Americans adopted a mantra early in the coronavirus pandemic that we need to “flatten the curve” shorthand to agree as much as possible on social distance to save lives.
The goal is to slow down the spread of the virus so that hospitals and doctors are not overwhelmed by a wave of patients. The strategy saves lives through mitigation and reduces the damage caused by the virus. It also makes it possible for patients to get the best possible treatment, because there are fewer of them in the hospital at any given time.
However, the approach still assumes that many people will still get sick and suffer – and many people will die.
Mitigation is important, but it’s not suppression – a strategy designed to bring the virus into submission. The United States has not conducted a serious public debate on this distinction. It must be.
The purpose of repression is to reduce the number of new cases and not just use the time it takes to appear. Social distance is part of an oppression strategy. But it is not enough. Oppression requires not only social distance, but also contact tracking and case isolation to keep numbers increasingly low – ultimately allowing society to return to a quasi-normal life, convinced it can destroy new outbreaks.
The US is debating how and when to reopen the economy. But first it must first come up with a response to public health repression. It is time to move beyond ‘smoothing the curve’ to a new mantra: ‘Suppress the virus’.
Mitigation: prevent the hospital system from becoming overwhelmed
In early March, when Seattle and the Bay Area had strict social distance policies but most of the country didn’t, many public health officials spoke enthusiastically about mitigation.
The goal, they say, was not so much to reduce the overall proportion of the U.S. population that became infected, but to slow the rate of infection.
Here’s how Dylan Scott and Eliza Barclay reported it for Vox on March 10:
The biggest uncertainty in the coronavirus outbreak in the United States now is how big it will get and how fast. Nancy Messonnier of the Centers for Disease Control and Prevention told reporters on March 9, “many people in the US will be exposed to this virus at some point this year or next year.”
According to infectious disease epidemiologist Marc Lipsitch it is ‘plausible’ at Harvard 20 to 60 percent of adults becomes infected with Covid-19 disease. So far, 80 percent of cases worldwide have been mild, but if the death rate is around 1 percent (which several experts say it may be), a scenario of tens of thousands or hundreds of thousands of deaths is possible in the U.S. alone.
But the speed at which the outbreak occurs is extremely important to the consequences. What epidemiologists are most afraid of is that the health care system is flooded with a sudden explosion of disease, requiring more people to be hospitalized than they can handle. In that scenario, more people will die because there aren’t enough hospital beds or fans to keep them alive.
A catastrophic flood of hospitals can probably be prevented with protective measures that we see more of now – schools closing, cancellation of mass gatherings, working from home, self-quarantine, self-isolation, avoid crowds – to prevent the virus from spreading quickly.
This was illustrated by the popular “flat de curve” graphs, which depict a massive campaign to prevent hospital tools from being overloaded by a sudden wave of cases.
Many media outlets developed versions of this map, all adapted from previous CDC work related to flu pandemics.
As Carl Bergstrom, a biologist at the University of Washington, tweeted, the strength of this image is that it shows how measures to reduce the spread of the virus can save many lives “even if you don’t reduce the total number of cases.”
After all, deaths are a function of both the number of cases and the death rate of the case. If the number of cases rises so high that the case death rate begins to rise – if a city lacks ICU beds or fans or well-trained nurses – far more people will die than if the curve flattens.
2. This single photo explains the concept of * flattening of the epidemic curve * to non-specialists without the need for additional text. Looking at the photo, you can see that even if you don’t reduce the total number of cases, slowing down the rate of an epidemic can be critical.
– Carl T. Bergstrom (@CT_Bergstrom) March 6, 2020
Mitigation can also reduce the number of total cases. People are hopeful that a Covid-19 vaccine will be available sometime in 2021. If that happens, the right arm of the curve would essentially be shortened and the total number of people falling sick would decrease.
But the power of curved smoothing as a concept is that it doesn’t rely on that. It shows that many lives can be saved, even if 60 to 70 percent of the population falls ill before the herd’s immunity stops the virus from spreading further.
A natural addition to the smoothing of the curve is to try to increase the line by increasing the capacity of the healthcare system to treat Covid-19 cases.
Raising the line is a multifaceted task: canceling or postponing selection procedures, training additional medical personnel, building improvised hospitals, stimulating the production of personal protective equipment, making additional fans and cleverly moving the country to where they are needed.
You can think of it as working in two different ways. One is that even a flattened curve can still peak above the capacity of the health system (this appears to have happened in Wuhan, China; Lombardy, Italy; and to some extent in New York City), in which case attempts to raise the line directly save lives.
The other is that the height of the line determines how many flattenings are needed to save lives. A higher line means that not much flattening is needed, so social distance measures can be less extreme without risking overwhelming the health system.
