Skip to Main Content

If the U.S. Covid-19 epidemic were a marathon, the country might have made it to Mile 20. It’s been through a lot, and already, there are signs things are getting better. But there are building leg cramps that could make this last push, which isn’t actually all that short, really painful.

The two existing vaccines are reaching more people, and soon, the country will likely have a third, from Johnson & Johnson, that’s just one dose and comes with easier transport and storage requirements. Cases and hospitalizations have fallen precipitously since their peaks last month, and now deaths — which are a lagging indicator — have turned downward as well. That will ease the burden on health systems and offer a reprieve from what had for months been worsening infection and death data.


While the numbers are going in the right direction, they are still at once unimaginably high levels. Even on the best days, more than 1,300 people die of Covid-19 in the U.S., and many more than that die on many days, according to the Covid Tracking Project. The country just logged fewer than 100,000 new confirmed infections in a single day for the first time since early November; some days in January had more than 200,000 cases.

Experts fear the decline might just be temporary. A more transmissible — and, as evidence increasingly suggests, a seemingly deadlier — form of the coronavirus called B.1.1.7 is starting to build up in the U.S., even as overall case numbers come down. It’s not clear whether the variant could cause cases — and thus hospitalizations and deaths — to increase once again, but the specter of it is adding pressure to vaccinate as many people as quickly as possible and to drive down overall cases as low as possible.

“We keep rolling the dice and keep letting the virus and its variants stay in the population at pretty high levels,” said Jason Salemi, an epidemiologist at the University of South Florida, in Tampa.


Salemi, a Tampa native, was speaking Monday morning, the day after the Buccaneers won the Super Bowl in their home city. He knew lots of people were out celebrating, and knew people traveled to the city for the game and were headed back home. Cases and hospitalizations in the state have been declining in recent weeks, but Florida also has the most confirmed cases of B.1.1.7 of any state.

Salemi cautioned that variant counts were greatly influenced by how much genomic surveillance a state conducted, and that Florida is the third most populous state, so it would normally have more cases than most places. Still, “I was so excited that we won [the Super Bowl], but all the while, knowing what could be coming, it’s bittersweet,” he said.

Cases don’t rise or fall for any one reason; infection trends are driven by policies, individual behavior, the virus itself, and its environment. And experts point to a number of human factors for the current decline in infections: We’re likely past whatever bump in transmission the holidays caused. As cases surged at the end of last year, more states and communities imposed tighter restrictions on businesses and gatherings. People who saw their local hospitals getting inundated by patients perhaps started being more cautious on their own.

“In California, I think people were getting the message that the state wasn’t doing well,” said Karin Michels, the chair of epidemiology at UCLA, in a city where the burden placed on hospitals was so intense there were struggles getting enough oxygen for patients. “They were waking up, and I think they got scared. We’ve had our New York moment, so I think people are more careful and more aware than they were before.”

The number of daily Covid-19 tests has also started to drop a bit, in part because local health departments are diverting attention toward vaccine campaigns. But experts say test numbers are not down enough to account for the full plummeting in confirmed cases; plus, other metrics like hospitalizations and deaths are also coming down.

Other potential factors for the decline in cases include the seasonality of the virus and population-level immunity.

We tend to think of cold and flu season as a stretch of five or so months, a period when cold, dry weather enables respiratory viruses to spread more easily and when we’re driven indoors. But depending on the pathogen, that overarching stretch of time is made up of a series of more compact spikes in viral spread. “It’s possible that for certain viruses that they have a shorter period of time” when they’re at peak prevalence, said Matthew Binnicker, the director of clinical virology at the Mayo Clinic.

For the four coronaviruses that cause common colds, for example, their cumulative period of maximum activity may last the whole season, but “if you look at each individual strain, you see that they’re each maybe two, three months,” said epidemiologist Michael Mina of Harvard’s T.H. Chan School of Public Health. Though SARS-CoV-2, the coronavirus that causes Covid-19, has shown it can spread at any point — think back to the summer outbreaks in the South — it’s possible that the country is coming out of its peak period.

Many experts expect that as more people get vaccinated, it will help limit transmission. But so far, about 10 million people in the U.S. have received both doses of their vaccines (the two immunizations authorized so far both require two doses).

However, there have been more than 26 million people who’ve had confirmed Covid-19, a number that suggests the true figure could be, by some estimates, in the 100 million to 125 million range, if not higher. Most of those people likely have some amount of immune protection, so even if herd immunity hasn’t been reached, the virus is still having a harder time finding new people to infect.

“We’re well on our march toward herd immunity thresholds,” Mina said. “And just like anything, it’s a continuum, so we start to feel the benefits of that the closer and closer we get.”

The drop in cases is evident at the local level, too. At the University of Alabama at Birmingham, there are some 175 patients hospitalized with Covid-19, down from the peaks in December and January and back to pre-Thanksgiving numbers, said Rachael Lee, an infectious disease physician. But when the hospital first started seeing Covid-19 in the spring, it had about 30 patients. “We’re still incredibly high,” Lee said.

If cases going down are like flood waters receding, the accelerating build up of B.1.1.7 could be like a congregation of alligators amassing below the surface. The variant could become dominant in the U.S. as soon as next month, and its snowballing comes as leaders are easing restrictions because of improving conditions at the moment. California lifted its localized stay-at-home orders, New York City is opening indoor dining this week, and Iowa has ended its mask mandate and restrictions on gatherings.

But ask experts how B.1.1.7 might reshape the U.S. epidemic and you’ll hear “it depends” or “it’s the wild card,” though there is a sense of a calm before the storm. Having so much population immunity and having vaccines reach more people could blunt its impact — perhaps the variant just slows the declines in cases — but data from other countries indicate severe measures have to be taken to control it and to stem outbreaks. Most places in the U.S. haven’t seemed inclined to impose those types of restrictions.

There are other variants of concern, too, called P.1 and B.1.351. They were first seen in Brazil and South Africa, respectively, but have since been identified in small numbers in the U.S. Studies have shown that mutations in the variants threaten the power of some vaccines, and in clinical trials, some vaccines performed less well against B.1.351 than against other forms of the coronavirus.

Scientists are still trying to sort out what helps P.1 and B.1.351 outcompete other variants. They could be more transmissible like B.1.1.7, but whereas B.1.1.7 has taken off in multiple countries, so far, huge spikes in P.1 and B.1.351 have only been seen in the countries where they each emerged.

What could also be happening, said infectious disease expert Kristian Andersen of Scripps Research Institute, is that the variants, which share some mutations, are just better at reinfecting people who already had a case of Covid-19.

Brazil and South Africa both had major outbreaks earlier in the pandemic, so there should be some decently high level of population immunity in those countries. Lab experiments have shown the variants can partially evade the immune response generated after an infection, so it’s possible the two strains are able to circulate to an extent among people who have already been infected. Other types of the virus might not be able to do that — they’re blocked by the existing immunity — and the result is that P.1 and B.1.351 grow in prevalence while other variants are curtailed.

Even if P.1. and B.1.351 are just 20% to 30% better at reinfecting people than other variants, Andersen said, because spread “is an exponential process, it will really quickly outgrow others.”