Our Next Steps on Coronavirus
I wrote about the Coronavirus in early March. After following it since mid January, I had been worried about its potential to impact us for over a month. As those close to me could tell you: I was advising people to prepare for a lockdown here, be cautious with travel, work from home, et cetera back in early-to-mid February. Here was my conclusion on March 7th:
Right now, the evidence is certainly pointing to a wakeup call for humanity about just how dangerous new infectious diseases can be. If it isn’t slowed down, we will almost all be touched by this virus personally, financially, and societally. Imagining the realistic downstream consequences of an pandemic like this simply make it too big to ignore.
So please: stay informed, be hygienic, prepare for the disruption of basic services, and avoid large gatherings. Being personally responsible will help slow the virus and allow medical personnel to respond as best they can.
I share that as a preface because without that context, my claim here would appear to some as misplaced, uninformed, or hypocritical. Yet as new evidence has come to light in the form of serological surveys, the answer is clear: it is time to open back up with some basic, common sense precautions in place.
The Evidence
Broadly speaking – it is time to open back up because Covid-19 is more contagious, more widespread, but far less deadly than we initially thought. The evidence for this comes from serological surveys. These tests reveal the presence of antibodies in the bloodstream that indicate whether this person’s body has already seen the coronavirus and responded to it.
These tests in a couple cases now have revealed a 8-10 fold undercount of reported cases vs. actual instances of infection. As noted above, this means the virus is much more contagious than we thought: one of the aforementioned surveys has measured New York City’s true case count at 21% of the population, while only 2% of the city has officially tested positive at this point.
Of course, this also means that the disease is an order of magnitude less deadly than initially thought. Death rates calculated off of confirmed cases have tended to cluster around 3 to 6% on average, so this would mean a proximate death rate from Covid-19 of 0.3% to 0.6%.
On face, this still seems quite high, but we have also clearly seen that any Coronavirus analysis that doesn’t consider age breakdowns paints an incomplete picture.
For this, we turn to Los Angeles County, which also did a serological survey and released the results by age. Importantly, this survey was designed with a random sample picked by a market research firm. While there is still selection bias in terms of who turned up, this mitigates that risk vs. an open survey.
When compared with current LA county numbers, those results show a 12.5 ratio of confirmed vs. total cases. This is similar enough to New York’s numbers to increase our confidence that these measurements are “in the ballpark.” While there are still measurement and bias risks, we can reasonably say that the Coronavirus is much more widespread than confirmed case counts suggest.
As noted though, the LA data does provide breakdown by age. And when cross-referenced with LA’s official case counts, we can construct the following table:
These numbers on their own don’t provide too much context – a comparison point or two is helpful. As it turns out, we could use two: the chances that an American dies in a car accident in a given year or the chances that an American dies from the flu in a given year are each about the same: 0.015%.
Thus, we can construct a new table – for each age bucket the number of “extra years of driving risk” that the Coronavirus poses:
Before we move on to what this means, I do want to acknowledge the multitude of risks and othssible. To wit, there are many reasons this analysis could be invalid:
Serosurveys could be biased upward by selection bias (those who had symptoms more likely to go for serosurvey)
Serosurveys could be biased upward by false positives. At the very least, they should be considered to have sizeable error bars.
Disease likely does pose long-term health detriments to some portion of population that does not die
Additional factors (existing conditions, gender, and more) complicate the risk profile for any individual
LA County and NYC aren’t representative of the rest of the country.
All that said – my read of this is that these numbers are in the right order of magnitude. With two sources providing similar estimates, it is the best evidence we have to work with for the time being.
Two Other Key Factors
Viral Load Matters – this has been written about for over a month, but seems to have been forgotten when it comes to understanding policy implications. Without diving too much into the science, it is well established that the Coronavirus acts similarly to other viruses: the more of the virus you are exposed to, the worse the illness. This would match basic intuition, but is worth considering when we talk about closed vs. open spaces and the benefits of wearing masks.
Sunlight is a Disinfectant – a recent Dept. of Homeland Security report suggests that the virus is rendered inactive by just 3 minutes of sunlight. This would also explain why large sunny states like California and Texas have faired substantially better than the Northeast. This is important for considering outdoor spaces and individual actions.
What This Means
Everything above is backed by a strong standard of evidence (sources linked at the end of this article): published medical work. While there is room for skepticism (as noted) on the broad spread suggested by the serosurveys, nothing above is controversial. And yet, figuring out where we go from here absolutely enters into the realm of political beliefs and landmines.
The state of our political discourse is so toxic, I’m tempted to leave the article here and let you draw your own conclusions. And yet, I do think there are meaningful statements that we need to draw from this that run against both political tribes. So in my ever-running trend toward masochism, here are some logical takeaways from the above analysis:
We cannot contain the virus – a 10x undercount ratio suggests that 11 million people have or have had the virus. This is past the point of containment. We should still reduce risk and slow the spread, but we also can’t act like we can eradicate it until we reach herd immunity or develop and deploy a vaccine to hundreds of millions of people.
