This week, Scott Atlas resigned from his controversial role as key pandemic situation adviser to President Donald Trump. During his tenure at the White House, Atlas—who has no background in epidemiology, infectious disease, immunology, or virology—loudly resisted mask science and argued that we are faced with a choice between the economy and public health. Atlas also promoted the idea that the solution to the coronavirus pandemic is to do nothing while massive numbers of people get sick so we can achieve herd immunity.
As an epidemiologist whose very specific area of expertise is herd immunity, I was glad when I saw that he had resigned. Perhaps now I can stop trying to explain to people why achieving herd immunity through massive infection is not a solution to the pandemic. But probably not, since Rand Paul and other politicians are still out there promoting these ideas. So one last time, please hear me: “Natural” herd immunity is a horrible, dangerous idea that would not accomplish what its promoters promise.
Getting to herd immunity is the goal—it’s the goal of a vaccination program. But reaching herd immunity by letting everyone get sick is the opposite of a solution; it is the problem. It’s doing nothing amid a crisis and calling it a plan. It’s buying into a fiction in which there are only two options: the health of our economy and pandemic control. In reality, we can’t have one without the other.
Herd immunity is what happens when so many people are immune that a virus cannot find any more susceptible hosts, so the number of new cases dwindles. Just how many people is “so many” is an open question, but most scientists agree that for COVID-19, it’s at least 60 percent—assuming no little pockets of society where far fewer people have immunity.
That means in the United States alone, we would need about 200 million people to be COVID-immune before we start to close in on the goal of herd immunity. There are two ways to become immune: either get sick or get vaccinated. It follows that there are, in theory, two ways to get a population to herd immunity: either let lots of people get sick or have lots of people vaccinated.
As of this writing, we’ve had 13.4 million people who been confirmed to have COVID-19, and we know from randomly sampled serological studies that many more infections have gone undetected. Let’s assume that for every confirmed case, there are four more undetected infections. Based on that reasoning, it works out that the United States has experienced around 67 million COVID-19 cases so far … which means just 130 million more to go!
Herd immunity through natural infection would take a painfully long time. Maybe you have heard proponents of herd immunity talk about how speedy the resolution of this mess could be via this route. I understand why that sounds appealing, because I, too, have been stuck at home for 10 months and would like it to end right now if possible. Unfortunately, it’s not true.
If we kept up the staggering pace that the United States experienced for the month of November—140,000 confirmed cases a day, and many more assumed unconfirmed—it would take another 186 days to reach wonderland. I know November felt like half a year, but can you imagine if the plan was to go on like that for another six months? Before you say that this seems pretty reasonable, be sure to check with your nearest hospital-based health care provider. This pace is killing them. And frankly, six months is optimistic. If there are fewer than four unconfirmed cases for every one confirmed, then it will take longer, perhaps much longer.
The next thing the herd-immunity herd will say is that we should just speed things up. Get this over with!
But how exactly do we plan to do that? There is, thank goodness, no ethics board that would approve a program where millions of people are exposed to a potentially fatal disease with no cure on purpose. And it certainly does not seem like there’s a coordinating body poised to mastermind this effort, especially with Atlas’ departure from the White House.
So I guess we’re going to have to rely on individual people to assess their risks, come to their own conclusions about what’s best, and … cooperate. The plan will only work if 3 out of every 5 Americans decide they should get a disease that can kill them for the sake of the greater good. Who wants to go first?
But wait—we can’t overshoot the goal! If too many people get sick at once, our hospitals will have to ration care, and the death rates from COVID-19 and other causes will skyrocket! So unless you want to tell your sister that she can’t deliver her baby in a hospital because there’s no room left, we’ll need to very finely balance new cases coming too fast and not fast enough. And if there’s one thing I’ve learned this year, it’s that getting Americans to cooperate on a complicated, unpredictable, greater-good project without any sort of coordinating body is a completely unrealistic plan.
Here’s yet another good reason to let this idea go: It means committing to millions of cases, billions of lost productive hours, and hundreds of thousands more deaths—even under a fantasy scenario where we can protect those at highest risk. To reach herd immunity via infection—if cases are distributed evenly through all age groups—we can expect to lose something like 2.7 million lives in the United States, with older people hit hardest. We would lose 1 in 20 people over age 64 and 1 in 3 over the age of 84.*
For this reason, proponents suggest that we should take the people who are most vulnerable and “shield them” while the plan unfolds. This means people who are vulnerable can have no exposure to general society for as long as this takes, while the rest of us run around getting COVID-19. That vulnerable group includes the 50 million people over the age of 65. But does it also include the 10 million Americans who are immunocompromised? What about people with heart disease or diabetes? What about people who are obese? Who decides?
