Coronavirus April 2020—Part 5 The Real Risk of Death

As with previous articles in this series, in this article we’re going to cover more of the emerging science around COVID-19. We’re also going to talk about the psychology of pandemics and how to manage mental health issues that may be arising as we all socially isolate ourselves. (Lots of science in this one, so in each section we’ve also highlighted the BOTTOM LINE in CAPS and bold.)

SYMPTOMS and clinical course

As we’re gaining more experience with the multiple ways COVID-19 presents, a few items are beginning to stand out:

  1. Anecdotal reports are beginning to surface that many patients are reporting a loss of sense of smell and/or taste with COVID-19. Studies need to be done to confirm this finding, but it’s beginning to seem robust.
  2. There does appear to be the potential for injury from SARS-CoV-2 to the heart, which may explain why patients with cardiovascular disease seem more likely to die from COVID-19.
  3. Fevers, body aches, and chills begin in many patients after a few days of mild illness and can be prolonged.
  4. GI symptoms (frequencies in parentheses) are being reported in one retrospective study, including loss of appetite (40 – 50 percent), diarrhea (2 – 50 percent), nausea (1 – 30 percent), vomiting (4 – 67 percent), rectal bleeding (4 – 16 percent), and abdominal pain (2 – 6 percent). You can see by the ranges that symptoms are highly variable from person to person.
  5. In some patients, especially patients sick enough to require hospitalization, the second week of illness can be worse than the first. Anecdotal reports also suggest that some hospitalized patients seem to improve temporarily before crashing and requiring oxygen and/or mechanical ventilation. The time to recover from mild disease appears to be on the order of 2-3 weeks, while for moderately to severely ill patients, 3-4 weeks.


Currently, the gold-standard for diagnosing COVID-19 is via something called reverse-transcriptase polymerase chain reaction (RT-PCR), where a swab from your nose or throat detects the RNA of the virus. Importantly, this test can remain positive for variable amounts of time even in the absence of infectious virus (say, after you’ve recovered). Merely having RNA of the virus in your nose or mouth doesn’t necessarily mean you’re infectious at the time you have the test.

Unfortunately, the RT-PCR for SARS-CoV-2 has, surprisingly, turned out not to be very accurate. In a study that compared rates of positive RT-PCR tests with chest CT scans looking for pneumonia (where a positive CT scan in a high-risk patient was considered the gold standard), the negative predictive value of the RT-PCR test (meaning the likelihood that when the test was negative the patient didn’t have the disease) was only 74 percent.

That means if you’re sick with symptoms that look like COVID-19 and the RT-PCR test is negative, there’s still a 26 percent chance you actually have COVID-19. Luckily, the positive predictive value in that study (meaning the likelihood that when the test was positive that the patient did have the disease) was 97 percent. That means if you’re sick with symptoms that look like COVID-19 and the RT-PCR is positive, there’s only a three percent chance the test is wrong.

Unfortunately, it’s not clear, even in a high-risk patient, that a CT of the chest with typical findings of COVID-19 is an accurate gold standard. Thus, the values above may not be—and probably aren’t—entirely accurate. They are, however, what we have right now.

BOTTOM LINE: if you have symptoms consistent with COVID-19 and test positive, you almost certainly have it. If you have symptoms consistent with COVID-19 and test negative, you might still be infected.

Our inability to test for COVID-19 accurately has enormous implications. Even once we have easy-to-access, rapid RT-PCR testing available so that anyone who’s sick can be tested, we’re at risk for mistakenly believing that patients who are infected with COVID-19 have something else. That means that anyone who might have COVID-19, regardless of their RT-PCR status, should be assumed to have it and self-quarantine. And that means anyone with any infectious symptoms at all. We need a better test.

One will eventually appear in the form of an antibody test. When a patient becomes infected with any virus or a bacteria, his or her body soon learns to make antibodies to it that typically confer immunity. In one study, 100 percent of patients developed antibodies to SARS-CoV-2 by day fifteen of their infection. Most likely these antibodies will confer immunity (though for how long we don’t know). Unfortunately, antibodies form as patients are already getting better so aren’t useful for diagnosing diseases until the patient has recovered. Obviously, antibody testing will still be enormously useful for letting patients know they’ve already been infected and are now (at least temporarily) immune. Yet much work needs to be done to figure out what level of antibody titer confers immunity and for how long.


We still don’t know everything we need to know about how SARS-CoV-2 is transmitted or which forms of transmission are most likely. In one study, the virus was found to remain on stainless steel and plastic for up to 72 hours, though importantly the concentration of virus on these surface rapidly declined over that time (by 20 hours the concentration was 1/1000 of what it was at the start). It was found to last only 4 hours on copper and only 24 hours on cardboard. (Luckily, in a recent study of respiratory viruses spread in an office environment, having employees wash their hands once during the workday and sanitizing high-touch surfaces like door handles and countertops once midday reduced the concentration of virus in the environment and on employees’ hands by 85 percent.)