Age-tiered Recommendations – our government should be providing three-tiered recommendations to people based on age. The implications for a 20 or 30-year old are *drastically* different than that of a senior citizen. We have to treat these situations entirely differently! While adding 3 years of driving risk for the youngest bucket isn’t ideal, it also isn’t prohibitive. For those under 40 without any risk factors, going out should be viewed analogous to driving: be careful and minimize risk, but don’t avoid the activity all together. Yet for those over 65 – this should be viewed differently: it’s more than a lifetime of driving risk! If you are over 65 and/or have risk factors, this disease can be quite deadly and you should be very cautious to try to avoid it!
Open all outdoor spaces immediately – we have seen certain Governors act like wannabe dictators in this crisis – most notably Gavin Newsom in California, Gretchen Whitmer in Michigan, and Tom Wolf in Pennsylvania. All three have tried to restrict the use of outdoor areas among other overreaches. Given what we know now, this is both ill-advised and counterproductive. The evidence is fast-moving so a slow reaction can be excused, but all three should open up all outdoor spaces ASAP. Along those lines – a BBQ or outdoor meetup of friends or family is quite safe, and no law should restrict consenting adults from doing this.
Indoor public spaces should open cautiously – this path is already being followed in several states, including my home state of Texas: many indoor locations have opened but with limits on how many people can go in. Restaurants are allowed to open at 25% occupancy, grocery stores still require masks to be worn, and things like nail salons are still closed. We can quibble about the details of what is open and how it is open. While those over 65 should still avoid these places, we have a sizable and growing portion of the population who has already had it and have developed antibodies, as well as a bunch of people under 40 who face very limited risk when masked and hygienic. Those people (and anyone else who wants to brave the risks) deserve the freedom to do so.
Wear Masks, Help the Elderly, and Support Local Businesses – I say these more from a personal level. We should not need (nor want) our government to take over anything that should be a personal responsibility. All of these are just that: personal responsibilities. Wearing masks helps slow the spread of the virus and puts social pressure on others to do the same. Helping the elderly – calling them, delivering food to them, etc. helps ease their burden of really needing to shelter-in-place. And supporting local businesses helps to combat the unemployment epidemic that is almost certainly our next big challenge.
The Conclusion: It’s not the conclusion
None of this means we’re at the end of the Coronavirus epidemic. It doesn’t mean the virus isn’t serious or that it’s “just the flu.” Even for the youngest age bracket, this is still a 3x as deadly and much more contagious flu. It doesn’t mean our initial reaction to this was hyperbolic or out of scale with what we knew at the time. Our reaction at the time was appropriate in scale but about a month too late.
But the serology surveys give us a better sense of where we go from here. And the answer sure isn’t pious social distance virtue signaling and calls for a nanny state that dictates every move citizens can and cannot take. The data shows us that stopping this disease is almost certainly a fool’s errand given how contagious it is. And they show us that anyone under 40 should be cautious, but doesn’t need to fear this like the plague. We should all continue to be cautious and careful. We should continue to gather new data and information. And we should continue to scale testing and push forward at 110% toward vaccines and more effective treatments.
But it’s also time to reopen America. To do so will require us to become more nuanced in how we express the risk and how we convey it to people. More importantly, it necessitates a revival of the American spirit and belief that people should be free to make their own choices and their own mistakes.
Sources
As noted: LA county’s data by age provides the basis for the estimated true fatality rates if you contract Covid-19 based on your age:
LA County – Overall Confirmed Cases and Deaths By Age
LA County – Serological Survey Results
New York City (and the state as a whole) provide corroborating evidence that LA’s ~12x undercounted case multiplier is in the correct range (NY State’s number was 8.7x, New York City was right around 10x). Sadly, their serological survey did not provide the data segmented by age.
NYC – Serological Survey Results
NYC – Overall Case Totals by Age
Note – there may be some rounding questions depending on the dates used. In order to get a low-end estimate, I am comparing the serosurvey’s estimates (1-2 weeks old) and comparing to the current confirmed case counts.
To address the other two pieces of evidence we’re factoring in:
The Importance of Viral Load in Severity of Coronavirus
DHS Report on Sunlight as a Disinfectant
Lastly – we use the comparison of driving a car to assess risk of Covid-19. The following links provide the fatality rate per billion miles and American’s average driving distance. Taken together, these tie out with the total number of fatalities from traffic in the US each year: approx. 40,000 deaths. Similarly, this approximates a normal flu season of ~40,000 deaths, which is why we can consider the risk of death by driving in a year roughly equivalent to the risk of dying from the flu.