[Comparison between several standard populations for age-adjusting]

To analyse the effect the choice of reference population has on different indicators derived from mortality rates age-adjusting.

The mortality rates for different causes of death in Spain from 1971 to 1992 were adjusted by using four reference populations: Spanish populations from 1971 and 1992, the Standard European population and the standard world population. The results obtained with the four populations were compared by applying three indicators: difference between the rates of 1992 and 1971, ratio between the rates of 1992 and 1971 and annual percentage change between 1971 and 1992.

In the majority of the causes of death studied, including the total, ischemic heart disease and almost all of the tumours, the ratio between the rates and the percentage of annual change are similar regardless of the standard population used. However, the difference in rates is very noticeable in relation to the reference population. In the infectious diseases and testis cancer, the opposite occurs: the different in rates is steadfast while the ratio between the rates varies with the standard population. Finally, the mortality due to pneumonia, Parkinson's disease and encephalic cancer show changes all three indicators used.

An analysis of the specific rates by age must be made before proceeding with their adjustment. This analysis will enable us to ascertain whether the adjustment by age is correct and which indicator (difference, ratio or percentage change) will be appropriate for making comparisons.