A fact drawn from the SARS-CoV-2 study above and widely publicized is that the virus can remain airborne for up to three hours. But this study has been widely misinterpreted. It didn’t find that the virus can remain airborne for three hours. The study used a device called a Goldberg drum whose function is to maintain infectious agents in an aerosolized form to study what happens to them when they’re suspended in air. It’s not used to determine how long an infectious agent remains aerosolized in real-world settings. In a Goldberg drum, an infectious agent will remain aerosolized as long as the drum is rotating. The question the study sought to answer was instead: how long does it take for SARS-CoV-2 to break down when it’s made to remain aerosolized? The answer: SARS-CoV-2 remains viable in the air for at least three hours (importantly, the concentration in the air was reduced to 1/10 of its original concentration within one hour). But that doesn’t mean when a person coughs, sneezes, talks, or breathes that the virus hangs in the air for three hours. Three hours was simply how long the researchers kept the virus hanging in the air through the artificial means of the Goldberg drum to measure how quickly the virus degraded.

What we really want to know is: when someone coughs, sneezes, talks, or breathes, how much virus is sent into the air, how much of that virus sent into the air must we inhale to become infected, and how long does that virus remain in the air for us to inhale before it falls to the ground?

We do know that SARS-CoV-2 RNA has been found in the air outlet fan of a hospital room housing a patient with COVID-19. This finding supports a new theory about how coughing, sneezing, and talking expels infectious agents of any kind, the turbulent gas cloud hypothesis, which does away with the black-and-white way of thinking about spread through respiratory droplets (from large infectious particles) compared to spread through aerosols (from small infectious particles). In the turbulent gas cloud hypothesis, smaller particles containing viruses can spread from a cough or a sneeze as far as 27 feet. No one has yet studied at what distance from a cough, sneeze, talking, or breathing a person infected with SARS-CoV-2 can spread a high enough concentration of virus to infect another person. If the turbulent gas cloud hypothesis is correct, though, the six feet currently recommended might not be enough for coughs or sneezes. It’s likely, though, that the distance you’d need to keep away from someone infected with SARS-CoV-2 who wasn’t coughing or sneezing but merely talking or breathing would be less.

As evidence in support of that, in the study mentioned above in which SARS-CoV-2 RNA was found in the air outlet fan of a hospital room of a patient with COVID-19, repeated air sampling in the room failed to find the virus. This patient had mild upper respiratory tract symptoms only and presumably wasn’t coughing, suggesting that being infected without coughing may not be enough to spread the virus into the air. (On the other hand, how then do we explain how RNA wound up on the air outlet fan? Perhaps the patient coughed or sneezed a little. The paper didn’t say.)

It is clear, however, from studies on influenza that infectious agents require a minimum concentration to cause an infection. If a person is exposed to a virus in a concentration below that critical level, he or she won’t become sick. Further, large droplets (> 50 micrometers) produced by coughing, sneezing, talking, and breathing fall to the ground almost immediately. Intermediate-sized droplets (10-50 micrometers) fall within minutes. Small droplets (<10 micrometers) can remain airborne for hours. Unfortunately, more than 99 percent of exhaled influenza particles are smaller than 5 micrometers. On the other hand, air temperature, humidity, U-V light, and other factors all influence the concentration of infectious influenza viral particles in the air. Finally, in a study that looked at how seasonal coronaviruses (not SARS-CoV-2) transmit into the air, in the subset of patients in the study who weren’t coughing but were merely breathing for 30 minutes during which samples were taken, there were no respiratory droplets or aerosols with virus detected, suggesting it may be hard to catch SARS-CoV-2 simply by being near an infected patient (who isn’t coughing). (Caveat: the study used a very small sample size!)

We don’t yet know the minimum concentration of SARS-CoV-2 required to infect someone. We don’t yet know how the virus is affected by environmental factors. It is, however, prudent to assume that, like influenza, some number of infectious SARS-CoV-2 particles are sent into the air even with breathing. Whether they’re small enough to remain airborne for more than a few minutes in a concentration sufficient to cause infection isn’t known.

Also interestingly, in a mouse model of influenza, the initial concentration of influenza virus mice were exposed to correlated with the severity of disease. If this also holds true for SARS0-CoV-2, it may explain why healthcare workers, who are typically around higher concentrations of aerosolized SARS-CoV-2, are not only more frequently infected but also may have worse outcomes. Thus, being coughed on directly by an COVID-19 patient may not only make it more likely that you’ll contract the disease than when you merely touch a contaminated surface and then touch your mouth, but may also increase your risk of dying from it. Further studies are required to confirm this hypothesis.

