This example is aimed at showing that whenever a precision of 5 percent (or better) is required, it is important to check carefully the statistical data. Even data coming from well known websites (in the present case the National Center for Health Statistics) can, for some reason, be wrong.