The excess mortality rate. A useful concept in cancer epidemiology

Abstract

Death cause registers and cancer incidence registers are often used to elucidate progress (or lack of progress) in the battle against cancer. Trends in the age-adjusted mortality rate of cancer or of specific cancer types may thus mirror the overall effect of anticancer interventions (prevention, early diagnostics, treatment), but are often influenced by changes in the death cause diagnostics or in the coding routines at the registers. Relative survival rate (or its inversion, relative mortality rate) is sometimes used in order to elucidate improvement due to treatment. It is independent of the death cause diagnoses but often seriously influenced by changes in diagnostics of incident cancer; earlier diagnosis and increased detection of non-fatal cases may thus give an improved relative survival rate, quite unrelated to any improvement in the treatment. In the present paper the excess mortality rate is introduced as a measure which can give additional information concerning effects of anticancer interventions. In contrast to age-adjusted mortality rate it is not dependent on death cause diagnoses or coding routines, and in contrast to relative survival it is independent of the rate of non-fatal incident cancer cases.