We are now in to the third year of dealing with this virus and pandemic, so we’ve got some long term trend data to digest despite the fact our real time surveillance and monitoring is terrible. It is amazing that in this technically advanced country it takes over a month to collect and collate reliable mortality statistics. But that said, here’s a go at a summary of where we stand. Note that the CDC mortality statistics are compiled on a weekly basis, which makes sense as day to day variations and record keeping anomalies like holidays and weekends get smoothed out. Here I’m looking at “excess mortality” – people who died in any given week who probably would not have if it had not been for COVID. Also note that the typical course of the disease is that death occurs 2-3 weeks or so after infection, so the “death” curve lags the “infection” curve by around three weeks.
First, we have to pick a “start date.” For that we can’t use the calendar year because in the US the pandemic really didn’t get going and become widespread until the end of March or early April. So let’s use the 13th week of the year, March 28 2020 as our reference. That means year one is from March 28 2020 to March 26 2021, and year two is March 27 2021 to March 25 2022, and we are now about 17 weeks into year three. Here is a plot of excess deaths due to the pandemic by week. The blue line is the first year, the orange line the second year, and the yellow line is this year so far. I’ll explain the green line further down.
The blue line, 2020, clearly shows the initial wave (with the sharp rise as the disease spread and vulnerable populations fell) as the infection spread across the country. The “summer peak”, which is unusual for a respiratory virus, is also clear, and happened between mid June and early September. Then the big one, the massive winter wave of 2020 between weeks 30 and 52 (early November to early March). Enter 2021 (the orange line). Here we have three factors: first, the various masking and social distancing measures were in full force, and second vaccination was kicking off. By week 1 of year two (first week of April 2021) 25% of Americans were “fully vaccinated” (the initial two dose sequence), and there was no spring wave or early summer peak as seen in 2020. Unfortunately by Summer 2021 restrictions were being lifted, and there was still a significant population that was either not exposed and unvaccinated. So the delayed summer peak hit with almost as many deaths as the initial wave in the spring of 2020. By disease week 40 62% of the population had been vaccinated, so perhaps 80% of the population had some kind of immunity from either having COVID or being vaccinated. So why was there a sharp uptick in the orange line? The third factor: variants like Omicron. While technically not as deadly, the variants spread more easily than COVID Classic. But, as the graph shows, the peak was not as high, and not as many people fell sick and died. So how bad was the first year? We lost 555,242 people who would not have otherwise died. SARS-COV-2 had a whole population mortality rate of 0.168%. The 2017 Flu season had a whole population mortality rate of 0.0187%. So it was far worse than a normal influenza epidemic, almost ten times as bad. However, it wasn’t as bad as the Spanish Flu of 1918, which had a rate of around 0.3%. The two year rate for Spanish Flu was 0.65%, SARS-COV-2 was 0.30%. However, a caution, in that the full story of COVID has yet to be told. Spanish Flu went away, and immunity was long lasting. it doesn’t look like this new virus is going to go so quietly.
Where do we stand? At first things looked pretty good this year. The mortality trend was going down. But the new variant (BA5) is more easily transmitted and appears to be more deadly. Starting at the end of March, the death rate in 2022 became higher than in 2021 – and the trends aren’t good. The yellow line is reported, but as noted above there is a problem. It takes at least a month for all of the mortality data to come in. Even so what we have the line is above 2021, and if we estimate statistically what the line might look like when all the data is in, we get the green line. I sure hope that’s wrong, but in any event it does look like we might be in for another wave.
One last note of interest, the consensus computer models (not the extreme ones that got all the attention in public) at the end of February 2020 estimated the mortality in the US for the first year of the pandemic at 0.1326%. That’s not too bad given what was observed (0.168%).
Caveats: this isn’t a formal scientific study, just a brief overview. I think it’s pretty close, it’s just plotting processed numbers so not terribly complex, but obviously not peer reviewed. The primary reference data for this analysis is the CDC’s Excess Deaths Associated with COVID-19 data set (link to CSV). Vaccination data also from CDC is located here (link). Both are raw data sets, be sure to read and understand the technical notes before you do any number crunching – like hurricane track data, spaghetti must be cooked before consumed 😛 .