The Office for National Statistics publishes weekly data at lower tier local authority for the number of deaths occurring in each week in the 2020 calendar year. From 31 March 2020 these figures also show the number of deaths involving coronavirus (COVID-19), based on any mention of COVID-19 on the death certificate. These are aggregated to upper tier local authority for this analysis.
Whilst mortality rates do vary each year, unless there is a specific event (such as a pandemic), we might expect a similar pattern to the number of deaths we see in an area each year (and there is usually a pattern of deaths across the year with fewer deaths in summer months than in winter). For health planning, we often take an average number of deaths over a longer time period (say five years) to help to smooth out annual variation and assess how different new data is to the recent few years' average.
The data used here comes from the Office for National Statistics. The five year average is calculated by taking the total number of deaths for each week over the period 2015 to 2019 and dividing by five. The average number of deaths for each week are rounded to the nearest whole number. Data are provided for weeks 1 to 52 for each year, although we know that some years have a week 53 due to the day of the week in which the new year falls. As such the ONS five year average data are not available for week 53, so we have repeated the value for week 52 at the end of 2020.
We have included a line on the figure to show the average number of deaths for each week using data from 2015-2019. This is a way of considering if the overall number of weekly deaths (so looking at the height of both COVID-19 and Non COVID-19 deaths bars) exceed what we might expect for that week in a typical year. Where the number of deaths is much higher than the line, this gives us an idea that there are a number of 'excess deaths' beyond those which we could reasonably have expected.
Using this method can be more useful than considering just the 'COVID-19' deaths, because there are debates about what is counted as linked to or caused by the COVID-19 virus.
You can clearly see from the figures below peaks where overall deaths exceeded the expected (five year average) number of deaths. This occurs first at the start of April 2020, then again at the beginning of 2021. For West Sussex, there are also a few weeks later in 2021 where deaths are higher than expected (although peaks are not as prominent as at the start of the pandemic).
Area:Data source: Office for National Statistics
In the first few weeks where deaths exceed the five year average, this included deaths that were not attributed to Covid-19. This may be indicative of Covid-19 deaths not being appropriately attributed although it could also be the result of deaths indirectly related to the pandemic other than the virus itself. This can include an increase in risk behaviours and stress as a result of isolation, and/or delays in seeking care for other conditions or lower capacity in the healthcare system.
Notice that there are some periods where deaths fall shorter than expected (particularly from April 2021, and at the start of 2022).
Data source: Office for National Statistics