[Comparison of a cohort mortality with the mortality of a reference population. Principle, necessary number of expected deaths and the given power of the test].

This note derives the power of the test comparing the mortality of a cohort to the mortality of a reference population. The principle of the comparison is recalled i.e. computation of the standardised mortality radio (SMR) and test of the null hypothesis: SMR = 1. It is shown how one can compute the power of the test. A numerical example is given.