A Reluctant Message About Masks

You may have noticed that I didn’t say a word about masks in my recent blog post. Because by this point people are already either on Team Mask or Team Naked Face and no one is switching teams. To Team Mask, masks have become an indicator of whether someone is informed/socially responsible or ignorant/narcissist. To Team Naked Face, they are an indicator of hysteria and infringement of basic freedoms.

To be clear, I have no sympathy for anyone who believes masks are an indicator of hysteria or infringement of freedoms. I believe stores are entirely within their rights to require customers to wear masks. But I do believe that the issue of masks has become more emotional than scientific. And the bifurcation of worldviews leaves little room for nuance. Which is unfortunate. Because a more nuanced understanding of what masks can and cannot do would be very helpful in the battle against COVID.

This is all I ask: that people approach masks the same way they approach acai berries. There’s no reason not to eat acai berries. It’s certainly possible these magical fruits reduce the risk of a heart attack. But we can all agree that mouthfuls of red berries shouldn’t substitute for actual lifestyle changes like daily exercise that are supported by real science. Keep in mind that there’s about as much science in support of homemade masks reducing COVID transmission as there is that acai berries will keep you from having a heart attack. Again, that doesn’t mean there’s no scientific support. It just means it’s not very strong. And just as you should be suspicious of any doctor that you can lie on the couch all day as long as you eat berries, you should be equally suspicious of businesses that try to convince you that your family is safe just because people are wearing masks. Masks are not sufficient. I have seen this message too many times: Masks that cover the mouth and nose–when worn by everyone in enclosed spaces–are the most effective means of reducing spread. It’s simply not true. Mask may be a supplemental help. Look instead for real organizational and systemic changes that reduce indoor human interactions. I know, it’s much easier to use masks as a litmus test. Don’t fall for it.

That said, no one should confuse the idea that cloth face masks may not be totally effective against COVID with the truly important problem of PPE shortages for healthcare workers. All masks are not equal. If you enjoyed shaming people for not wearing masks and still want to scratch that itch, maybe direct that energy towards the organizations responsible for healthcare workers needing to reuse PPE.

So why do we overweight our cloth masks? It’s not just because it’s a simple checkbox. It’s also just human nature. Humans are more visual than abstract thinkers. A person wearing a mask provides a stronger visual cue compared to subtler things like the absence of people, invisible contact networks, or ventilation.

So if you think the visual of people wearing masks could be interfering with your risk perception and giving you a false sense of security, do an easy thought experiment. Just pretend that no one around you is wearing a mask. Would you still be in that yoga class? Or that socially distanced happy hour? If the answer is no, trust your instincts and think hard about whether you should be there.

Want to read more about the science of masks? A nice review is provided by the National Academy of Sciences, Engineering, and Medicine’s Expert Consultation on the Effectiveness of Fabric Masks for the COVID-19 Pandemic.

TL;DR: Masks have become a centerpiece of the US strategy to reduce COVID. But scientists know that it will take a lot more than masks to control the pandemic. Masks should be encouraged, but not as a substitute for more important changes in behavior, mobility, and social engineering that are the real key to driving down virus transmission.

The Good News and the Bad News about COVID

Congratulations, America. You have survived Stage 1 of the Tour de COVID. You pulled together, and learned all kinds of new tricks. How to Zoom. How to order groceries online. How to homeschool kids. And the trickiest of all: how to walk in public with an awareness of your space that requires not burying your face in your phone. You did it!

The nightmare version of Leg 1 only hit certain places like nursing homes and New York City. Most of us have our families intact and only have secondhand stories of people who died. America is collectively having a huge sigh of relief. We’re ready to hit the beach and see old friends again.

I do want to emphasize that there is a lot of good news about COVID. So I’ll try to highlight that before dipping into the bad news.

Good News: New Zealand eradicating COVID means that the island country can return to normal life, complete with rugby games, schools, beach parties, and whatever else Kiwis do for hedonism. But this is even bigger than that. This proves to the rest of the world that you don’t need a miracle cure or vaccine for life to go back to normal. You can beat back COVID entirely with pure and simple public health and testing. Sure, there will continue to be imported cases from travelers who go abroad. But strict quarantine and testing can stamp out imported cases before they spread to the community. In theory, any country, rich or poor, can do this. It sure helps to be an island nation with secure borders. But Rwanda is taking notes.

Bad News: America ain’t New Zealand. Yesterday Arizona had over 1,000 new COVID cases. We managed to blow up our economy, spike unemployment, and we still can’t play baseball.

Science Sidebar: There are two particular features of the SARS-CoV-2 virus that explain why the pathogen is so difficult to control. The first, symptom heterogeneity, is something we’re all familiar with. The just means that some people have zero symptoms and some people die. If you wanted to bioengineer the worst possible pathogen for humanity, you wouldn’t want the virus to kill everyone it infected. That would just be like Ebola: really nasty in the people it infected, but not good at transmitting around the world. You would want it to be stealthier, transmitting silently in large numbers and only killing selectively. That’s COVID.

