Since my mid-December forecast, more person-level data has arrived. Denmark analyzed 2,225 Omicron cases. Canada, about 6,000. The UK analyzed over 200,000. Each found about half the amount of hospitalizations compared to Delta.
Research from Denmark found fewer hospitalizations and went further at examining the chains of transmission. The study found that unvaccinated were twice as likely to transmit Omicron and twice as likely to become infected. The implication for the US is that areas with lower vaccination rates will be hit harder as Omicron spreads.
There has not yet been an effort at a Meta Analysis to combine the studies, so I’ll replay the key results of the largest among the studies, the UK Study.
Overall: UK study suggests Omicron is less severe than Delta with 53% fewer hospitalizations per 100,000 cases (remarkably close to the 57% decrease observed in the aggregate data trends).
With Omicron, Vaccines reduce:
Hospitalization (by 65% with 2 doses, and 81% with three)
The UK study showed that by 20 weeks after the most recent vaccine dose, protection wanes.
The research shows Omicron is proficient at evading immunity, especially that induced via prior infection as opposed to induced by a vaccine booster.
Because there are so many more vaccinated people than unvaccinated people, expect to know more vaccinated people with a breakthrough infection – even though the odds of infection are higher for an unvaccinated person.
Omicron has shifted the ratio of infections from unvaccinated and vaccinated. Before Omicron, unvaccinated were 5x more likely to test positive for COVID and 14x more likely to die of COVID (10x more likely to test positive compared to booster 20x more likely to die compared to a booster). The gap will narrow with Omicron, perhaps to 2x, based on the data from Denmark. Still, even with a 2x difference in infections and 2x difference in transmission reported in Denmark, if this ratio is similar in the US, about 70% of all infections and transmissions will be driven by the unvaccinated.
With skyrocketing cases in the US, we are likely to see hospitals fill up, even though an Omicron infection, overall, is less likely to require hospitalizations.
Keep in mind, the UK has higher vaccination and booster rates than the US, so the US may see less of a decline in hospitalizations. I did not create a vaccination to hospital rate regression because five countries were too few. The five countries to “keep an eye on” have all seen the surge continue, and therefore the 10-country cohort may provide enough data to refine the projection of hospitalization rates as a function of vaccination rates in a country, state or county.
Considering that Red State and Red Counties in the US are half as likely to get the booster, we can expect higher hospitalization and deaths in these Red State regions.
The Denmark study showed vaccination and boosters reduce transmissibility. An unvaccinated person is 41% more likely to transmit and infect another household member than a person with two doses. An unvaccinated person is twice as likely (196%) to transmit and infect another household member compared to a boosted person. Therefore, a model should factor in the population of unvaccinated, vaccinated and boosted to predict where transmissions will grow faster. This suggests Red States and counties, with lower vaccination and booster rates will fare worse.
Coming off the Holidays and back to school/work, I expect cases to continue to climb in the US and peak around mid to late January. I expect this peak to offset the lower rate of hospitalization, causing a net increase in hospitalizations, and likely causing hospitals in the harder hit areas to run into capacity challenges, and therefore worse outcomes for patients. My hope was there might be some policy actions that would avert what I initially labeled “The Denmark Scenario.”
Is it really possible for US hospitals to fill, and the number of dead Americans from COVID to double in a matter of a few months?
Consider where the US is in this curve today relative to the UK and Denmark, and what hospitals might look like if we double again to approach where Denmark is now.
Chart 7: US cases per million people relative to counties with Omicron surges