Comparison of the mortality of a cohort with the mortality of a reference population in a prognostic study

It is standard practice in epidemiological studies to compare the observed mortality of a cohort to the mortality expected in a reference population of the same age and sex distributions computed from national statistics. The same methods can be used in a prognostic study where the effects on survival of potential prognostic factors are studied in a cohort of patients with a given disease. Writing the hazard function in the cohort as the product of the hazard function in the reference population and a function of some characteristics of the patients, one can derive simple estimates and statistical tests of the standardised mortality ratios (Breslow et al. 1 ). The study of a cohort of patients with thyroid cancer will provide an example.