A feature of the SARS-CoV-2 virus that may be less familiar is transmission heterogeneity. This means that there’s a lot of variability in how many people an infected person actually transmits the virus to. You’ve probably heard about R, the reproduction number that represents how many secondary cases arise from an index case in a population, on average. This number is central to the growth dynamics of an outbreak. But you probably haven’t heard as much about the parameter k, transmission heterogeneity. When R = 2 it doesn’t mean that each person transmits to exactly 2 people. The vast majority of people could transmit to 0-2 people with only rare outliers transmitting to 10-20 or more. High transmission heterogeneity means a lot of superspreading events that introduce a great deal of volatility into the system. It means that lots of virus introductions into a given location won’t transmit at all. But occasionally you’ll get introductions that spread explosively and cause sudden outbreaks. This explains why you see communities have no cases for long periods of time followed by sudden explosive growth. It also explains why it’s hard to predict why the virus takes off explosively in some locations and not at all in others. A lot of it is just random. You can reach false conclusions if you don’t account for transmission heterogeneity. You can look at a community that never closed schools and has few cases and decide that schools aren’t important in transmission. When really it’s a lot of random chance whether a virus introduced into the community takes off or not. You just need to roll the dice enough times.

Good News: You can relax on the disinfectants. No need to wipe down your groceries (I never made my mom do this). Back in March I listened to a well known NYC ER doc advise people to Purell after every touch of a door handle or elevator button. I told Aaron this was nuts.

Bad News: Wiping down tables is far easier than coming up with new systems to minimize human contact and density and move activities outdoors. But restaurants, schools, day cares, gyms, or churches cannot simply make Purelling the hell out of the place the centerpiece of their control strategy.

Good News: I’ll continue to be bullish on vaccine development. I believe we will have a safe, effective vaccine in 2021.

Bad News: Once we make a vaccine, there’s a little problem called distribution. You need to get that vaccine into the arms of people. Millions of people. That requires logistical planning about where vaccine doses should go and when. Because we’re not going to have 300+ million doses all at once. We’re going to need to prioritize what groups get vaccine first. You may think this is obvious. Old people. Healthcare workers. But what about other essential workers on the front lines who put their lives at risk? Firefighters? Police? Grocery store workers? Utility workers? School teachers? Political leaders? And what happens when lobbying and political interests get involved? Should baseball players get vaccine before high-risk workers in meat-packing plants? Lordy…..Make yourself a big bucket of buttery popcorn and watch the food fight unfold. And remember, letting the neglected classes get crushed is just the American way.

Also…..as long as we’re getting the bad news out of the way. When you put a new vaccine in millions of people, some people will get sick shortly after. You can be sure that the stories will make media headlines way before epidemiologists have a chance to figure out whether there’s any connection. Most connections will be bogus (like autism and measles vaccine). But there is still debate over whether the adjuvanted 2009 H1N1 influenza pandemic vaccine caused narcolepsy in some kids in Europe. America is already a vaccine-skeptic country, so my intention is not to fan any of those flames. I and everyone I know will be getting the COVID vaccine. But be prepared for media headlines about vaccines causing disease. But as seen with the H1N1 narcolepsy story, it can be very hard to tell when there is a very small number of cases. Especially when publicity in the media artificially increases clinical diagnoses. Conditions like autism and narcolepsy can go years before being diagnosed, so higher awareness in a community can dramatically accelerate detection rates.

Good News: There has been an incredibly steep learning curve about how to better treat hospitalized COVID patients. At the beginning ER docs faced a novel disease and had very little information. While it’s tragic that there have been over 250,000 patients hospitalized for COVID in the US, at least we’ve used those experiences to learn a great deal about what works and what doesn’t. The debate about whether to treat with hydroxychloroquine was just the tip of the iceberg. We’ve learned a lot about how and when to administer oxygen, including less invasive and risky procedures than intubation. Simple steroids have also been shown to be effective. So if you have an occupation that puts you at high risk for contracting COVID, know that every month that you hold off infection makes a good clinical outcome more likely.

Bad News: This is a really nasty virus. COVID is more than a respiratory disease. It has high tropism, meaning that it can infect the brain, blood, kidneys, etc. The virus destroys the body in so many different ways. And many people who survive will still have severe longterm problems we’re just beginning to grasp.

Good News: Kids are still not getting severe disease. Furthermore, kids are half as likely to get as infected as adults.

Bad News: Kids that do get infected turn out to transmit just as well as adults. Even if kids are half as likely to get infected per exposure, kids typically have larger playgroups and more contacts than adults, so this kind of evens out. Especially when schools and day cares are operating. This means that if you want schools to be open in the fall, you’re going to need to really drive down COVID levels in your community over the summer.

Good News: America has all the ingredients needed to control COVID this summer. We have a coordinated, data-driven national strategy grounded in intensive testing and an ability to rapidly respond to new flare-ups. That includes consistent, transparent messaging with the public and businesses about how to safely, strategically reopen parts of the economy in a controlled, closely-tracked way.

Bad News: I made that up.

Good News: At least it’s true in Vermont.

TL;DR: America believes it has weathered the worst of COVID and is ready to return to normalcy. But the pathogen is not going away any time soon, and a rush to reopen businesses this summer will make it difficult to drive the disease down to levels low enough to reopen schools in the fall.

How the West Was Lost

Question: Why does the Bible command us to Be Fruitful and Multiply?

Answer: Because half of the kids weren’t going to see their fifth birthday.

For the last 10,000 years of human history, and as recently as my grandmother’s generation, children grew up with foundational experiences that included death of friends and family by infectious agents. Growing up in Boston in the 1930s, my grandmother was infected with tuberculosis as a teenager, along with her father. After witnessing her father’s death, she spent two harrowing years in a sanitorium for TB patients watching her friends get picked off one by one.

