and twice a year rather than yearly? Surely the point is that there's a continuum, for any person, and for the community generally. That is, testing identifies cancers which would not be otherwise identified until too late to treat effectively and permits their effective treatment. At some point on the continuum most everyone agrees testing is warranted and at another point it's not. Same sort of thing men face with prostate cancer, though I gather from the first link there might be a more straight-forward link between a positive test and treatment.