March 10 was just a few days ago, but it already feels very far away — before the NBA season was suspended, Tom Hanks tested positive, and a national state of emergency was declared. On that date, Think Global Health, which is a project of the Council on Foreign Relations, produced a report that includes a data table that tells us how many people in the U.S. might ultimately die from COVID-19 under a variety of different assumptions. In the top-left corner, the table shows a scenario where 0.1 percent of people in the U.S. contract the virus and 0.1 percent of those die from it, leading to a bit more than 300 deaths. That’s the best corner of the table. We like that corner. What we don’t like is the bottom-right corner of the table, which contemplates 50 percent of the American population contracting the virus and one percent of those dying. If we end up in that corner, about 1.6 million of us will die.
We could land anywhere in the table. Beyond that, unfortunately, I can’t offer much more specific guidance — in fact, depending on which experts you ask, we could land outside the table, too. But our knowledge about our lack of a knowledge is a kind of knowledge. The coronavirus endgame depends on a series of unknowns. We keep getting more data, but we still don’t know how inherently infectious or deadly this virus is. We don’t know how well we as Americans will respond and are responding to those risks of infection and death: how effectively we will reduce its spread, ensure that our hospitals are capable of handling a flood of sick patients, and heal those who are in the hospital. The early public response was abysmal, but we don’t know how quickly, or dramatically, that is changing. And, in large part because of the appalling failure of the Centers for Disease Control and Prevention to ensure an adequate supply of usable testing kits, we don’t know how bad the virus outbreak in the U.S. is already. However bad it is, things will get worse.
So let’s talk about what we know about some of those known unknowns.
Roughly, there are two ways to address an epidemic that threatens to spiral out of control and infect much of the world. Plan A is to stop it by imposing public-health measures that reduce the rate of transmission, such that each person who gets infected ends up infecting less than one person on average, and the epidemic begins to die out. Dr. Cyrus Shahpar, one of the creators of the table that tells us so much about what we don’t know, told me the more benign outcomes proposed in its leftward columns are based on the apparent success China has had in doing that: Outside the Wuhan area, infections so far have made up less than 0.1 percent of the population. In the Wuhan area, the disease has hit approximately 0.5 percent of the population. In both areas, new infections are declining, not trending toward half the population.
But there are good questions about whether this success can be replicated in other parts of the world and even whether it can be maintained in China itself. China’s measures to slow the spread of the virus have been draconian, and the country is seeking to ease up on them and return to normal. But since most of the population has never been infected, most people remain “naïve” to the virus — they don’t have antibodies and are vulnerable to a resurgence of the epidemic, at least until there are effective antiviral treatments or a vaccine available. And over time, practices that shut down large fractions of the economy and society are likely to prove less sustainable.
It is possible, as countries ramp up their testing and surveillance capacity, we will increase our capability at managing the outbreak by identifying and isolating individuals with the infection, allowing us to reduce our reliance on population-level containment measures — this has been key to South Korea’s success at slowing the growth of new infections with less-severe restrictions on social activity than we are seeing in China or Italy. On Friday, the president held a press conference to tout expanded testing but dodged a question about when those tests would actually become available. And even at South Korean levels of social disruption, epidemic containment would require a more intense and prolonged change to the American way of life than a lot of people and even policy-makers have yet to recognize.
The even less palatable option is to try to let it run its course in the most orderly manner possible. After a large fraction of the population is infected — 40 to 70 percent are the numbers I tend to hear thrown around — many people will have developed antibodies, making it increasingly difficult for the virus to find vulnerable hosts to infect and causing the epidemic to ebb naturally. The big problem with this outcome is that it would entail a very large number of deaths even if the infection fatality rate proves to be on the low end of the estimates.
And then there is the Italy problem. There, as in the U.S. and everywhere else, hospitals have limited capacity to treat patients with severe lung illnesses. If the health-care system gets overwhelmed with an enormous number of COVID-19 cases requiring intensive care, the quality of care will deteriorate and a larger fraction of patients will die than would die in a well-functioning health-care system. This is why people keep talking about the need to “flatten the curve”: Merely slowing the growth of the epidemic, even if it doesn’t reduce the ultimate number of cases, would lessen the pressure on the medical system is considerably. But as I have reported this story, I have gotten an increasingly sinking feeling about the “flatten the curve” discourse for reasons that have to do with the other variable in the table: the fatality rate per infection.
U.S. hospitals contain about 900,000 hospital beds, of which fewer than 100,000 are for critical care. Of course, most of those beds are already in use for people with other medical problems — U.S. hospitals admit approximately 36 million patients annually. If 165 million Americans ultimately contract the novel coronavirus, how much would we need to “flatten the curve” to fit all those who need hospital care into properly equipped beds to avoid a situation like in Italy, where patients are being denied treatment for lack of available resources?