Our grandparents’ generation grew up with an intuitive, common sense understanding of infectious disease risk that modern Westerners sorely lack. Citizens can mobilize rapidly to to existential threats that are familiar: Israelis to terrorism, Japanese to tsunamis, or post-SARS Hong Kongese to pathogen outbreaks. No one in the West wants to return to an era when Little League also included funerals. But by growing up in sanitized environments, modern Westerners have lost not only their fear of pathogens, but an intuitive sense of how to respond to infectious disease threats that has caught us badly flatfooted during COVID. And now that COVID is here to stay, our inability to intuitively evaluate the risk of invisible bugs is highly problematic, sending us flailing from one reactionary extreme to another, from mask-shaming joggers to group protests against social distancing restrictions.

Stark differences in familiarity with infectious disease risk aren’t just divided generationally by time, but also geographically over space. My aunt recently did a clever survey of her friends, asking them to guess the answer to this question:

My New England seaside town Gloucester has 30,000 people and 15 confirmed deaths from COVID-19 so far. My daughter lives in Rwanda, a country with 12.8 million. How many deaths has Rwanda had?

No one guessed the right answer: zero. The trajectory of COVID-19 within Africa is highly bifurcated between countries that responded early and decisively, and those like Tanzania that sat on their hands. Well-governed countries like Rwanda have succeeded remarkably in containing the virus early through swift, stringent lockdown. Because swift, stringent lockdown is exactly what differentiates the COVID winners (New Zealand, South Africa, South Korea, Israel) from the COVID losers (US, UK, Brazil, Italy, Russia). It’s why there have been more deaths in Connecticut than in Japan, South Korea, Taiwan, Hong Kong, Singapore, and Vietnam combined. You know what happens when you win at COVID? Ask the New Zealanders. You get a dream world in which you go back to flipping pencils in the office while your kiddos go back to school. You get to not tank your economy.

Why was Rwanda better prepared for COVID than France or the United States? Isn’t the West supposed to be the standard bearer when it comes to international disease crises? Wasn’t it just a few years ago that US CDC was the one bailing out Africa during the Ebola outbreak?

The US has vastly more wealth and technology, but Rwanda has something proving to be more important. You could call it citizen mindset or infectious disease IQ. Not only do Rwandans need to manage endemic malaria, but also sporadic flare-ups of Ebola in their neighbor the DRC. Throughout the developing world, people regularly contend with a range of pathogens for which there is no vaccine available: Chagas disease (American trypanosomiasis), Llasa fever, Japanese encephalitis, chikungunya, dengue, Rift Valley fever, Crimean Congo hemorrhagic fever, Nipah, sleeping sickness, leishmaniasis, schistosomiasis. Throw in the pathogens Westerners have actually heard of like HIV, measles, SARS, and malaria, and you realize why cultures spanning most of the non-Western world know a thing or two about bugs.

What happens when the generations currently residing in the West are the first in human history to be raised in a bubble that doesn’t include regular experiences with infectious disease mortality among the young? Soccer moms in Orange County start refusing to give their kids the measles jab. Knowledge of infectious disease atrophies like a vestigial limb. France may have the Sorbonne, but its citizens score 40 points lower than Nicaragua in response to the question Do you believe vaccines are safe?  

Wealthy G8 countries tend to lag behind poorer countries in Asia, Africa
and South America in vaccine trust.

How did Western society do a 180-nosedive from the March of Dimes to anti-vaxxers in a single generation? Of course we know who’s to blame: Barack Obama. Funny joke, right? Well, there’s actually an underlying grain of truth. The reason why Americans and other Westerners have lost their intuitive sense of infectious disease risk is because Western governments and research institutions in the latter half of the 20th century were too effective in achieving victory after victory against the world’s deadliest pathogens . In the 1970s we wiped one of the deadliest pathogens (smallpox) off the face of the earth. A CDC-led team had to track the final cases hidden in the remotest corners of the Earth. Only Indiana Joes or superheroes in capes are supposed to do stuff like that.

The New Superheroes: CDC directors of the Global Smallpox Eradication Program celebrate their success, 1980.

For decades the US has been undisputed leader in global outbreak response. GW Bush is no favorite of the left, but his PEPFAR program gets enormous credit for delivering antiviral drugs at scale in sub-Saharan Africa and turning the tides of an HIV epidemic that was spiraling out of control. CDC was critically involved in helping countries in West Africa control their Ebola outbreaks, and more recently in the DRC, using a new Ebola vaccine developed at the US NIH.

Why did a nation with a legacy of fighting the rest of the world’s infectious disease problems fail to control COVID within its own borders?

Countries with higher GDPs still have a lot of advantages against COVID over the long run. We’ll develop a vaccine faster and have the funding to administer them at scale to our citizens. We’ll discover drugs that reduce disease severity, and churn those out as well. I’m no expert in finance, but the US seems to have a miraculous bottomless bank that creates trillions of dollars out of thin air so we can sustain social distancing without collapsing the economy.

But vaccine is a long way off, and SARS-CoV-2 was doing loops around the world at a time when all we had in our toolkit was human intelligence and willingness to mobilize. Government leaders needed the knowledge and foresight to swiftly grasp of the scale of a new disease threat and understand that the only way to contain the pathogen was through strong, early decisive lockdown, despite economic consequences. And citizens needed to consent. With less experience in managing infectious disease outbreaks than humans from any other time or place in history, the West simply could not mobilize in time.

In early March, when COVID was appearing in Europe and Asia but didn’t seem to have a foothold yet in the US, I had a conversation with my aunt from Gloucester, MA. She knew the city’s mayor personally and wanted expert advice. I told her just to focus on 3 things: (a) build testing capacity, (b) make plans for nursing homes, and (c) cancel large gatherings before the first case is detected. With a highly capable mayor in a small city with a longstanding culture of social welfare, I thought Gloucester, MA could be a model for US COVID response, and possibly avert a single death.