BOTTOM LINE: people infected with SARS-CoV-2 spread the virus into their immediate environment. Almost certainly, there is transmission of virus in the asymptomatic stage, but to what degree that occurs via turbulent gas compared to direct contact (infected hands on surfaces) remains unclear.


The next logical question to ask—especially given the news that the CDC is reconsidering its recommendation on the topic—is whether or not people should wear masks when going out in public. Other countries that seemed to have flattened the curve have seen widespread use of masks by the public. But was that mask use responsible for the flattening of the curve?

The answer is that it may have contributed but probably hasn’t been a major factor. To prevent a mask-wearer from contracting SARS-CoV-2 when an infected person coughs in their environment requires a properly-fitted N95 respirator mask. Have you seen Facebook pictures of healthcare workers showing the bruises on their faces after a day of wearing an N95 respirator? That’s the kind of seal these masks require to be effective. The masks must be fit tested, meaning proven to seal the mouth and nose from the outside world correctly, before healthcare workers can rely on them. Many fit tests fail (one study found that 65 percent of healthcare workers didn’t achieve an effective seal). Even if enough N95 respirator masks were available to the general public, without being trained on how to wear them and without being test fitted, the percentage of people getting benefit from wearing them would likely be even lower.

On the other hand, non-N95 masks, like regular surgical masks, might provide a modest societal benefit. If someone is coughing, it may limit the spread of the turbulent gas cloud and reduce the risk of the mask wearer infecting others. But if someone is coughing, they should already be self-quarantined at home! Indeed, the likelihood of encountering an infected person who’s coughing in public is low with all the social distancing measures now in effect across the country.

The circumstance in which it may make sense for someone to wear a mask is if they’re confined at home with someone who has COVID-19 and is coughing. We’re recommending in that situation that the infected person be confined to a room that no one else enters given the evidence that the entire room will be covered in virus. Every time that person coughs, virus is sent into the air. We don’t know how long it lingers there, but best not to be in the room when it happens. There is, however, controversial evidence that non-N95 masks are effective at preventing infection in people who wear them while remaining in close quarters of infected patients.

BOTTOM LINE: wearing a non-N95 mask probably won’t appreciably decrease your risk of getting COVID-19, but if you’re especially concerned about the possibility that merely talking and breathing may spread the virus into the air—which remains an open question—it might be of modest benefit. If there’s any rationale for the population wearing masks in public en masse, it’s that patients who don’t know they’re infected might not be as likely to spread infection to others.


Can pets transmit the virus? Two dogs have been reported to have positive RT-PCR tests. One, a Pomeranian, never had a positive viral culture and never developed antibodies to the virus, suggesting there was never an active infection. As yet, we have no results for viral culture or antibody detection on the other dog, a German shepherd. The CDC hasn’t received reports of pets infected with SARS-CoV-2 or of animal-to-human transmission from pets, but absence of evidence doesn’t constitute evidence of absence.

BOTTOM LINE: keep your pets away from others in case the virus can use pets as an intermediate host. If your pet must contact others (in doggy daycare, for example), wipe your dog down with disinfectant wipes if feasible.


As COVID-19 was beginning to appear in the U.S. in significant numbers only three short weeks ago, most people we talked with about it seemed sanguine. Yet the following week we received over 70 calls regarding COVID-19 in the space of two days. What happened? Why did everyone suddenly become so anxious?

We’d argue it’s because people think anecdotally rather than statistically. That is, our emotions are aroused more by stories than numbers. Not only has the news coverage in the last three weeks been almost exclusively devoted to COVID-19, ensuring it remains top of mind for all of us, but also the stories have been almost exclusively about people who’ve contracted it and/or died from it (this is especially activating when the person is young or a celebrity, whom we often feel we know). What we’re not seeing in the news are all the stories of those who got COVID-19 and survived. The result of this negatively biased reporting is to make us all feel that the likelihood of our dying from COVID-19 is much higher than probability suggests.

The antidote to this anxiety may lie in reminding ourselves of the data. From a recent study in China, we find the following risks of death from COVID-19, stratified by age:

0-9 0.0094%
10-19 0.022%
20-29 0.091%
30-39 0.18%
40-49 0.4%
50-59 1.3%
60-69 4.6%
70-79 9.8%
80+ 18%


Further, the negative emotional impact of a statistic is always worse when we focus on the outcome we don’t want (death from COVID-19). This is how the chart looks when we provide the same information but focus on the outcome we do want (the likelihood of surviving COVID-19):

0-9 99.99%
10-19 99.97%
20-29 99.90%
30-39 99.82%
40-49 99.6%
50-59 98.7%
60-69 95.4%
70-79 90.2%
80+ 82%


Of course, if we’re talking about millions of people becoming infected, the absolute numbers of deaths will be terribly high. But for an individual facing his or her own course of illness with COVID-19—with the caveat that these statistics may vary by country and even by regions within a country depending on many factors—these statistics are heartening.

