Skip to main content

Changes in COVID-19-related mortality across key demographic and clinical subgroups

Changes in COVID-19-related mortality across key demographic and clinical subgroups: an observational cohort study using the OpenSAFELY platform on 18 million adults in England

How to cite: Changes in COVID-19-related mortality across key demographic and clinical subgroups: an observational cohort study using the OpenSAFELY platform on 18 million adults in England The OpenSAFELY Collaborative, Linda Nab, Edward P K Parker, Colm D Andrews, William J Hulme, Louis Fisher, Jessica Morley, Amir Mehrkar, Brian MacKenna, Peter Inglesby, Caroline E Morton, Sebastian CJ Bacon, George Hickman, David Evans, Tom Ward, Rebecca M Smith, Simon Davy, Iain Dillingham, Steven Maude, Ben FC Butler-Cole, Thomas O’Dwyer, Catherine L Stables, Lucy Bridges, Christopher Bates, Jonathan Cockburn, John Parry, Frank Hester, Sam Harper, Bang Zheng, Elizabeth J Williamson, Rosalind M Eggo, Stephen JW Evans, Ben Goldacre, Laurie A Tomlinson, Alex J Walker medRxiv 2022.07.30.22278161; doi:

Lay Summary


In the early stages of the COVID-19 pandemic, we learned that certain groups of the population were at greater risk of death due to COVID-19. Older age, belonging to a racially minoritised group, socioeconomic deprivation and learning disability were found to be associated with increased risk of death, as well as several pre-existing medical conditions such as diabetes, blood cancer, kidney disease, heart disease and having an organ transplant. However, as the pandemic has progressed, several strategies have been adopted to reduce rates of death. These include treatment for people who are sick enough to be admitted to hospital for COVID-19 (such as steroids), population-wide vaccination, and early community-based treatment with new drugs for people who are considered to be at high risk (such as sotrovimab).

In this study, we looked at whether risks have changed across different groups of the population over time, and which groups are at the highest risk at the current stage of the pandemic.

Using healthcare records and death registration data across England on the OpenSAFELY platform, we compared the risks of death due to COVID-19 across three time periods, each corresponding to COVID-19 waves:

  • Wave 1 - 23 March 2020 to 30 May 2022
  • Wave 2 - 2 September 2020 to 24 April 2021 and
  • Wave 3 - 28 May 2021 to 14 December 2021

Each of the study periods included over 18 million adults. Our study largely does not include the Omicron wave of the 2021-22 winter.


We measured rates of death for every 1000 person-years. If you followed 1000 people of the same age and sex with the particular condition for a year (while the risk of catching COVID-19 remained the same) this is the number of people we estimate would have died from COVID-19. We looked at the proportion of people who died of COVID-19 out of all of the adults in the study, not just those who had caught Covid.

The groups we considered included age, sex, ethnicity, BMI, geographical region and a number of medical conditions. Overall, for the whole study population, after taking account of age and sex, COVID-19 death rates decreased over time from 4.6 per 1000 person-years in wave 1, to 2.8 in wave 2 and 0.7 in wave 3. Reassuringly, the rates of death consistently fell over each wave in each of the groups we studied. Rates of death fell across waves considerably in people over the age of 80 years, those with brain conditions, learning disabilities and severe mental illness. The amount by which they fell was lower in people with severe obesity, kidney disease, blood cancers and other immunosuppressed conditions.

We also measured the relative risk of death in groups across each wave to allow us to compare the risk of death from COVID-19 for people with a condition compared to people without the condition. In some groups, relative risk remained roughly consistent across the waves (e.g. the risk of death in males compared to females was 1.7 times higher in wave 1, 1.6 times in wave 2 and 1.9 times in wave 3). In other groups, there was a decline: for example, in people aged over 80 years the risk of death was 42 times greater than people aged between 50 and 59 in wave 1, whereas it was 15 times greater in wave 3.

However, in several groups, there was an increase in the relative risk of death. For example, amongst those on dialysis compared to people not on dialysis, the risk of death increased from 8 times greater in wave 1 to 12 times greater in wave 3. In people with a kidney transplant, the relative risk increased from 7 times higher compared to people without a kidney transplant in wave 1, to 26 times in wave 3. Other groups with similar patterns of increased relative risk include people with non-kidney transplants, younger people, those with severe obesity and those with blood cancers and other immunosuppressed conditions.

The marked reduction in COVID-19 death rates across all groups, including in those who are immunosuppressed, is encouraging. This is likely to be due to high uptake of vaccination, which was introduced on 8 December 2020 in the middle of wave 2 as well as improved hospital treatments, with a contribution from immunity from those previously infected. While rates of death fell in all groups, the amount by which they fell was less in those with kidney disease, blood cancers and other immunosuppressed conditions. The importance of these results is that they highlight groups who remain at relatively higher risk of worse outcomes from COVID-19 compared with the rest of the population and whose needs must be carefully considered.