Inadequate Explanation

The analysis by the Robert Koch Institute (RKI) on deaths due to COVID-19 in Germany (until 18 January, 2021) (1) shows an average residual life expectancy for people who have died from the disease of 11.0 and 8.1 years for men and women, respectively. For this purpose, the authors use “the highest empirically derived age-specific residual life expectancy among all the German Federal States” based on the life tables 2016/2018. They do not explicitly correct for previous illnesses.

The residual life expectancy of patients who died of COVID-19 still appears to be relatively high. The median age of persons who died of COVID-19 until 29 December 2020 was 84 years (2). The life tables 2016/2018 reveal a residual life expectancy in the general population of only 6.0 years for men, and of 7.0 years for women.

Earlier situation reports from the RKI show that the median age of the COVID-19 deaths is close to the average age. In the last situation report of which I am aware with both reported parameters (13 September, 2020) (3), the average and median ages of deceased cases were 81 and 82 years, respectively. Thus, an asymmetrical distribution of years of life lost to COVID-19 does not seem to be a crucial reason for the differences in residual life expectancy between the median age and the average of all deaths recorded in the study.

Further, “the highest empirically derived age-specific residual life expectancy among all the German Federal States” seems to explain only a small part of the differences in residual life expectancy. According to the life tables 2017/2019, the difference in residual life expectancy for 80-year-olds between the national average and the federal state with the highest residual life expectancy for men and women is only 0.25 and 0.47 years, respectively.

A plausible explanation to be considered is that a (too) crude division of deaths into 10-year groups may have been used as a basis for calculating the average residual life expectancy (the use of 10-year groups is implied, for example, in eFigure 2 of the article).