Vaccines Are Still Mostly Blocking Severe Disease

Our original-recipe shots are holding up against new variants. But we may need to improve them, and soon.

a mountain with a coronavirus particle peeking over it
Getty; The Atlantic

For the past year and a half, since the COVID-19 vaccines first became available—even as last summer’s reprieve gave way to Delta’s surge, then Omicron’s; even as the coronavirus continued to rack up mutations that lifted its speed and its stealth; even as millions of vaccinated Americans caught the pathogen and passed it on—there’s been one huge slice of solace to cling to: The shots we have are still doing an excellent job of staving off severe disease and death.

Billions of people around the world have now been dosed at least once, twice, or thrice; the shots have saved hundreds of thousands, if not millions, of lives, in the United States alone—and they probably could have saved hundreds of thousands more, had more people rolled up their sleeves. “We’re so much better off than where we were in 2020, when nobody had any immunity,” says Donna Farber, an immunologist at Columbia University. It feels, in some ways, like gazing down the side of a mountain we’ve been trekking up for a good 30 months: A nice, stubborn buffer of elevation now lies between us and the bottom, the sea-level status of no protection at all. The body’s defenses against severe disease are immunological bedrock—once cemented, they’re quite difficult to erode. Even as the fast-mutating virus pushes down from above, our footing has, for more than a year now, felt solid, and the ground beneath us unlikely to give.

The shots aren’t perfect: They can’t completely block infections or keep the debilitating symptoms of long COVID at bay. Still, against the severest outcomes, “I think vaccination is holding up,” Ali Ellebedy, an immunologist at Washington University in St. Louis, told me. “It provides a lot of comfort, just knowing that layer is there,” says Natalie Dean, a biostatistician at Emory University.

As SARS-CoV-2’s shape-shifting shenanigans continue, though, widening the evolutionary chasm between its current iteration and the version that inspired 2020’s vaccines, our position is starting to feel more precarious. Say our immune defenses weaken, and cause us to slip; say the virus ups the ante again, and delivers a particularly powerful blow. A rapid tumble down to the trailhead—a total immunological reset—still seems very, very unlikely. The further away we stay from that juncture, though, the better off we’ll be. If minimizing severe disease is a summit of sorts, it’s one we have to keep striving for, likely by revaccinating, and hopefully with updated shots. Knowing when to dose up again, and with what, will require keeping close watch on local conditions, trying to anticipate how the virus might shove us, and maintaining our gear in tip-top shape. It’s a long way to the bottom, but backslides are possible.

Stopping severe disease and death is the first goal of any vaccine. But it’s not necessarily the first protective pinnacle the world set its sights on. Back when the vaccines were new and a near-perfect match for the circulating strain, many people felt hopeful that we’d quickly clamber up to some Symptom-Free Vista—maybe even dart up to No-Infection Point.

As the vaccines got further out from their debut, however, it became clear that we weren’t going to be camping at those outcrops long-term. Which is pretty expected: For any immunization to sustainably and reliably keep people safe from all infections is rare. In the months after people get their shots, levels of infection-blocking antibodies naturally drop off, making it easier for pathogens to infiltrate the body and reproduce. At the same time, the virus is only getting better at knocking us down—it strikes a new blow each time it tacks on another mutation that distances it from the version of itself that inspired our shots. That wild card worries experts far more than any immunological stumble. “Virus evolution is always my biggest concern,” Ellebedy told me. Members of the Omicron clan—the most formidable branch of the SARS-CoV-2 family to date—have proved themselves deft at infecting even the multiply vaccinated, slipping around shot-raised antibodies with ease.

Immunity is too multifaceted, too broad, and too flexible for SARS-CoV-2 to shove us all the way down to the mountain’s base; although speedy defenders such as antibodies decline in the short term, other soldiers such as B cells and T cells can stick around for years, even decades, stowing intel on the virus so they can rise up again. These veteran fighters aren’t fast enough to stop a virus from breaching the body’s barriers. But when it does, they can trounce it before the infection gets too severe. They’re also far harder to stump than fickle, fragile antibodies; even weird morphs like Omicron are familiar-looking enough to evoke the ire of most vaccine-trained T cells and an appreciably large fraction of B cells. “That protects us, even if antibodies are lost,” says Hana El Sahly, an infectious-disease physician at Baylor College of Medicine.

