More data will soon be coming in. But how much do they really matter?
By Sarah Zhang, The Atlantic
DECEMBER 18, 2021
A lot has changed for Omicron in just two weeks. At December’s onset, the variant was barely present in Europe, showing up in 1 to 2 percent of COVID cases. Now it’s accounting for 72 percent of new cases in London, where everybody seems to know somebody with COVID. In the U.K. and Denmark, Omicron case numbers are doubling every other day. The same exponential growth is happening—or will happen—in the United States too, just in time for the holidays.
What seemed likely earlier this month is now quite certain: A big Omicron wave is coming, on top of an already substantial Delta wave. There are still some unknowns about the variant, such as exactly how severe these cases will be. But we know enough about Omicron to understand that the time to act is now. “If we wait until our hospitals look like they’re starting to fill,” says Lauren Ancel Meyers, the director of the UT COVID-19 Modeling Consortium, “then it will be too late.”
The most intriguing unknown—the one in which we might like to place our hopes—is whether Omicron could be milder than Delta. But a milder, more transmissible virus can easily sicken so many people that it ends up increasing hospitalizations and deaths on the whole. Here is some simple math to explain the danger: Suppose we have two viruses, one that is twice as transmissible as the other. (For the record, Omicron is currently three to five times as transmissible as Delta in the U.K.—though that number is likely to fall over time.) And suppose it takes five days between a person’s getting infected and their infecting others. After 30 days, the more transmissible virus is now causing 26, or 64, times as many new cases as the less transmissible one. Exponentials are one hell of a growth hack. If we are banking on the idea that Omicron is more mild to get us through winter, then we had better hope that it’s really, really mild.
Vaccines will lower the proportion of hospitalizations quite a bit in those extra cases, especially because Omicron is infecting lots of vaccinated people. But it’s a long climb down that exponential curve. Moreover, when so many cases pile up all at once, their effects start spilling over into the lives of those who aren’t sick. If Omicron runs through a workplace it may present a temporary inconvenience. But if that workplace is a school, then the school will have to close, disrupting the lives of every child and parent. If that workplace is a hospital, then doctors and nurses are unable to work. This has been an issue in South Africa, where Omicron is already dominant and nearly 20 percent of the health-care staff have COVID. Even if most of these cases are mild, huge numbers of people getting sick all at once will alter everyday reality.
Not every case will be mild, though, and even a small hospitalization rate on top of a huge case number will be a big number. With Delta, “we were already headed for a bad winter,” says Roby Bhattacharyya, an infectious-diseases physician at Massachusetts General Hospital. Now, as my colleague Ed Yong reports, Omicron could push a collapsing health-care system further into disaster. Hospitals are already dealing with the flu and other winter viruses. They’re already canceling elective surgeries. After another year of pandemic burnout, they simply may not have the staff to create the surge capacity that barely got us through last winter. Overtaxed hospitals mean care gets worse for everyone with COVID—but also everyone with a broken hip or a stroke or a baby that urgently needs to be delivered. Omicron’s transmissibility is a danger because high levels of COVID cases come with these second-order consequences that transcend the risk to individuals.
If there are no changes to behavior or policy, this year’s winter wave would peak at about double the hospitalizations of last winter at its worst, and 20 percent more deaths, according to the most pessimistic of projections from Meyers and her team at the University of Texas at Austin. The team gamed out a total of 18 scenarios, based on different guesses for the variant’s inherent transmissibility and immune escape, booster uptake, and the vaccines’ effectiveness against hospitalization and death. The most optimistic projection sees a caseload similar to last winter’s, but hospitalizations and deaths at about half of where they were back then, assuming the vaccines keep up their very high protection against severe illness.
Vaccine protection against severe illness should be more durable than it is against infection, but may still take a hit. Very preliminary data from South Africa’s largest health insurer suggest that two doses of the Pfizer-BioNTech vaccine were 70 percent effective at preventing hospitalization from Omicron infections, down from 93 percent before. If that holds, it’s a “huge decrease,” Meyers says, and one that matches the assumptions of her team’s grimmer—but not grimmest—projections. When they modeled scenarios where vaccine effectiveness against hospitalization dropped by about that much, they saw a difference of tens of thousands of deaths.
The available evidence on Omicron’s inherent severity is likely to be biased in ways that make it appear more promising. First of all, hospitalizations lag infections. “Omicron has been around for three weeks,” Bhattacharyya says. “But so many of those infections have happened in the last one week of those three because of exponential growth.” Second, the first people infected may skew young and are thus more likely to have mild cases regardless of the variant. And third, some of the mildness attributed to the virus may result instead from existing immunity. In South Africa, where doctors are reporting relatively low hospitalizations compared with previous waves, many cases are probably reinfections, given that the majority of people there have had COVID before. The South Africa health-insurer data suggest that Omicron might carry a 29 percent lower risk of hospitalization than the original virus, when adjusted for risk factors including age, sex, vaccination status, and documented prior infection—but many prior infections may be undocumented, which would make the reduction in risk seem bigger than it really is. (A recent analysis of early U.K. cases found “at most, limited changes in severity compared with Delta.”) Meanwhile, Omicron is “going to spread so fast that to wait until we have definitive answers will be to wait too long,” Bhattacharyya says. “If it’s anything but the best-case scenario, and we wait to find out, it’s going to be too late to mitigate the worst.”
Another unknown is where Omicron’s tendency to spread more quickly than Delta comes from. Is it inherently more transmissible, better at evading immunity, or both? And how much of one or the other? “We don’t understand the new equilibrium,” says Lone Simonsen, an epidemiologist at Roskilde University, in Denmark. In the long run, if Omicron’s advantage is largely based on immune escape, then Delta and Omicron could co-circulate like multiple lineages of the flu, says Katia Koelle, an evolutionary virologist at Emory University. Getting sick with one variant might not give you much protection from the other. But if Omicron has advantages in both immune escape and transmission, then Delta could eventually go extinct.
Either way, in the short run, we will have a massive number of Omicron cases on top of a massive number of Delta cases. Together they will infect huge numbers of people, vaccinated or not, and burden an already overburdened health-care system. Boosters, social distancing, rapid testing, and masks can slow down this impact. We will know more about Omicron soon, but we already know enough.
This article originally misstated the University of Texas’s most pessimistic projection for COVID-19 deaths this winter. According to updated information from the university’s modeling team, the projection shows only 20 percent more deaths than last winter at its worst, not twice as many.
The Atlantic’s COVID-19 coverage is supported by grants from the Chan Zuckerberg Initiative and the Robert Wood Johnson Foundation.
MOST POPULAR
How Long Does Omicron Take to Make You Sick?
The new variant seems to be our quickest one yet. That makes it harder to catch with the tests we have.
Omicron Is the Beginning of the End
No matter the severity of the variant, the appetite for shutdowns or other large-scale social interventions simply isn’t there.
Omicron Is Our Past Pandemic Mistakes on Fast-Forward
We’ve been making the same errors for nearly two years now.
