There has been a lot of discussion in relation to “excess deaths” and COVID-19. This term is not specific to this pandemic and has been used by statisticians globally for many years. This explainer article updates our previous analysis of excess mortality, otherwise known as “excess deaths”.
What are “excess deaths”?
“Excess deaths” can be defined as: the difference between the observed number of deaths and the expected number of deaths. Different methods are used to determine the expected number of deaths; for this article, we used the method of deaths during a similar period in previous years. For example, the Northern Ireland Statistics and Research Agency (NISRA) uses a five-year rolling average.
As NISRA explains, it is not possible to identify if an individual death was an excess death. Instead, excess deaths are a mathematical comparison of all deaths over an interval of time. For example, in a particular area or section of the population, excess deaths can be negative (average this year lower than previous five-year average) or positive.
Excess death figures are a more comparable measure across countries than counting deaths from COVID-19 (or any other causes), due to the different ways countries count their deaths. It is also a useful measure of a wider impact of events, including those who died from a consequence of the pandemic (such as lack of access to health care, etc.).
COVID-19 and excess deaths in Northern Ireland
NISRA publishes weekly death statistics, which include detail on Northern Ireland’s excess deaths over time and by demographics, including age and sex. Some of these statistics are available on an online dashboard.
Compared with the five-year average, there have been 3,424 excess deaths in Northern Ireland since the beginning of the pandemic in Northern Ireland (illustrated in the chart below): 1,929 excess deaths in 2020, and a further 1,495 in 2021 (according to NISRA’s bulletin for week ending 31 December 2021).
Chart 1. Weekly number of deaths registered above or below previous 5-year average* from the start of the pandemic in Northern Ireland
A correlation between COVID-19 registered deaths and excess deaths can be seen in the next chart taken from the same NISRA bulletin.
Chart 2. Weekly number of deaths registered in Northern Ireland
The relationship between COVID-19 and excess deaths has been analysed by NISRA, which suggests that further research is needed to explain the gap between cumulative excess deaths and cumulative COVID-19 related deaths, to understand the direct and indirect effects or impacts of the pandemic.
How does Northern Ireland compare with the rest of the UK?
The age structure of populations differs across jurisdictions. To compensate for this, when comparing different regions, statisticians use “age-standardised mortality rates” to show the number of deaths observed per 100,000 people. As a measure of excess mortality, “relative age-standardised mortality rates” are used, which can be expressed as the percentage difference of one week in a current year from the five-year average of the same week.
The following chart shows these percentages for the four jurisdictions of the UK.
Chart 3: Relative age-standardised mortality rates, all ages, weeks ending 3 January 2020 to 3 September 2021, UK countries
The above chart suggests that since 2021, the relative percentage of excess deaths for those aged under 65 has been greater than those aged 65 and over.
What about Ireland?
The lag between the occurrence of a death and its official registration can affect the accuracy of the reporting of excess deaths, particularly for the most recent time in any given reporting period. In Ireland, death registrations may be completed up to three months after an individual’s death. (The average lag between date of death and registration of death was 63 days in the last quarter of 2019.) Due to the strong correlation between death notices and official mortality statistics, the Central Statistics Office is using the RIP.ie website as an alternative source to calculate more up-to-date estimates of excess death figures, which are also delivered to Eurostat to be included in its statistics.
The following chart suggests that during 2020 and 2021, Ireland’s monthly excess mortality rate has generally remained lower than the EU27 average. Comparisons can also be made with individual countries.
Chart 4. Monthly excess mortality, all ages, January 2020 to September 2021, EU27 average and Ireland
How does the UK compare elsewhere?
The EuroMOMO project collates and publishes statistics on excess deaths across Europe gathered from 30 partners. The following chart shows cumulative excess deaths in the data-providing EuroMOMO partner countries for the past years, for all ages.
Chart 5. Cumulative excess deaths, data-providing EuroMOMO partner countries, all ages, 2017 to 2021
The above chart suggests that for the 29 participating European countries, cumulative excess deaths for 2021 (339,282) is less than for 2020 (397,929).
Our World in Data
Our World in Data presents global statistics on excess mortality, with data from the Human Mortality Database, the World Mortality Dataset, and The Economist. One statistic compares the cumulative number of excess deaths in a selection of countries.
Chart 6. Excess mortality: Cumulative number of deaths from all causes compared to projection based on previous years
The above chart suggests that since January 2020, Russia has accumulated an estimated 1,004,000 excess deaths, followed by other countries, such as the US (909,000), Brazil (703,000), and Mexico (610,000). Within Europe, higher cumulative excess death figures are suggested for Italy (151,000), Poland (145,000), the UK (134,000), Ukraine (122,000), Spain (99,000), Germany (82,000), and France (71,000).
What will be the total excess deaths associated with COVID-19?
Several circumstances will affect the figures for both absolute deaths and relative excess deaths associated with COVID-19:
- how quickly the virus spreads (its natural virality);
- the effectiveness of the population to minimise the spread of the virus (voluntary efforts and mandatory intervention policies);
- the morbidity of the virus, especially for those at risk;
- the quality of health care for those with COVID-19; and
- the quality of health care for those without COVID-19.
It is important to remember the difference between excess deaths (i.e. deaths from all causes compared to a historical average) and confirmed COVID-19 related deaths. Both metrics inform the total impact of the pandemic on deaths, but a final global figure of excess deaths associated with COVID-19 may never be known, only estimated after the crisis has passed.
FactCheckNI is Northern Ireland’s first and only dedicated independent fact-checking service and a verified signatory to the International Fact-Checking Network’s Code of Principles. You can learn more about about FactCheckNI, our personnel, what our article verdicts mean, and how to submit a claim.