Justin Lessler, an associate professor of epidemiology at the Johns Hopkins Bloomberg School of Public Health, told me he believes that flattening the curve will only work to manage hospital capacity if the fatality rate from COVID-19 is on the low end of the range of estimates. “I was discussing this with a colleague, and I had been operating under the idea that maybe the infection fatality ratio was around one in a thousand,” Lessler told me. “And under that, things are very, very bad, but you can envision a world where we spread out the epidemic and slow it down and never overrun our ICU capacity.
“My colleague was saying he thought the evidence was closer to one percent. If that’s the case, that’s ten times as many, and then it’s hard to envision a scenario where health systems are not overrun. ”
Available critical-care capacity is not a totally fixed figure. The Johns Hopkins Hospital, for example, has the capability to provide critical care in non-ICU areas. As feasible and medically appropriate, it will free up ICU space by moving patients to standard floors. The hospital will delay elective surgeries, reducing demand for the surgical ICU and allowing for redirection of resources.
Still, there are limits, and the uncertain timeline is also a problem: We don’t know how long the epidemic will place additional strain on hospitals and the extent to which that time could outrun the sustainability of some of these measures. As Dr. Robert Scott Stephens, an ICU physician at Johns Hopkins, noted to me, elective surgeries eventually cease to be elective.
If the lethality of this virus turns out to be on the higher side of estimates, Lessler says the only public-health response that seems likely to keep any country’s hospitals from being overwhelmed is one that is “very, very, very, very intensive,” using whatever infection-control measures are necessary to keep the rate of new infections per new infection below one. Countries that take early and effective action, like South Korea, may be able to achieve that with more moderate social-distancing measures, while those that wait too long or fail to take the right steps may find themselves in the same position as Italy, with a complete lockdown as the only viable option to get a handle on the spread of the disease.
What position are we in, I asked? Is it still early here?
“It’s earlier than it will be,” he said. We had this conversation on March 11.
This public-health crisis will also produce an economic crisis of some scale. The stock market has fallen more than 20 percent from its record highs in February. Disruptions caused by efforts to slow the spread of the virus are having particular impacts on the transportation, hospitality, and energy sectors. United Airlines announced Tuesday that net bookings for domestic travel — new bookings minus cancellations — had fallen by about 70 percent in recent days. Net bookings to Europe and Asia had fallen 100 percent, meaning as many people were canceling trips as were booking them, and that was before Wednesday, when President Trump announced a policy restricting travel from most of Europe to the U.S. With fewer people expecting to travel, demand for oil has weakened and prices have fallen sharply. A recession in the U.S. has gone from improbable to very likely in a period of weeks.
As for what happens next to the economy, I can’t tell you. As New York’s business columnist, I have devoted a lot of time to thinking and writing about the economic-policy responses this crisis will require, but it wasn’t until March 10 that I realized both I and the economists I routinely check in with had been making an error in thinking about responses to the crisis: We’d been focusing on the light- to medium-bad versions of how this epidemic might go, without giving much thought to the bottom-right corner of that infernal table, in which this epidemic is worse than any other we’ve seen in our lifetimes.
“The more I think about this, the more this reminds me of post-9/11,” said Doug Holtz-Eakin, a former top economic adviser to George W. Bush and John McCain, when I put the worst-case-scenario question to him. “You had to respond to the terrorist attack by attacking terrorism, and so the analogy here is we have a public-health mission that I continue to think is not robust enough. I think they ought to be giving the states a lot of money for test kits and ventilators and beds and lodging and all sorts of stuff — first responders’ protection. But you also then have to essentially harden the economy against the impact of the threat in the future, and in the case of 9/11, that took years.”
As with any crisis, there is also overreaction among the population. But on balance, there is still more underreaction. There is also the problem of delayed reaction. Lessler, the Johns Hopkins University epidemiologist, noted a way in which this epidemic tricks people into panicking when it’s too late. “If people are only going to start taking the actions they should when they start to see a lot of people dying around them, it’s already too late,” he says. When you combine the substantial period from infection to death with exponential growth in infections, the number of deaths you see around you is likely far lower than the number of deaths you are about to see. The people who stand to die within the next 30 days may not even be very sick yet. And when they get very sick, the hospitals may be overwhelmed and ill-prepared to respond. This is the corner Italy backed itself into. We might be headed there, too.
*This article appears in the March 16, 2020, issue of New York Magazine. Subscribe Now!
We’re committed to keeping our readers informed.
We’ve removed our paywall from essential coronavirus news stories. Become a subscriber to support our journalists. Subscribe now.