But it turns out it takes a lot of guts for government leaders to act preemptively to COVID. Americans know the drill when it comes to responding preemptively to hurricanes and natural disasters. Local leaders don’t have to sit on their hands and wait for the breeze on their cheeks before ordering an evacuation. But Americans have no experience with responding preemptively to invisible bugs lurking at their borders. You might as well have said space aliens were coming.

At face value, the upcoming months of relaxing lockdowns in the West should bring sunnier skies. However, reopening society would be smoother in a population with a better grasp of infectious disease risk. By their nature, humans have inherently flawed perception of risk. We tend to be afraid of anything considered foreign or exotic (like terrorists or sharks). But risks of familiar activities rapidly become normalized (like car accidents).

We are not aware of how rapidly we have normalized our sense of COVID risk. This is an actual conversation I had with a highly intelligent person:

Highly Intelligent Person: I’m going to join an outdoor group exercise class.

Me: Sure, if you feel comfortable. But you recognize that the risk here hasn’t actually changed much since a month ago? Not here.

Highly Intelligent Person: Well, I’m feeling a lot more comfortable having made it this far without getting infected.

Me: Yes, but you got this far because you social distanced.

Many of us are feeling pretty good about COVID. It’s become less scary. Our inner circle of friends and family probably haven’t been infected. But we need to be highly aware of how rapidly we normalize risk. And particularly how we bias our perception of disease risk. We still think we’re going to get infected by sink in a dirty public bathroom, not by our gorgeous wife.

And we need to realize that no matter how much New York Times we read, very few of us have an intuitive sense of how and where viruses transmit. To make it worse, we’re horrific at statistics. It’s not out fault. For some reason we have never updated our 1950s math curriculum that makes you take geometry and trig and maybe some calculus to get a high school diploma, but not a whiff of stats. The one damn thing in school that might have been useful.

While this post gives a mile-high view on why some the richest and most educated countries in the world have been the least equipped for COVID, there’s still a lot of guidance needed on the ground. Here’s my rapid-fire take on a couple burning issues:

Antibody tests. Everyone and their dog wants to know if their recent or distant cough meant they were infected by the virus. Detecting antibodies in the blood can theoretically tell you if you’ve recovered from a SARS-CoV-2 infection. Antibody tests are going to be crucial for assessing how many people in a community were infected from different age groups, and how many never experienced symptoms. These surveys will be helpful in assessing the role of children in transmission and the risks associated with reopening schools in the fall. But I’d be skeptical of serology tests being used to assess individual-level infection histories. The tests can be inaccurate (false negatives, false positives), particularly new serology tests that haven’t been completely validated. Statistical methods can account for these inaccuracies when pooled over a population, particularly when combined with PCR data. But I wouldn’t expect antibody tests to be particularly meaningful for individuals.

Mutations. You’re going to hear a lot about the blasted mutations. Yes, viruses evolve rapidly and the viruses circulating now have genetically drifted from the original strain that emerged in Wuhan in December. But if a country/region/city wants to explain to its citizens why COVID is exploding, they either need to admit mistakes or explain complicated statistics around stochasticity and random super-spreading. A far simpler strategy is to look at the viruses you observe locally, notice that there’s a new mutation, and blame that. But the vast majority of mutations will have little phenotypic affect. And it’s perfectly expected to observe new mutations crop up wherever the virus sustains transmission locally. For some reason people find mutations scary, so the media bites everyone time. Mutations are not scary. If it weren’t for mutations we would all be clones of our parents. Now that’s scary.

Yelling at runners not wearing masks. To Wear, Or Not to Wear. That is the Question that seems to be gripping American communities like no other. It’s silly. The science just isn’t there one way or the other. Ask the hamsters. If you personally like wearing a mask because you think there’s a chance it’s helpful and at very least it telegraphs to the world that you take COVID seriously and consider yourself a socially responsible citizen, then wear your mask with pride. If you don’t like wearing a mask, and are in a setting that doesn’t require it, don’t feel guilty. Focus instead on other kinds of running etiquette, and giving advance warning and lots of space as you pass. The key to COVID running is to not be such a hurry. Be prepared to slow down stop in your tracks entirely to let someone pass safely. Trust me, you’re not training for any races~

Should we be reopening? A big source of confusion is COVID risk is highly variable across the country. My brother lives in Burlington, VT, where there’s barely been any COVID and it makes complete sense to begin reopening businesses. But reopening needs to be data driven. Not motivated by politics or social distancing fatigue or a conviction that cold viruses simply can’t transmit when it’s hot and humid (just ask Singapore). So it’s important to understand what’s happening specifically in your community. You can find information on local health department websites. I find this site to be useful for state level trends. Here’s an example from Virginia, where there is little indication of cases coming down at a state level. At best the peak has simply leveled off.

There’s also high variability within a city. Here’s DC by neighborhood:

You’ll notice that COVID cases tend to crop up in lower income neighborhoods where people often have to report physically to their jobs, whereas higher income neighborhoods have more options for teleworking.

What began as a disease of the wealthy (travelers returning from cruises and Austrian ski vacations) is now evolving into a disease of the poor and neglected. The new epicenters will be in impoverished Indian reservations, prisons, slums, poorly run nursing homes. The virus is opportunistic and will exploit the vulnerable, especially where densities are high. Targeting such corners for disease control and testing is not merely altruistic. They provide breeding grounds for viruses that can reseed outbreaks in other settings and locations.

