Comparability of cancer mortality data in different regions of Yugoslavia.

The standardized cancer mortality ratio based on well-defined causes of death (MSMR) was analysed. It was shown how the use of MSMR gave wrong results just because differences related to the age structure of compared populations were not eliminated. In this paper, the presented calculation pointed out that, actually, MSMR only standardizes age distribution of well-defined causes of death. The age standardization can be obtained with MSMR only if compared populations have the same age specific rates from all causes of death. At the same time standard population must also have the same age specific rates from all causes of death or have the same or similar ratios of age specific rates from all causes of death as populations under comparison.