We don’t know how severe Omicron is, but we do know it’s spreading very fast.

A lot is still unknown around Omicron, but a worrying trend has become clear: This variant sure is spreading fast. In South Africa, the U.K., and Denmark—countries with the best variant surveillance and high immunity against COVID—Omicron cases are growing exponentially. The variant has outcompeted the already highly transmissible Delta in South Africa and may soon do the same elsewhere. According to preliminary estimates, every person with Omicron is infecting 3–3.5 others, which is roughly on par with how fast the coronavirus spread when it first went global in early 2020.

In other words, Omicron is spreading in highly immune populations as quickly as the original virus did in populations with no immunity at all. If this holds and is left uncontrolled, a big Omicron wave lies ahead—bigger than we would have expected with Delta. Cases were already surging ahead of winter. The U.S. already had a too-low vaccination rate. And now Omicron threatens to eat away at the immunity we thought we had.

To be clear, this does not mean the pandemic clock has reset to early 2020. Vaccines and previous infections can blunt the virus’s worst effects. Even if protection against infection is eroded, which experts expect, given Omicron’s heavily mutated spike protein, protection against severe disease and death should be more durable. Hospitalizations, rather than cases, might be a better measure of the virus’s impact, as I and others have argued. But if cases balloon dramatically, even a tiny percentage of patients becoming seriously ill can turn into too many hospitalizations all at once. Therein lies the danger possible with Omicron. “That small proportion of severe disease, if it’s multiplied by millions of cases, that will be bad,” says Jeffrey Barrett, the director of the COVID-19 Genomics Initiative at the Wellcome Sanger Institute. “I’m pretty worried.”

This is the simple math we have to keep in mind: A tiny percent of a huge number is still a big number. A largely mild but uncontrolled Omicron wave could cause a lot of pain, hospitalizations, and death across a country.

The ultimate impact of Omicron will depend on how tiny that tiny percent is and how huge that huge number is. We can’t say for sure, but we have some hints. Given the early trends out of South Africa, the U.K., and Denmark, a large Omicron wave is very possible, though not guaranteed. If we wanted to reassure ourselves, we could note that the absolute numbers of Omicron cases detected so far are so small, they may be skewed by chance, and we could be overestimating the variant’s growth by specifically searching for it. But Omicron is consistently increasing in the three countries looking hardest for it and therefore likely increasing quietly everywhere else.

At the same time, Omicron doesn’t appear terribly virulent so far—but this observation comes with even bigger caveats. Doctors in South Africa, where Omicron is already dominant, have not seen as many severe cases as in previous waves. Other countries with small numbers of Omicron haven’t found many very sick patients either. But there are several reasons to believe that the news on severity could turn out less rosy than it currently appears. First of all, it’s early. Infections take weeks to progress to severe infections and eventually to death. Back in 2020, the first COVID case in the U.S. was confirmed on January 20, 2020; the first official COVID death was not reported until February 29. The picture may change with time.

The early severity data are also confounded by who is getting sick. People who catch the virus early in a wave may be disproportionately young and healthy. “They’re probably taking fewer precautions than an elderly person or someone who’s immunocompromised,” says Vineet Menachery, a virologist at the University of Texas Medical Branch. South Africa’s population is itself fairly young, with a median age of 28, compared with the U.S.’s 38.5. And although vaccination rates are low in South Africa, where less than a quarter are fully inoculated, immunity from previous infection is very high, with one estimate suggesting 62 percent. A good number of Omicron cases are likely to be reinfections. Cases in people who are young or have been previously infected or both should be largely mild. If Omicron cases in this population were mostly severe, that would be a catastrophic sign. The fact that they’re not right now is merely a not-bad one.

Scientists are now working furiously to understand Omicron’s effect on vaccinated people. Even if most breakthrough cases continue to be mild in the vaccinated, a small uptick in how many are not mild can still impact hospitalizations by the “tiny percent of a huge number” rule.