Monogamous Friendships. My chief advice for people who want to dip their toe into a world that includes (gasp!) friends is to approach it methodically and stepwise. In the upcoming months, I think it’s reasonable to make a single, highly selective step in expanding your contact bubble. Just for your own sanity. But think about which friend not only is someone whose judgment you really trust, but also someone who really matches you and your family in terms of risk profile. It needs to someone you’re so comfortable with that you can ask direct questions about their private habits and contacts. Because you’re putting the lives of your family in their hands. So you need to know if someone in their family must report physically to work. Or has a medical condition that requires visiting doctors. Or if they have to rely on babysitters for child care. Ideally they have similar shopping habits. And when you identify someone you are confident is a perfect match, be completely frank about offering a couple weeks of ‘monogamous’ friendship where you occasionally visit each other but agree not to see any other friends. Otherwise your contact network grows exponentially, since each contact has their own set of contacts. After a couple weeks or months you can swap and choose a new monogamous friend. But don’t cheat~

When will this end?

In the upcoming months, there will be intense pressure to relax social distancing, get back to work, and let your children run outside. And never return. For at least a week. Tension over when to open sectors of the economy will explode into a ThunderDome Death Match between politicians, economists, and medical experts.

Expect government officials and pundits to make a lot of noise about reopening sectors of the economy this month. Even though we have not yet reached the beginning of the end of the COVID pandemic. Even though in less connected cities and towns, we haven’t even reached the end of the beginning.

It might be helpful to think of social distancing as the country being on a ventilator. Meaning a life-saving intervention that is currently the only thing keeping us from a complete national medical catastrophe. There may be signs that the country (patient) is on the upswing. And that the most life-threatening period has passed. But would you ever want a loved one taken off a ventilator before a medical expert gave the okay? Even if your favorite politician thought she looked pretty good? The same should go for social distancing. Just as most doctors have a good idea of what would happen to a patient if taken off a ventilator too early, the scientists who study infectious disease patterns for a living can provide a pretty good picture of what will happen if we prematurely let up on social distancing.

Because, just to be clear, COVID will not just naturally disappear as the weather warms. Some of you may recall that the spring wave of the 2009 H1N1 influenza pandemic peaked in June in the United States. That does not mean that COVID will also peak in June. H1N1 has different transmission dynamics and was introduced later into the US, so there is no reason to expect COVID to follow the same curve. But it’s a potent reminder that novel pandemic viruses do not behave like seasonal viruses and are perfectly capable of causing severe outbreaks during warm weather.

This doesn’t mean we shouldn’t put economic considerations into the equation. If the ventilator for your loved one cost $1 trillion a day, you might decide to take the risk even if your doctor advised against it. The point is just that the economic considerations should be scrupulously and ethically weighed against well-calculated risks on both sides. It shouldn’t come down to a politician noticing a little improvement in a patient’s vital signs and deciding enough is enough.

The upshot is that, politics aside, everyone should be mentally prepared for intense social distancing to continue at least through the spring. But, the good news is that as we learn more about the virus, we can be more strategic about how we go about social distancing and hopefully improve our quality of life. The more we understand how the virus transmits, the more we can cut needless paranoia and allow ourselves new little freedoms.

For example, this is purely a hunch, but I think it’s possible that we’re going to find that talking is going to be important in the spread of the virus. We tend to focus on coughing and sneezing, but one of the things that makes COVID so dangerous is that it is transmitted early in the infection cycle. The 2003 SARS virus mostly spread to other people after a person had already been sick for a while. Same thing for Ebola. COVID is very different in that it spreads readily right at the onset of early symptoms, and to some degree even before symptoms. This makes it much harder to control. And suggests that while sneezing and coughing certainly helps transmission, talking may also contribute to transmission at those early stages of infection. In which case, better understanding of this could inform exactly how we go about social distancing. It could be that extended periods of talking with someone, even at a properly social-distanced 6 feet away, could be far riskier than opening a box of delivered groceries. (By the way, 6 feet is a good baseline number to shoot for in grocery stores, but is not entirely protective — the Washington State choir outbreak suggested that people singing loudly might shoot virus a lot farther than 6 feet and infect each other).

But the point is we still have a long way to go in understanding the virus and what activities put you at most risk for infection. So there is cause to be optimistic that even if social distancing continues into the summer, that greater knowledge of how the virus transmits could at least tell us how to do social distancing right and with less stress and uncertainty.

When we start to think about reopening parts of the country, it will be useful to look closely at how things play out first in Asia. A few months ago, I was able to be pretty confident in my early predictions about the arrival and intensity of COVID not because have great foresight, but because Asia (and later Italy) provided a vivid roadmap in real-time of how the virus was behaving, and it was obvious that America had no special sauce that would lead us to follow a different trajectory. If anything, America was less prepared. Since Singapore and other Asian countries that initially managed COVID far better than the US have already started to reopen society, we can get another sneak preview of how things are playing out. So far, their struggles with resurgences of COVID outbreaks as economies reopen sends a pretty clear message of how fraught reopening society is even in places where the virus has been brought under control and where intensive testing is available. This seems to be because, as long as there is trade and movement with other areas with high COVID activity, there are too many opportunities to import new cases. This will be particularly difficult in a country as large and heterogenous as the United States. It will be particularly interesting to see whether countries like New Zealand and Iceland can successfully reopen sectors of the economy in the upcoming months, as these would be the best candidates given their gold-medal responses to controlling COVID through high testing and early, intensive controls as well as innate geographical advantage as islands.

