[Frequency analysis for achieving health goals--II: Analysis of mortality tables for cause of death, expected mortality].

The article gives algorithms to calculate mean ages at death for specific causes, based on life table models. Parameters of interest are death due to a specific cause, dying after the "elimination" of the specific cause, or the transition into a "new" structure of mortality when a specific cause of death vanishes. The setbacks of widely used calculations are discussed in this context. These different approaches are demonstrated by means of data pertaining to accident and cardiovascular mortality of the male population of Berlin (West). The estimated impact on life expectancy varies according to both the mathematical model and the specific cause of death: differences are negligible in mortality due to accidents, whereas the results differ considerably in cardiovascular causes of death. The algorithm suggested allows to constrain the calculations to specific age groups. Topics such as "avoidable death" and "health objectives" have to be aware of these different methods.