And anyway, where do these vulnerable people go? The “plan” seems to assume that most older adults live in nursing homes, but that’s not the case. And for those who do, are nursing home staff also shielded? Or do they have to move out of their homes so that they aren’t exposed to their own families? What about the millions of children whose primary caregivers are their grandparents? Should those children be separated from their guardians? To where?
What happens when we reach 60 percent recovered in the general population? COVID wouldn’t just go away. We would still face outbreaks, and we’d have a big problem with the groups with no immunity—all those who were “shielded.”
And who will pay for the hundreds of thousands of expected hospitalizations and ICU visits? Are people without insurance going to get COVID-19 and then pay for hospital care out of pocket, or is the state going to launch a new program to cover them?
And who will staff all those fabulous retail and dining establishments to keep our economy bursting at the seams while so many of us are sick?
Even if we could answer all those questions in ways that seemed reasonable, we still have to face the fact that the risk of death for people who get COVID-19 and are under age 65 is low, but it’s not zero. We’d expect between 312,000 and 412,000 deaths in the United States in people under age 65. That’s nearly 1 in 1,000 working-age adults, or the total population of New Orleans.*
And if the totally unrealistic shielding plan fails, or there are too many sick people at once, the number of deaths would skyrocket, suddenly and unstoppably. I very much doubt that our economy would be booming under such conditions.
Proponents argue that the weight of social problems caused by restrictions on activity—including hunger, child abuse, and suicide—outweigh the need to control the pandemic. But the fact is, we do have a pandemic and it’s a big problem. There is no alternative where everything is completely normal right now. In a let-it-burn scenario, what kind of collective trauma would we all face? What will happen to children who lose their parents or grandparents? What will the mental health costs be among our health care providers under such circumstances?
I have one more argument to make. We still do not know how long COVID-19 immunity lasts after we’ve recovered. We’re only a year into this. Cases of reinfection have been rare so far, but they exist. What if immunity lasts 15 months and then people who survived infection are once again susceptible? If that’s the case, we would never reach herd immunity no matter how many people were infected.
We also don’t know anything about the long-term consequences of having had COVID-19, because it’s so new. We do know that as many as 1 in 10 survivors have a very long recovery. Some survivors’ lungs are so badly scarred that they require a lung transplant to survive.
We’ve only seen one place in the world reach such a feverish peak of infection that scientists have debated whether it has reached localized herd immunity: Manaus, Brazil. In this little city in the middle of the Amazon, as many as 1.4 million of a population of 2.2 million were infected with COVID-19 in April. News reports about coffin shortages, mass graves, and hospitals turning patients away are a warning to the rest of us, not an invitation. And in spite of all that suffering, it didn’t even work. COVID-19 cases started surging in Manaus again in September. Sweden—often cited as a herd immunity success story—has now abandoned its version of this plan in the face of failure.
Some localities in the United States have seen a similar pattern: Cases begin to fall after a massive wave of infections, and do-nothing-about-the-pandemic proponents would like to attribute this to the success of herd immunity. However, no place in the United States has yet come close to the threshold required—no, not even South Dakota. Ups and downs in an epidemic are so typical that scientists have been referring to these curves as “waves” for decades. It’s literally in the textbooks. A temporary slowing of the epidemic’s progress is not evidence that herd immunity has been achieved. And by the way, if you’re wondering how that shielding strategy is working out so far: Eight percent of nursing home residents in South Dakota have already died of COVID-19.
Vaccination, unlike infection, introduces our immune system to the virus without risking our lives or a whopping bill for an ICU stay in the process. Vaccination also provides stronger and more lasting immunity, thanks to controlled and repeated dosing. With vaccine approvals just around the corner, I’d like to stop talking about letting the pandemic rage out of control as a strategy. That isn’t a strategy, it’s a failure.
I’m not arguing for “lockdowns.” Asking millions of people to shelter in place until a vaccine is widely distributed is equally unrealistic. I’m saying we should take targeted actions to reduce the spread of COVID-19. We should wear masks and avoid crowded indoor spaces. Restrictions should be stronger when disease rates are high and more lenient when they are low. We should protect our health care workforce—and our economy for that matter—from the disastrous effects of millions of cases of disease at once.
Success is keeping COVID-19 from destroying our country until we can reach herd immunity through a safe and effective vaccination program.
Correction, Dec. 4, 2020: This article originally misstated the proportion of people 65 and over whom we could expect to die under a “natural herd immunity” approach to the pandemic. It is 1 in 20, not 1 in 3. (It is 1 in 3 for people 85 and over.) It also misstated the proportion of people under 65 who would die. It is 1 in 1,000, not 1 in 100.