Mitigation still means many deaths
If the sole purpose of U.S. policy is to prevent hospitals from becoming overwhelmed, it may be on track in most of the country. Even if the United States has become the country with the highest number of confirmed Covid-19 cases, the rate of growth – as represented by bending the line on a log scale – has clearly slowed.
A few states – mostly in the urban Northeast – use almost all of their ICU beds and need the number of cases to drop further to maintain the health system’s capacity. However, much of the country, including major blue states such as California and major red states such as Texas, is well below the level of hospital depletion and skyrocketing deaths.
California has been that, for example export fans to other states with greater needs. And even in New York, where hospital capacity was under pressure, there were emergency measures to increase the line ad hoc field hospitals and the shipping of the USNS Comfort to the city turned out to be largely unnecessary.
One reason for this is that in practice, the hospital admission rate for Covid-19 cases appears to be lower than model builders estimated five or six weeks ago, although the mortality rate is roughly in line with expectations.
And again, the * mortality rates * were about as bad as models feared / assumed. (Emphasis on roughly because there is some discussion.) However, hospital admission rates were lower than most forecasts. In general, there is some reluctance to talk about this.
– Nate Silver (@ NateSilver538) April 29, 2020
Unfortunately, we have seen that even if hospital systems are not overwhelmed and people can get good care of Covid-19, many people still die. And if more people get sick, more people will die.
At the current level of infection, the United States has no major hospital capacity issues. And it is plausible that the measured reopening steps underway in many states, if implemented carefully, are compatible with the country more or less plating with the current infection rate and avoiding the hospital’s nightmare scenario.
But every day, more than 1,500 Americans die from Covid-19, and research into “excessive deaths” makes it pretty clear that the real number is likely to be more than 2,000. Project that up to a year and the US death toll will land around 800,000.
Like Bergstrom and Natalie Dean, biostatistician at the University of Florida, explained in a recent New York Times opinionit would likely lead to a significant number of transmission infections and more than a million deaths if enough people become infected to achieve herd immunity.
Before that happens, there may be a vaccine or a super effective treatment drug. But the widely cited 18-month vaccination schedule is already very optimistic. Barring some unforeseen breakthrough, even a successful mitigation strategy can result in a death toll that is dramatically higher than what anyone active in politics is considering. The alternative strategy is to not only smooth the curve, but reduce the number of cases to zero.
Some countries follow the curve
Hong Kong, on the other hand, in the past three weeks has seen zero new cases most days.
Under the circumstances, the island moves steadily to ease restrictions. As of May 1, a rule has been lifted that requires restaurants to remain semi-empty, and now there is a slightly more lenient policy, allowing dining as long as the tables are 1.5 meters apart. And not only are restaurants open, people are screaming to go out and eat.
Not to say HK get going again but it’s a Wednesday night (albeit before a holiday) and I was just told it’s a 90 minute wait for a table in a pizzeria.
– Mike Brrrrrd (@Birdyword) April 29, 2020
With the disease seemingly reduced in the interior, Hong Kong is now able to shift its focus to a strategy focused on border controls. With the pandemic still raging worldwide, the city cannot completely abandon its vigilance. And because Hong Kong is so small and dependent on international trade, opening up the domestic service economy cannot really save the city from serious economic problems.
But the city has a clearly articulated strategy they call “suppress and override”: now limit the constraints when things are zero, then clamp them back if necessary to push things back down when they surface.
Taiwan has also had no new cases for several days, and since April 6, all reported cases of Taiwan have resulted from the mission of one naval vessel to the island of Palau instead of spreading the community. New Zealand has not fared so well, but the government believes it has successfully identified and isolated all of the country’s coronavirus cases and removes restrictions on the claim that the virus has been “eliminated” in the country.
The outbreak of South Korea is now thanks to single-digit numbers of new cases each week, and a key test for the country is whether an expected wave of vacation trips this week to Jeju Island (a major Korean tourist destination) will lead to a new one golf or if the peninsula can suppress the spread of the disease. South Korean professional baseball also resumed this week, but without fans in the stands.
The United States, meanwhile, is opening up on the basis of a vaguely formulated assumption that settling for mitigation is good enough.
One reason for the pressure to open is that while widespread shelter orders have clearly managed to slow the spread of infection, they do not quickly reduce the number of cases. Authorities fear the economic pain of lengthy shutdowns and it seems that the general public is getting impatient and starting to bend the rules.
But the reality is that the United States hasn’t really tried it the strategies that have made oppression successful. To achieve that, America should invest in expanding the number of tests, invest in more contact tracers, and create centralized quarantine facilities so that infected people aren’t just sent home to infect the rest of their household.
Since the US was not spending April on that, striving for repression – along the lines of Taiwan, Hong Kong, Korea, and New Zealand – would necessarily mean more slowdown and more economic pain. But this would potentially save tens of thousands or hundreds of thousands of lives and would almost certainly lead to better economic outcomes by really restarting activity.