As my last three newsletters have emphasised, it is important to calculate age-adjusted measures of mortality when attempting to quantify COVID-19’s lethality in Western countries. This is because Western populations are ageing, meaning that the absolute number of people at risk of dying is increasing each year. (Another, less important reason is that people dying of COVID-19 appear to be slightly older than people dying of other causes, meaning that COVID-19 deaths involve the loss of slightly fewer life years.)
It is commonly asserted, based on the comparison between 2020 and an average of the last five years, that there have been around 62,000 “excess deaths” in England this year (61,764, to be exact). Since the cumulative total of average deaths over the last five years (i.e., the sum of the five-year monthly averages from January to November) is 455,360, this implies that deaths in 2020 have been 13.6% higher than “expected”. However, this comparison does not take into account population ageing or the older average age of people dying from COVID-19.
A more appropriate way of computing excess mortality is to compare age-standardised mortality rates in 2020 to the average age-standardised mortality rates for the last five years. The chart below shows the two comparisons side-by-side: the one based on absolute number of deaths is on the left, and the one based on age-standardised mortality rates is on the right.
Both charts display the same basic pattern. In 2020, mortality was substantially higher than expected in the spring; was slightly lower than expected in the summer; and was somewhat higher than expected in the autumn. However, there are important differences. In the right-hand chart, the differential between observed and expected mortality is much larger during January and February (with observed mortality being lower), and is somewhat smaller during October and November (with observed mortality being higher).
These small differences add up to a non-trivial overall difference in excess mortality. As noted above, the comparison on the left implies that mortality has been 13.6% higher than expected this year. By contrast, the comparison on the right implies that mortality has been only 7.6% higher than expected. The latter figure is 44% less than the former.
Another instructive conclusion that can be drawn from the right-hand chart is that, so far, England’s second wave of COVID-19 has been much less lethal than the first. In April, the observed age-standardised mortality rate was 91% larger than the expected one. But in November, it was only 9% larger. (As before, none of what I have said should be taken to imply that I believe COVID-19 is “just the flu”.)
Image: Vilhelm Melbye, From the coast of Cornwall, 1880
Thanks for reading. If you found this newsletter useful, please share it with your friends. And please consider subscribing if you haven’t done so already.