Excess deaths, viral surges, and vaccines: Looking at UK data.

Excess deaths analyses is a useful calibration point for seeing how the pandemic has affected mortality that should be uncontroversial -- it does not depend on death attribution to COVID-19 or vaccine or otherwise, and it does not depend on SAR-CoV-2 PCR testing thresholds or counts or practices. It simply plots ALL deaths in a place from all causes each day over time, and in the same plot includes a "baseline death rate" computed typically from a 5-year average of deaths on that day over the past 5 years pre-pandemic.

The UK has published this in a report they just released, as follows:

Note how nearly all excess deaths are concentrated in the two massive viral surges in the UK -- one spring 2020 and the other Janurary 2021.

Also notice how the deaths quickly returned to baseline as vaccination was taking off in February, and how it remains really low even now that they are in the midst of a Delta surge. Given the long delay between first and second doses in the UK, most people got their second doses after February, and most people <60 had their vaccinations after February.

These data are also presented on ourworldindata.org that has these data for most countries in the world. If you peruse these, you can see the following trends:

1. Excess deaths tend to cluster in spikes that correspond to surges of confirmed cases. This pretty clearly highlights the danger of the pandemic, both in direct viral deaths and deaths from other conditions that occur because hospitals are overwhelmed and preoccupied with COVID-19 cases. This contradicts the point some make suggesting the virus is not really dangerous and pandemic is blown out of proportion.

2. Times with low viral counts but with lockdowns in place, the excess deaths tend to be low. This contradicts the claim some make that excess deaths are driven by lockdowns themselves -- depression, suicide, drug overdose, stress. Surely there are some of these but by the timing it is clear these are not the driving force.

3. In places with high vaccination rates using the highly efficacious vaccines, the death rates have tended to quickly come down to baseline once vaccination is underway. This contradicts the claim that the vaccines are inherently dangerous and leading to more deaths. If there are vaccine-related deaths, this must be more than counteracted than deaths prevented to end up with deaths at or below baseline.

Of course, we can never draw precise causal inference from time series of deaths, but the fact that these avoid death attribution (to vaccine or virus or other) or SARS-CoV-2 testing (what threshold, who is tested, what is reported, ..) that are controversial to some, it is a clean and controversial data point that can give us some insight.