What seems to be most frustrating to people right now is the uncertainty about the future. The darn models keep changing. But the models aren’t changing. Human behavior is changing. The original models with really bad projections were based on what would happen if we didn’t do social distancing. New models have lower projections because they include current levels of social distancing. That doesn’t mean the original models were wrong. The models project much lower numbers because this is a virus with exponential growth potential, so changes in human contact patterns have really pronounced effects on the trajectory.

So why can’t models be smart and incrementally increase social distancing parameters as the number of cases and death increase, since this is how humans would naturally respond to a building threat? Because a smart model requires uniformly smart humans. And anyone who has been following the news knows just how much variability there has been in the responsiveness of leaders at different levels of US government, from the White House down to mayors of small towns. So instead models have to be updated manually according to real-world human behavior. If social distancing relaxes, the models will get updated again and show things getting worse.

The models also have a lot of uncertainty because they are only as good as the data fed into them. And, at least in the United States, the underlying data is unfortunately not so good. For a country that is a global leader in biomedical research, there will be a long reckoning at some point in the future about why our testing flopped.

This is what happens when testing flops:

  1. We are always chasing the epidemic from behind. Localities still seem to still be under the impression that they don’t need to do anything until a positive case pops up. But at our slow pace of testing, by the time even a single positive case is detected, the virus is already transmitting like wildfire in a community.
  2. We don’t know fundamental characteristics about COVID transmission, and these are needed to inform any decisions to relax social distancing. Things like how many people are asymptomatic or mild cases, and how important are they in overall transmission. Things like how much transmission occurs within households versus in the community. For all you parents out there, things like how important are children in transmission, and what would happen if we reopened schools?
  3. Not testing mild or asymptomatic cases means we don’t know how many people actually have had the virus. Those numbers you see of lab-confirmed cases are just a fraction of the real number. And therefore we don’t know what proportion of the American population will be immune in the upcoming months. In theory, the greater the number of people who are immune, the lower the rate of transmission in the community (Re) and the sooner we begin to relax social distancing. It becomes very difficult to model the effect of reopening sectors of the economy without a baseline estimate of existing immunity in the population and well as individual-level knowledge of

There are certainly blockbuster events that could accelerate the reopening of society. If a new or existing therapeutic is proven to be effective (with the emphasis on proven). If a vaccine becomes available at some point in 2021. A high-quality antibody test would be very useful in determining who has been infected and is now immune (I mention high-quality because serological tests often have too many false-positives or false-negatives to actually inform policy).

I do want to emphasize that the COVID pandemic will officially end at some point, likely in 2021. That does not mean that the virus will be eradicated from the human population. It just means that through natural infection or vaccination, there will be enough herd immunity* that the virus is controlled. It remains to be seen whether the coronavirus behaves like an influenza virus, recurring in seasonal epidemics that affect the global population and vary in intensity year-to-year. Or whether it is more like a measles virus, causing intense outbreaks in localized pockets where vaccination levels are low. The answer will depend on the effectiveness of the vaccine and the evolutionary rate of the virus, specifically in the spike protein on the exterior of the virus that will be targeted by human antibodies. [You would think we would have a little more baseline knowledge of how coronaviruses evolve year-to-year in humans, since there are at least four strains that circulate widely and cause common colds in humans (e.g., HKU-1, OC43). But since they don’t tend to kill people, there’s not much funding for research.]

So the upshot is that social distancing is not the new normal. The COVID pandemic will eventually end and we will eventually see the return of baseball and senior bingo nights. And, if we have the political will, it is possible that, after decades of underfunding pandemic research and preparedness, we will be better prepared next time.

*Note that we do not know how long natural immunity to the virus lasts. It is likely that people who are infected with COVID and recover will have strong immunity to it for at least several months, probably longer. In terms of opening society back up and improving quality of life, having a proportion of the country that is known to be immune would be extremely helpful logistically. But if we don’t test enough, there will be many people who are immune but don’t know it.

COVID book club

In case the COVID epidemic has piqued your interest in the wild world of viruses, there are some wonderful books that manage to pack scientific detail about viruses, emerging pathogens, and the threat of global pandemics into vivid narratives aimed for a lay audience. Here are 5 non-fiction books I recommend. Excellent for bored teenagers.

  1. Spillover, David Quammen (2012)

2. The Great Influenza, John Barry (2005)

3. The Viral Storm: the dawn of a new pandemic age, Nathan Wolfe (2012)

4. The Coming Plague: Newly Emerging Diseases in a World Out of Balance, Laurie Garrett (1995)

5. The Planet of Viruses, Carl Zimmer (2015)

Return of the COVID blog

I apologize for the three-week hiatus in the COVID blog. A few Saturdays ago a horse and I had a little disagreement about whether I or not I should be on its back. The horse won. The Suburban Hospital ER was empty except for a few COVID cases, and scans showed that I had experienced a concussion but nothing life-threatening. It was a strange twist of fate that after preparing my entire career for something like COVID, I was suddenly in a brain fog and incapable of viewing screens or even following the COVID situation by radio. For all of you moaning about social distancing, try doing it for a week plus without the distraction of work, Netflix, or cat memes.

So after a three-week semi-dream state I’ve woken up to a brave new world. Some things have stayed the same, including the curious age structure that is mostly sparing children from severe disease. And of course the absence of toilet paper on any store shelves. But in my three-week haze America raced to the top of the global COVID case count. Part of this is increased testing, which we finally seem to have gotten up and running after a botched start. However, I will continually emphasize that the true epidemic is far worse than the case counts and death tallies suggest. We are still massively undercounting cases due to lack of testing. And even the reported death toll is just a fraction of the true number of deaths (which will be statistically calculated, within a certain degree of uncertainty, only after the pandemic is over).