These statistics don’t, of course, address what’s happening to our economy and the record-setting rise in unemployment claims we’re now seeing. We want to acknowledge that there’s plenty of reason to be anxious about that.

BOTTOM LINE: whenever you find yourself becoming overly anxious about contracting COVID-19, remind yourself of the actual statistics—and not of the risk you have of dying from it but of the likelihood you have of surviving it.

Weight gain

The strategies we’re using to prevent ourselves from contracting COVID-19—social distancing, social isolation, and self-quarantine—have unfortunately in many instances placed limitations on our ability to exercise and may have changed our eating patterns in a way that increases our risk for weight gain. What can we do to prevent weight gain during this new normal?

First, we need to openly acknowledge our worry about the effects that SARS-CoV-2 may have on our health, our finances, and our families. Especially when left unacknowledged, anxiety can lead to stress eating. Food is a source of comfort for many. This needn’t be a bad thing, however, if we make sure to snack on healthy foods, like fruits and vegetables.

The key to avoiding eating unhealthy foods is to make they aren’t in our environment. Because dining out in most cities in the U.S. is now unavailable, the only environment we need to control is the one in our own homes. Tried and true strategies will still work: 1) don’t shop when you’re hungry, and 2) don’t buy unhealthy foods in the first place (if they’re in your home, you’re going to eat them eventually).


No one will want to hear this, but the COVID-19 race is going to be a marathon, not a sprint. Flattening the curve does nothing to reduce the total number of cases destined to occur. It only slows the rate of new cases. Only two things will speed our ability to return to normal faster: 1) an effective vaccine (by some estimates, at least 12-18 months out), and 2) a medicine that significantly reduces the risk of death from COVID-19. Randomized, double-blind, placebo controlled trials take time, but they’re the ONLY way to know if an intervention actually helps more than harms. Maybe we’ll get lucky, and one of the agents currently being investigated will turn out to be a game changer. We can all hope. In the meantime, set your expectations realistically. What we’re experiencing now is going to be the new normal for some time.

  1. For previous posts related to COVID-19, see:
    1. Coronavirus February 2020—Part 1 What We Know So Far
    2. Coronavirus March 2020—Part 2 Measures to Protect Yourself
    3. Supporting Employee Health During the Coronavirus Pandemic
    4. Coronavirus March 2020—Part 3 Symptoms and Risks
    5. Coronavirus March 2020—Part 4 The Truth about Hydroxychloroquine

[jetpack_subscription_form title=” subscribe_text=’Sign up to get notified when a new blog post has been published.’ subscribe_button=’Sign Me Up’ show_subscribers_total=’0′]

Leave A Comment

Your email address will not be published. Required fields are marked *

  • On the subject of anxiety about COVID, there is a good discussion by Esther Perel on the multiple ways it challenges to our ways of socializing.

    On the subject of masks, these short notes I wrote will go on an economics research & policy blog. (slightly edited)
    I think that we economists should remind policy makers (such as WHO) that testing face masks in clinical settings will always miss the substantial shift in behavior when consumers are given a “gift” of advanced technology expanding people’s available opportunities across riskiness and daily (fun) activities. Substitution by consumers in response to the lower price of risky (fun) behaviors affects the new cool tech’s effectiveness in use. The fancier the construction & materials, the lower is the incentive to avoid risky behaviors; thus more socializing is done if considered safer wearing a fancy mask.

    Simple cloth masks function far better in life than in clinical comparisons, because they are first and foremost reminders to use best practices. All the desired practices, such as avoidance of going out in company while sneezing/coughing, hand washing after contact with other people or communally touched surfaces, and other cleaning practices remain VERY COSTLY with mere cloth than some (overly?) trusted and possibly incorrectly used respirator.
    *** grt

  • These articles are tremendously thoughtful, clear, and useful. I appreciate the evidence-based approach. Thank you.

  • Doc, thanks for your synthesis. This is well needed to cut through the noise.

    My mantra during these times if TFY (Think For Yourself)!.

    Even government officials are confused ‍♂️


  • Varying death rates among countries reflect, in part, each country’s level of care. In the US, if death rates can be driven down with enough healthcare workers, respirators, etc…, could we go the herd immunity route and save the economy somewhat? Thank you in advance.

    Probably not. The likelihood that our healthcare system could grow sufficiently in the next few months to handle all of the severe cases of COVID-19 we’d expect all at once if we returned to business as usual so that everyone who needs a ventilator would get one is extremely low.