?
Действительно ли ожидаемая продолжительность жизни при рождении является наилучшим измерителем уровня смертности населения?
It is generally considered that the most adequate indicator of mortality in any population in a certain calendar period is life expectancy at birth (LE). Yet this indicator has serious drawbacks, and its use leads to a number of unjustified difficulties. The main limitation is that the method for calculating life expectancy is still not unified and most likely never can be. As a result, LE for different countries and periods may not be comparable. The data of the international Human Life Table Database makes it possible to demonstrate that in LE comparisons it is often impossible to use the first decimal digit.
Computation of LE requires the calculation of life tables for a hypothetical cohort corresponding to the calendar period under consideration. When mortality declines rapidly, as has been the case in most countries since the mid-20th century, life tables for a hypothetical cohort say little about the age patterns of birth cohort mortality, and instead form false impressions. Counting how many people in a hypothetical cohort survived or did not survive to a certain age has no relation to any real population.
The transition from age-specific mortality rates to life expectancy is described not by a formula but by a computational procedure. This creates serious difficulties when trying to assess the impact of mortality at certain ages from certain causes of death in certain regions and population groups.
All these drawbacks are not shared by the standardized death rate (SDR), the calculation of which becomes the only option as soon as the standard population is specified. Data of the Human Mortality Database allowed us to show that mortality level estimations based on SDR almost coincide with those based on LE.
SDR is a linear function of age-specific mortality rates; therefore, it is not difficult to calculate the effect of mortality in certain ages, regions or population groups and from different causes of death on SDR.
Non-professionals perceive LE on an intuitive level as the duration of a human life, which is rather a drawback of the indicator. The use of SDR requires more explanation. But in practical research and in professional publications, focusing on SDR will facilitate the work and protect against unwarranted emotions.