Somehow my concussion perfectly overlapped with the critical early growth period of COVID cases in the United States.

Americans also seem to have all overnight become experts in viral epidemiology. Who imagined a day when droplet transmission and reproductive number would be trending on Twitter? There are some consequences to the proliferation of armchair epidemiologists. The internet has become a cesspool of misinformation. At this moment Americans desperately need reliable information. To know how to stay safe. And to know that they’re tanking their economy for a good reason.

One of the problems is that it can be difficult to evaluate the quality of information, especially on social media. A source of confusion is figuring out who is really an expert. A professor from an eminent US university with a fancy title (e.g., laureate) seems like they should know what they’re talking about. So should a top infectious disease doctor. I can perfectly understand that when COVID is taking over people’s lives there is an irresistible urge for anyone with a loose connection to the biomedical field to weigh in on aspects of COVID. Even in areas far beyond their specific area of expertise. You just make some reasonable assumptions, plug in the little numbers (death rates, attack rates, our favorite reproductive rates), and voila! Anyone can make a pretty graph.

I wish it were so easy to make a good COVID model. I wish we had really good underlying data drawn from intensive testing so we could nail down even simple parameters like rates of mild and asymptomatic infection. Across all age groups so we could know whether children were important in transmission. I wish we had a finely-tuned model that could be more prescriptive about the kinds and intensity of social distancing is needed and for how long. Right now, in the absence of good data, we’re using social distancing as a bludgeon rather than a scalpel. It’s like an elimination diet where you just stop eating everything — gluten, eggs, nuts, dairy — because you don’t know the specific culprit yet.

At this time, we have enough information to know that the situation will be dire if we don’t do anything. That’s not up for debate. But we don’t have a more detailed model that could fine-tune our approach to social distancing. At least not at this late stage in the epidemic (more targeted contact tracing was an option early when the virus was just entering the US but at this stage could not be done without a massively higher intensity of testing). The current unavailability of a good model that can answer our most pressing questions is not because US epidemiologists aren’t any good. There is a relatively small and tight-knit community of experts in the field of mathematical modeling of emerging pathogens. The community has grown over the last decade, as H5N1, Ebola, Zika, and other emerging infectious diseases have increased funding and scientific interest. However, neither H5N1, Ebola, nor Zika every truly invaded America, leading us to become complacent, and major funding networks across the US government for infectious disease modeling have lapsed.

So the list of people who have any business building models to predict the trajectory of COVID-19 would fit no more than a single page. What makes building a COVID model so difficult is how many uncertainties there are, not just about the virus but also human behavior. It doesn’t matter if you’re a professor of epidemiology at an esteemed university or one of the top infectious disease doctors, it takes years of specific study of modeling pathogen dynamics needed to accurately account for these kinds of uncertainties.

Just, to be clear, I am not a modeler. I am an evolutionary biologist who happens to work closely with mathematical modelers. I have great respect for how difficult it is to make a good COVID model. There are enormous gaps in data and information needed to parameterize it. And the parameters are constantly changing as humans modify their behavior. And being a good modeler is a thankless job. Either everyone ignores your model and you fail to help anyone at all. Or politicians take necessary actions and avert a full-scale epidemic, effectively making all your original projections wrong. Which is of course a good thing. But it leaves people with the impression that modelers chronically over-hype.

My area of expertise is how viruses evolve. I’m the person who knocks down the rumors that there are genetically different strains of COVID circulating that cause different severity of disease. Or that the virus will mutate over time to become less lethal and more like a typical cold virus. (It’s quite the reverse. Over the next year or so humans are the ones who will be changing, developing natural or vaccine-induced immunity that makes re-infection less severe. Over the next decade, rather than mutating to become less severe, coronaviruses, like influenza viruses, may continue to evolve to evade human immune responses and cause recurring seasonal epidemics.)

Okay, I’m still limited in my daily allowed screen time. But I want to clear about one final thing that seems to be tripping people up: masks. I am absolutely heartened that so many Americans are willing to don masks to help #flattenthecurve. While it’s been part of Asian culture for a long time to wear a mask when you’re sick, in the West masks are only for doctors, nurses, and Halloween. So, should we imitate the Asians? First off, there is currently a severe shortage of medical-grade masks for doctors and nurses, so if you have a commercially made mask you should donate it instead of wearing it. Even if you’re a risk group. Because the mask cannot protect you, it can only reduce the likelihood of you infecting someone else. Second, if you’re considering making a homemade mask, there’s been a line of thinking that this should be encouraged because keeping your mouth from spewing droplets and infecting other people, and at worst it couldn’t do any harm. But I do think people should be aware of potential harm. An important part of the fight against COVID is training people not to touch their faces. So just be aware that masks that are itchy, uncomfortable, or ill-fitting could actually draw the hand the face, especially for people who aren’t used to wearing them. Or subconsciously give people a false sense of security that emboldened them to do activities they otherwise wouldn’t do without a mask. So it’s not a scientific question, but a human behavior question. But if you’re in a position where you simply can’t follow social distancing at all times (e.g., an essential worker who needs to ride the subway), a mask may be appropriate for a limited time period of commuting. But just keep in mind that the mask is not protecting you.

Hmm, I wanted to finish on a more positive note. I’m going to stretch my screen time a little further to mention something people should do that I don’t think has received as much attention as it deserves. Certainly not compared to masks….. Pre-symptomatic transmission can occur in the days before a person’s first signs of infection (fever, cough, shortness of breath). So if you get a positive test result for COVID, right away you need to inform any people that you had close contact with in the days prior, even before the onset of symptoms, so those people can self-isolate. You should also consider informing people who regularly share surfaces with you know, for example residents in your apartment building. The key is to inform people right away – not days or weeks later.