Still, the virus’s assaults on our position on the flanks of Protection Peak are getting stronger. We’ve had to dig our heels in far deeper to stay the course. Two doses of mRNA vaccine, for instance, were enough to hold the line against SARS-CoV-2 in the Delta era and before. Omicron’s mutations, though, upped the ante and made the mountain more formidable. The latest estimates provided by the CDC, which run through the first part of winter, suggest that adults who have received a duo of mRNA doses can cut their risk of serious sickness from Delta by at least 80 to 95 percent; the numbers drop into the 50ish to 70ish range with the original iteration of Omicron, or BA.1, subbed in. So experts recommended an equipment upgrade to keep the summit in sight: another dose, which can restore the body’s ability to stave off severe disease from BA.1 at rates of about 75 to 90 percent.

What’s not certain, though, is how long SARS-CoV-2 will continue to rest on its laurels. Faced with growing population immunity, the virus is being forced to repeatedly switch up its appearance. In the span of just a few months, Omicron has already sprouted several new alphanumeric offshoots—BA.2.12.1, BA.4, and BA.5—that can dodge the defenses that even a tussle with their sibling BA.1 leaves behind. And it’s not entirely clear how wild SARS-CoV-2’s costume changes could get. Parts of the virus that scientists once thought were unlikely to change much have since transformed. This coronavirus, like others that have come before it, has shown a remarkable capacity to shape-shift when faced with immunity blockades, says David Martinez, a viral immunologist at the University of North Carolina at Chapel Hill. “There’s a lot of real estate left in spike,” he told me, “for the virus to continually evolve.” Our vaccines, meanwhile, remain modeled on a version of the virus that first infiltrated the population more than two years ago, and that has since disappeared. “Yes, vaccine effectiveness remains really high” against severe disease, even months out, even against Omicron, says Saad Omer, an epidemiologist at Yale University. But that doesn’t mean there isn’t room to relace our boots and attempt to ascend again.

Just how well protected we are right now—how close to the bottom or the top of the peak—isn’t totally clear. “Our ability to capture this is clouded,” Dean told me. So much of the landscape we’ve been meandering upon has shifted in recent months; we, the travelers, have also changed.

To really get a good grip on vaccine performance, Dean said, researchers need to carefully track large groups of people who have gotten different numbers of shots—anywhere from zero to three or more—over long periods of time, carefully tabulating who’s getting infected, sick, hospitalized, or killed. Metrics like these were relatively easy to monitor during the shots’ clinical trials. But the real world is far messier, and gathering data is much more difficult now. That’s especially true in the U.S., which lacks a nationalized health-care system, and has no single, uniform way to record-keep. Americans’ immunization options also keep splintering. Shots are being administered to different people at different intervals, in different combinations, against different variants; successes and failures against the virus are now much harder to tie directly to the potency of the injections themselves. Vaccinated and unvaccinated people have also gotten way more challenging to compare: They belong to very different demographic groups, split apart by when (or if) they became eligible and how vulnerable to the virus they are, as well as the allegiances that might have swayed some of them toward opting into or out of shots.

It’s also getting rougher and rougher to statistically account for how the pandemic, and the response to it, has evolved. Vaccines work better against all outcomes when the bodies they’re protecting aren’t constantly being taxed by heavy, frequent exposures to a pathogen; in recent months, mask mandates have lifted, and crowded indoor gatherings have gone back into full swing. On the flip side, treatments such as Paxlovid have become more available, muddying patterns that could help clue researchers in to what interventions are saving the most lives. Effectiveness estimates can also be obscured by how severity is defined. Even tracking who gets hospitalized can be a pretty coarse metric. Different hospitals use different criteria to admit patients, especially during surges, when capacity gets stretched. And parsing out the severity of an infection isn’t easy in someone who’s battling another ailment, says Westyn Branch-Elliman, an infectious-disease physician at VA Boston Healthcare System and Harvard Medical School. Nor are all serious COVID cases the same: Some people might be discharged after just a couple of days, while others end up on ventilators—nuances that get lost when all cases that meet the bare minimum criteria for “severity” are lumped together. Add to that the complexities of actual infections—which happen more commonly in the unvaccinated and layer on their own patinas of protection—and it’s that much tougher to figure out how well the shots are performing on their own.

Our estimates are also always a few steps behind. The most up-to-date effectiveness numbers in the books still largely reflect how the shots are faring against BA.1, which, thanks to its speedier siblings, has now almost entirely blipped off the American map. If we’re having trouble figuring out where we stand on this mountainous trail, forking paths exacerbate the confusion, along with a compass that gets more challenging to calibrate by the day.

