Understanding excess mortality: comparing COVID-19’s impact in the UK to other European countries

The coronavirus (COVID-19) pandemic has so far killed at least half a million people worldwide and has led to over 64,000 excess deaths in the UK. And although it is a global pandemic – triggered by the same SARS-CoV-2 virus – the impacts of the pandemic and the responses to it have been anything but the same across countries.

Early in the pandemic countries were at different stages of their outbreaks, making it difficult to make robust comparisons. Initially it looked as if Italy would be the country hardest hit, though this has not turned out to be the case. But as the first wave across Europe is coming to an end, meaningful comparisons have become possible. These comparisons can help us understand and learn from the experiences of different countries, so we are better able to manage a potential second wave or a future pandemic.

In this analysis we use excess deaths (the number of deaths in a given period less the usual number) over the pandemic period. This is a more comparable measure across countries than deaths from COVID-19, because different countries count COVID-19 deaths in different ways. It is also a measure of the total impact of the pandemic, including both COVID-19 deaths and other deaths that may have been a consequence of lockdowns.

The pandemic’s weekly impact across countries

This chart shows weekly excess deaths for selected European countries. In each case the pandemic profile forms a wave with the number of excess deaths rising to a peak and then declining to zero when the first wave of the pandemic has receded. It is clear from the height and breadth of the curves that different countries have had different experiences:

• Spain had the highest peak, with deaths in the peak week being 155% more than usual
• the UK also had a high peak (109%) and has had a slow descent
• until the first week of April, France and the UK seemed on the same trajectory, but this was actually the peak week in France and the curve flattened quickly thereafter while the UK’s deaths kept on growing
• Germany is notable for the flatness of its wave.

The overall impact on mortality across the pandemic period

One metric of the overall impact is the ratio of excess to usual deaths over the pandemic period. This measures how much the relative risk of dying has changed and takes account of population size and, to some extent, factors such as the age composition of the population and the prevalence of morbidity.

Another metric is the number of excess deaths per capita, which measures how much the absolute risk of death has changed.

The table shows these measures for each country. To make the measures comparable we calculate excess deaths and usual deaths over 11-week periods spanning the worst weeks of the pandemic in each country. These 11 weeks may not be the same for each country.

On both measures, Spain is the country most affected. The risk of death increased by 56% over the 11-week pandemic period. This equates to an additional 1,023 deaths per million of the population (or just over one in a thousand). The UK is next most affected, though both measures are very close to Spain’s. The impact of the pandemic in France is less than half that of the UK. In Germany it is less than a tenth.

What impact does adjusting the countries for differences in age structure make?

One challenge that could be directed at the use of excess deaths per capita as a comparator is that it does not take account of a country’s underling age structure.

Two countries could have the same rates of excess death per capita for each age group (for example those aged 80–84) but have different rates of excess death per capita if one had a higher proportion of older people. One way of addressing this is to age-standardise the rates – in effect answering the question: if countries had the same age structure how would they compare?

To help us understand what difference this makes to our comparisons we can apply age adjustment and compare the UK and Spain. The table below shows that if the UK’s excess death rates by age are applied to Spain’s population, Spain would have 54,217 excess deaths, rather than the 48,399 it actually has. This would suggest that Spain fared 11% better than the UK’s and make the UK the most affected of these countries.

In general, countries with older populations will show reduced impact from age-standardising their rates of excess deaths per capita. As the UK has a relatively young population, and because the differences between the UK and other countries’ per capita rates are relatively large, age adjusting other countries’ rates would not change the position of the UK as the most affected country.

Conclusion

As countries emerge from the first wave of the pandemic meaningful comparisons have become possible. We have compared the UK with France, Germany, Spain, Italy and Sweden. Among these countries, the impact has been greatest in Spain and the UK, and has been at least a third greater than in Italy – the next most affected country. Spain fares slightly worse than the UK in terms of the ratio of excess deaths to usual deaths over the 11-week pandemic period, and on excess deaths per capita. But taking account of Spain’s older population would make the UK the hardest hit of these countries. We should look at these not as final comparisons but comparisons about how countries fared during the first wave.

In order to learn the lessons we need to understand not just how countries have differed but why. There are five major areas for investigation:

• The proportion of the population infected before lockdown. This depends on where and how widely the virus was introduced, how quickly it spread and the timing of the lockdown measures.
• The spread of the virus through the population after lockdown. Some of the factors affecting this are lockdown policies and adherence, the availability of PPE, how testing was used, and the protection given to care homes.
• How at risk the population was, for example, the prevalence of risk factors such as obesity.
• Access to and the quality of health care received by patients with COVID-19.
• Health care for non-COVID-19 patients.