Thread by @OpenSAFELY

🚨New blog post🚨 on our recent The Lancet Public Health @TheLancetPH paper describing changes in #COVID-19 related mortality from 2020 to 2022. Read it here:
bennett.ox.ac.uk/blog/2023/05/c…

Let's do a short thread! 🧵

In the paper, we compared #COVID-19 related death rates in the first pandemic wave with death rates in the subsequent waves in 19 million people in England🕵️‍♀️.

The full paper is available here #OpenAccess🔓thelancet.com/journals/lanpu…

The graph below depicts subgroup specific sex- and age-standardised #COVID-19 related death rates and how they changed over time👇.

The first column shows the wave-specific death rate 📉. There is one dot for each wave for each subgroup, with dots shifting to the left indicating a decrease in death rate.

The three columns on the right show the fold-change in death rate of wave 2 (alpha) vs 1 (WT), wave 3 (delta) vs 1 and wave 4 (omicron) vs 1, respectively 📊.

Overall, death rates were lower in wave 2 and wave 3 vs wave 1 (compare the orange dot with the blue dots). In wave 2 (alpha) vs wave 1, the decrease in death rates were broadly evenly distributed across population subgroups (first column of the right) 📊.

In wave 3 (delta) vs wave 1, we observed larger decreases in death rates in groups prioritised for primary SARS-CoV-2 vaccination.

Smaller decreases in COVID-19 death rates were observed in younger age groups, organ transplant recipients, and people with CKD, haematological malignancies, or immunosuppressive conditions.

Death rates in wave 4 (omicron) were higher compared to wave 3 explained by a higher force of infection in wave 4 (compare the dark blue dot with the lighter blue dots) 📈.

In wave 4 vs wave 1, the decrease in death rates was smaller in groups with lower vax coverage (including younger age groups) and conditions associated with impaired vaccine response, including organ transplant recipients and immunosuppressive conditions (last column) 📊.

In conclusion, throughout time, there was a substantive decrease in COVID-19-related mortality 📉. The decrease in mortality rates was not evenly distributed across population subgroups over time 🦠.

We additionally looked at relative risk profiles and how they changed over time. Relative risk levels persisted or worsened in some demographic and clinical groups, particularly those with lower vaccination coverage or an impaired immune response 👉 thelancet.com/journals/lanpu…

The paper is accompanied by a nice editorial available here 👉
thelancet.com/journals/lanpu…