Taking this another step further, I would even encourage people to let contacts know before you get a test result, as soon as you have been declared a suspected COVID case by your doctor based on symptoms. Many people are still not getting tested, either because they have a mild case or because tests in their area are still unavailable. I understand the desire not to make people unnecessarily anxious, but personally I would rather know if a contact of mine was a suspected COVID case, so I could decide for myself how to act on the information. You could decide to entirely self-isolate if you live with people in risk groups. Or simply postpone going to the grocery store for a period of time.

Most of us are social distancing and hopefully won’t have many contacts. But there are essential workers, whether working in hospitals, grocery stores, or the government, who are still reporting to work, which is why transmission is still occurring. Recognizing the risk for pre-symptomatic transmission and rapidly sharing information about early symptoms and test results with contacts could be a simple, relatively non-draconian way we empower each other to make informed decisions to momentarily intensify our social distancing and potentially break onward transmission of secondary cases.


  • American is experiencing a catastrophic disease event unlike anything we’ve experienced since 1918. The virus is deadly and transmissible.
  • Other countries in Asia have demonstrated that an early, aggressive campaign of intensive testing, targeted isolation and contact tracing, and social distancing can dramatically reduce COVID transmission. But America missed that early intervention window and we are now the global epicenter. Ahead is a long period of economic hardship coupled with sickness and death.
  • The good news? This is not the new normal. Eventually people will gain immunity, either through natural infection or vaccination. This time next year I am cheerfully optimistic we will be enjoying spring baseball again.

Question for next time: what are the most likely scenarios for the rest of the year?

Good COVID Ideas

Across the world, people are coming up with great ideas for how to deal with the COVID emergency. I’ll keep a running list in case others want to copy.

Senior hotline for COVID information, isolation (CA)

Drive-through testing clinics (NY)

Free COVID testing (including emergency room visit), regardless of health insurance (Fed)

Paid sick leave (Fed)

Immune Corps! Enlisting people who have recovered from COVID and are immune to serve on the front lines of the response for people who are sick or in need (I made this up)

Develop hospital surge capacity plans

Helpful COVID Resources

Map of global COVID cases/deaths, Johns Hopkins University

Good summary of several aspects of COVID outbreak

Incubation Period


CDC info for preparing your home

Deeper into the Science

Modeling tool to simulate how interventions could alter the course of a COVID epidemic, Richard Neher

Molecular epidemiology, Trevor Bedford

If you think you’re infected

Week 11 (March 9-15)

Do I have COVID?

The most tell-tale signs of a COVID infection are:


-Dry cough

-Shortness of breath

More details on symptoms here.

A snuffy/runny nose is a sign that you probably have just a regular respiratory infection.

What do I do now?

If you are severely ill, go to the hospital. If you think you can rest at home, call your doctor. Your doctor will probably tell you just to stay home, hydrate, and self-isolate for 14 days. There are no treatments available at this time. This is a virus, not bacteria, so antibiotics are not helpful.

Should I try to get tested?

There are many advantages to knowing if you’re positive or negative for COVID. You can inform people you had prior contact with, since there does seem to be transmission in conjunction with the very first mild symptoms.

When you recover from COVID, you will have immunity for at least the near term. When all hell breaks loose, it would be highly useful to have a cadre of people who know they are immune who could be on the front lines to serve vital roles in the community. For example:

(a) Delivering groceries and essentials to older people at home

(b) Interacting with patients arriving for COVID testing at drive-through clinics

That said, there is a severe shortage of testing capacity for COVID. Most people are being told to not even try.

How Afraid Should I Be?

Week 11 (March 9-15)

There’s a simple reason people can’t decide whether they should be scared out of their wits or shrugging off COVID. It’s because human brains are not good at thinking at abstract population levels. Even doctors’ brains. But you cannot understand the current COVID situation if you don’t have a simple understanding of contact networks and population-level risk. So I’m going to give it a try:

Why you will be scared. COVID is frightening at a population level. At some point, in the upcoming year, someone you know personally is going to die of COVID.

This is because:

In a likely scenario, around 40% of the population will be infected. So if you know 1,000 people, 400 will get the virus.

Maybe half will show symptoms (200 people). Of those, 0.5-1% are estimated to die (1-2 people).

Social networks are complex and variable. You might not have 1,000 close friends, but if you have a couple hundred friends on Facebook, and know many of their friends and family members, plus people you went to school with, people your kids go to school with, and their parents….well, pretty soon your contact network starts to grow exponentially and a thousand is probably a pretty low-ball number.

Why you should not be panicked. The natural instinct is to panic when you hear about people you know getting severely ill and dying. But no matter what happens around you, you must remember that the circle of people you know is probably larger than you realize, and your personal risk is still very low if you are a young, healthy adult or child. Even if you get infected, you have a very slim chance of severe illness.

In the near term, what this means is that your decision to limit your social contacts over the next weeks and months is not so much for your personal health, or that of your children, but for your community and your circle of contacts, especially those at high risk.

That’s why it makes strong mathematical sense to upend your life in many ways, to protect your larger community, but not to panic, recognizing that your personal health is probably not at great risk.

Also, expect that there will absolutely be news reports about children and young people getting severely ill and dying. If large enough numbers of them get infected, there will inevitably be a small number that have poor outcomes. But these news reports do not change the fact that you are low risk.