Note: Briggs Analysis, based on a conservative 10x ratio of deaths among unvaccinated compared to vaccinated. CDC data shows greater than 10x ratio.
In the US, September has been the most deadly COVID month among the unvaccinated since the beginning of the pandemic. Fortunately, less than half of the US is unvaccinated, and those who are vaccinated have significant protection. I knew September would be bad, but no major forecast predicted September would be this bad.
Specifically, for September, in the US, I forecasted deaths would increase from August’s toll of 27,600 to 44,677 -- a substantial increase driven almost entirely by the unvaccinated. This was much higher than IHME’s forecast of 40,163 incremental deaths, and slightly lower than COVID Forecasting Hub, which estimated 46,714. The actual for the month was 57,469 - more than 15% higher than any forecast (15% is the threshold I use to classify a forecast as “within range”).
How did we all underestimate September’s death toll? Since deaths had increased from 8,563 in July, to 27,600 in August, I expected the message would get through to the unvaccinated to adjust behaviors to protect themselves and their loved ones from COVID-19. That’s what happened in previous waves, and I used prior response patterns to forecast this wave. But the people that are unvaccinated this wave are different than those unvaccinated last year. As it turned out, every major forecaster significantly underestimated the unvaccinated unwillingness to internalize the reality of hospitalizations, deaths and the role of vaccinations, masks and social distancing in preventing hospitalizations and deaths.
If one is looking for silver-linings, here are a few:
1) Vaccines provide significant protection from deaths and hospitalizations. By my calculations, vaccines have averted 759,254 deaths in the US, reducing the negative financial impact of the pandemic by approximately $6.6 Trillion.
2) Among the unvaccinated, my model suggests we may be at peak, or nearing peak in most states. My overall forecast expected the unvaccinated to peak at 100 deaths per million per week by mid October, but we hit that level last week. Considering the even higher levels of deaths per million among unvaccinated in certain states, I expect the signal of hospitalizations and deaths to get through to more of the unvaccinated - or perhaps to the vaccinated will pass policies requiring vaccination or testing.
3) Because of vaccination over and above the infection recovery, it is highly likely that the pandemic will not carry into 2022. Even if reckless behavior continues among the unvaccinated, there are only so many people left to infect. As I present my forecast for October through the end of the year, consider there are seven states past the 85% herd immunity threshold based on my calculation of combined vaccinations and infection-recovery immunity. Higher than average vaccinations propelled 71% (5 of 7) to these states to herd-immunity levels. Several other states are approaching these herd-immunity levels, again, mostly propelled by vaccinations. More employers are requiring vaccinations, and expansion to 5 to 11 years old’s is expected to be approved soon.
On the other hand, there are 17 states with bans on mandates for masks or vaccines (or both), as I reported in last month’s update. Not a single one of these states is above the 85% herd immunity level based on their reported COVID-19 deaths. It is unlikely vaccinations will increase significantly in these states. Deaths and hospitalizations, at great cost to society and to the families, will be the path out of the pandemic for most of these states. My advice, steer clear of these states if you can.
In my work with Brown University, we’ve developed a risk score to help you navigate which states and counties have higher levels of infections and lower levels of cumulative immunity, vaccinations, mask use and testing. A downloadable table is available from Brown University’s www.GlobalEpidemics.org. We hope to add a map UI soon.
Chart 10 shows my forecast peaking in early October and following a declining rate over the next six weeks. My model forecasts that Thanksgiving will result in new infections as it did in 2020 (the blue line in the chart). I would therefore suggest being cautious this year for the Holidays. This is a pandemic of the unvaccinated, though we all suffer the consequences of an enduring pandemic at some level. The yellow line shows the deaths per million among the unvaccinated. The red line is the national total. The blue line is 2020, as a reference.
Chart 10: Forecast deaths per million among vaccinated and overall