Practically, that all makes assessing when, or whether, to intervene very difficult. A catastrophic, cliff-like plunge in effectiveness—especially one tied to the emergence of a new variant of concern—might prompt a scramble to revamp our vaccine recipes, stat. That doesn’t seem to be what we’re seeing, though, and experts are still working on timing our next steps just right. Scientists are used to strategizing against other viruses, such as the ones that cause seasonal flus: Scientists reformulate and readminister those shots every year, in an attempt to counteract both waning immunity and viral mutations before most people hit a winter surge. By comparison, this coronavirus’s spread is still too haphazard, too unpredictable. If we can barely gauge which mile marker we’re at, it’s hard to know how often we’ll need to update our approach.

It has, at least, become clear that protection can fall off far faster in some vulnerable populations, who may need immune refurbishings more often: older people, immunocompromised people, people with certain chronic health conditions. In the rest of the population, though, the extent of the decline feels far murkier. Branch-Elliman told me that some studies are likely overestimating drops in effectiveness against severe disease: Some aren’t accounting for the immunity building up in the unvaccinated; others are failing to disentangle the true severity of infection in each case. Adjust for them, she said, “and the evidence for waning becomes much weaker.” Still, as the virus continues to change, and people’s most recent doses recede further into the past, “I think we are seeing a little bit” of a performance drop, says Huong McLean, an epidemiologist and a vaccine researcher at the Marshfield Clinic Research Institute. The dip isn’t massive—maybe a few percentage points over several months, in those who are up-to-date on their shots. “But,” Omer told me, “it is there.”

The experts I spoke with couldn’t point to a single threshold at which they’d really start to worry—when the dreaded bottom of Protection Peak might feel just too close. One tentatively offered a dip below 50 percent effectiveness against severe disease as an obvious bad-news-bears benchmark. A couple said they’d start to worry at about 70 percent, while another told me that anything below the range of the 80s would “be clearly problematic” (which, maybe, means now). Martinez, of UNC Chapel Hill, framed it by outcomes: If rates of hospitalization or death among people up to date on their vaccines were to approach “an uncomfortably high rate”—maybe half what experts are seeing among the unvaccinated, he told me—“that’s when you want to patch this.”

The clearest and most immediate intervention available would be administering an additional vaccine dose. And ideally, it would be tailored to better match the circulating strains du jour, which, for now, requires at least some nod to Omicron and its offshoots. Ellebedy suspects that the U.S. may be on track to revaccinate against this virus with a new vaccine recipe as often as each autumn, as we do for the flu—an annual adjustment that may become as necessary as redrawing trail maps to account for shifting mountain terrain. But it will be a behavioral hurdle as much as a technological one. Just a third of Americans are boosted, and uptake on future doses might not fare much better.

None of this means our vaccines have been a bust. “Protection against severe disease and death is what you want out of a vaccine in a public emergency—that’s the most important thing, and that’s what the vaccines have done,” Farber, the Columbia immunologist, told me. And the shots continue to trim down symptoms, transmission, and infection, even if they do not completely stave off those outcomes; the illnesses that do occur among the vaccinated also tend, on average, to be notedly less severe, Branch-Elliman points out. But our current crop of immunizations has its limits; shots won’t be enough to end the pandemic on their own, especially not with uptake lagging, and global vaccine equity still in a disastrous state. “I feel worried, just looking ahead,” Farber told me. “As long as the virus can infect, it’s going to change and adapt.” The virus could get more chances to transform into something more troubling; it may become even tougher to hold our ground against it, should we grow complacent now.

Longer-term solutions, better at tackling infections, transmission, and less-severe disease, may be in the pipeline. Martinez is one of several researchers hoping to cook up a universal vaccine that will teach the body to arm itself against a panoply of coronaviruses at once; others are working on up-the-nose vaccines that could post up defenses in the airway, to head off future infections at the pass. There may even someday be an immunization recipe that better cements the durability of defense, so future generations wouldn’t need so many shots. Those innovations might be years away, if they appear at all. For now, we’ll likely need more than vaccines to stay on top of protection—to reach for the zeniths where not just severe disease is minimized, but infections and transmission can stay low, too. Fortunately, it’s a path we’ve charted before, with tools and footfalls that are by now familiar: masks, ventilation, antivirals, tests. The way back up to a peak isn’t always the way we came down. Sometimes, we just need to blaze a better trail.