Protection against infection after two doses is not looking very good. “Omicron was a huge jump in evolution,” says Jesse Bloom, an evolutionary virologist at the Fred Hutchinson Cancer Research Center, in Seattle. In what seems to be just a few months, the virus has changed as much as Bloom says he and many researchers expected it to change “over the span of four or five years.” In a slew of recent lab studies, the potency of antibodies that can neutralize the virus dropped anywhere from five- to sevenfold against“pseudoviruses” that have been engineered to carry Omicron’s spike mutations to 41-fold in a study with live Omicron viruses, which is the gold standard. (In the Beta and Delta variants, we saw drops of about six- and threefold compared with the original virus, respectively.) A 41-fold drop in neutralizing antibody activity after two doses does not mean a 41-fold drop in vaccine effectiveness. The real-world impact is hard to predict, but the effect is big enough that protection against infection might be quite low, says Florian Krammer, a virologist at Mount Sinai’s Icahn School of Medicine. “I think you’re dealing with a variant that has no problem infecting vaccinated individuals,” he says.

Not all is lost, because protection against severe disease is likely to hold up much better against Omicron. The first glimmers of real-world data will probably come from the U.K., which is closely tracking Omicron’s spread. Protection against severe disease generally tends to be more durable because of how the immune system works. The first-line defenses of neutralizing antibodies might wane, but other, slower parts of the immune system, such as T cells, can still hold against severe disease. A booster can also strengthen the immune response, says Ali Ellebedy, an immunologist at Washington University in St. Louis. The preliminary data so far bear that out: Two doses plus infection or three doses get people to a higher baseline of neutralizing antibodies, which can better withstand the erosion from Omicron.

Currently, however, fewer than half of Americans over 65 have gotten a booster shot, even though they were prioritized because the elderly tend to mount weaker vaccine responses. And 13 percent are still not fully vaccinated. The early glimpses of Omicron severity data cannot tell us how the variant affects an older and unvaccinated group, but everything about our experience so far with COVID suggests that there’s an extreme age skew to risk. Hospitalization trends this winter will likely track with how many older people remain unvaccinated. And the size of this group is another “small percent of a big number” problem: 13 percent of the 54 million Americans over the age of 65 translates to 7 million people at risk for requiring hospitalization if they get COVID. The unvaccinated population remains vulnerable to Omicron, as do immunocompromised people who don’t mount a good response to the vaccine. “Once you have spread, then you start bringing in all those populations that are inherently more susceptible, and that’s a problem,” Ellebedy says.

Omicron is also arriving on the cusp of the holiday season, when Americans are gearing up for holiday parties and travel. “It’s an especially bad time for a new variant,” says Matthew Ferrari, who studies infectious-disease dynamics at Penn State. “People are already going to be hanging out. They already have plans. It’s going to be hard to disrupt those plans.” He points out that other seasonal respiratory illnesses such as the flu, which can also burden hospitals, are rising too. Nearly two years of pandemic have left many hospitals understaffed and backlogged. Health-care workers are quitting in droves. The level of tolerable COVID hospitalizations in a potential Omicron wave depends on the capacity of our health-care system to absorb them, and hospitals are already running with little slack.

This “tiny percent of a huge number” problem has been with us since the very beginning of the pandemic. The coronavirus is much less deadly than other emerging viruses that have rung alarm bells in the past—SARS, MERS, or Ebola—but it is a whole lot more transmissible. Across the population, this still added up to so many severe cases, it overwhelmed our health-care system. COVID patients got worse care, as did anyone unlucky enough to get sick or injured during these big surges. We don’t want to get close to this point again.

But we aren’t in the same position as in early 2020 because we now have the tools to control Omicron. And thanks to the scientists in South Africa who saw the risk of this variant very early, we have time to put them in place. Vaccines will likely keep protecting against severe infections, and a third shot is likely to boost that protection. Manufacturers are working on an Omicron-specific booster. We better understand the virus’s airborne transmission and how to stop it with masks and ventilation. We have antivirals on the horizon. We have rapid tests, though they should be easier to get. We know social distancing has curbed the virus before. Omicron is spreading fast, but we know how to slow it down.