This post appeared is written by Daniel Hanson and appeared originally at: https://www.facebook.com/photo.php?fbid=10222159161191103&set=a.2392855263881&type=3
It’s the Monday between Palm Sunday and Easter, a moment for reflection on life and death. Here’s a 10 minute quick summary of some salient corona news.
The curve is bending and flattening. Social distancing is working, and as a consequence, we’ve gone from doubling the number of deaths in the US every three days to doubling them every four. That’s great news, and it should reduce the total number of deaths from this disease meaningfully. Still, more than 10,000 Americans have died from this disease as of this morning.
Hospitalization is a leading indicator; people tend to be hospitalized before they die from the disease. Typically, death from coronavirus comes about 16 days after being admitted to the hospital, and in this respect, it seems that New York has turned the corner. NY had fewer deaths yesterday than it has since mid-March, but that shouldn’t be the headline. The real story is new hospitalizations, which dropped from 2,879 on April 3 to 1,069 yesterday. Such a drop is a seismic change in the so-called second derivative of this disease, a dramatic leading indicator showing progress in prevention.
New York isn’t alone in bending the curve. Georgia is doing better, and so is California. Washington State has improved so much that it’s actually shipping ventilators to New York to help. There are, unfortunately, hot spots that are getting worse – including Philly, Miami, New Orleans, and Detroit – so this week will likely see another 10,000+ deaths from covid-19.
Perhaps the most important development of the last week, however, is that the uncertainty around covid-19 is going away. There are essentially eight reasons public health officials have given for why such dramatic action has been necessary in this pandemic; we’ve made progress on seven of them. Let’s look at each:
--Scientific mystery of the virus: Notable progress. We didn’t know how the virus reproduced, what its spike coat looked like, how transmission occurred, how symptoms would show, or the rate at which asymptomatic people were infected. Each day, we get more data on these questions, and we’re finally starting to see reliable vector models that can help us target the disease’s problems better.
--High mortality rate: No progress. The disease is a killer; it’s something like 10 times more deadly than the seasonal flu and we have no real treatment regimen to fight it.
--Public nonchalance: Solved. Initially, neither the public nor policymakers were taking the virus seriously and therefore were not engaging in proper hygiene or social distancing habits. This caused some of the massive spread in urban centers, nursing homes, and college parties. Now, the public seems to grasp the seriousness of the threat and is taking appropriate precautions.
--Limited testing capacity: Limited progress. We had virtually no tests five weeks ago; now we can process more than 100,000 tests per day. We need tests to become even more widely available, and eventually we will need to scale up serological testing capacity, but from a national perspective, we’re on the right path. Roche, Abbott, and Cepheid all control their supply chains end to end, so the low commodity component shortages in tests, like swabs, are becoming more widely available as the companies scale down flu testing and scale up covid testing.
--PPE Shortfalls: Notable progress. We didn’t have enough masks, gowns, and gloves to protect workers at the outbreak, and while there are still shortages, capacity is growing quickly to provide front line workers and the general public with appropriate equipment.
--Lack of hospital capacity: Notable progress. Hospitals could not and still cannot handle a giant surge in cases, and one of the chief concerns was that hospitals would be totally overwhelmed with cases. A lack of sufficient ventilator capacity made this problem worse. Hospitals are strained, especially in urban centers, but nationwide capacity is not a problem at this moment. Makeshift hospitals in urban areas have helped relieve strain, and major steps have been taken to increase ventilator access.
--Contact tracing: Limited progress. Because health departments were not equipped to trace where people who tested positive had been, they were forced to keep millions of people at home. Rural health departments are now doing a decent job of contact testing and tracing, though urban areas are continuing to struggle. Target quarantines will replace full lockdowns as more contact tracing is done.
--Therapeutics and vaccines: Limited progress. More than 70 therapeutics are being examined and two vaccine candidates are beginning trials. Vaccines are still 18+ months away, so we’ll need to find other ways to move past the present moment.
The reasons outlined above highlight a key paradox of fighting the virus: the more progress we make in containing it, the less likely Americans are to tolerate measures that contain the virus because they will not directly observe the benefits of these actions. It’s a huge victory that NY hospitals didn’t become totally overwhelmed, and it kept thousands of people from dying. But because those people didn’t die, or even get sick, or because they were concentrated in NY, the 295 million Americans currently under stay-at-home orders are likely asking if all this trouble is worth it.
Anecdotes are not data, but talk to a person who has been hospitalized with covid-19, and you start to quickly think the pain is worth it. An anesthesiologist I talked to yesterday had a fever for 11 days, lost 26 pounds, and was in the ICU for 17 days. He’s a marathon-running doctor in his 30s who eats well, has no known medical conditions, and has two young kids. Another friend spent 22 days in the hospital with a fever that didn’t drop below 102 degrees for 16 of those days. He was put on a ventilator on day 11, and he’s 22 years old and works out five days a week. This is not the seasonal flu.
While anecdotes are not data, data are. A CDC MMWR last week reported on pre-existing conditions in people with covid-19, looking at nearly 75,000 cases. The narrative around covid-19 – which was reinforced by media reports on the CDC data – is that the people who die from the disease are super old and very sick, and that in places like Italy, they’re overweight, 50 year chain smokers who eat poorly. This narrative is wrong. It is true that about 40 percent of the cases had one co-morbidity factor, like being overweight or having diabetes or lung disease, and that you were far more likely to be hospitalized or die if you had one or more co-morbidity factors. But a third of the remainder (20 percent of the total hospitalizations) were in totally healthy people, and, much more importantly, 70 percent of Americans have at least one major co-morbidity factor like diabetes, heart disease, lung disease, or cancer. This puts a huge number of Americans in a place of pretty serious risk. (See: https://bit.ly/2Rf66hv).
Whatever one thinks of the disease, it seems clear that the public is becoming antsy with lock downs. The economic carnage of lock downs is serious. A 2018 study found that a 10 point rise in the unemployment rate lowered life expectancy by a year and half. Life expectancy in the US is 78.6 years and declined by a 0.1 percent during the last recession, then declined another 0.1 percent in the intervening decade, mostly due to addiction and suicide. In actuarial terms, a 1.5 year decrease in life expectancy in the US would mean 200,000 more deaths per year, which, weighed against the current covid-19 death figures, looks incredibly bad.
It shouldn’t come as a shock that economic insecurity and the psychological scarring of unemployment and quarantine comes with a hefty price tag. It’s incredibly important to remember, though, that if you are able to work from home – as an employee or a student – you’re incredibly lucky; if you’re able to quarantine, you’re rich. About 33 million Americans with homes don’t have broadband; another half million people are homeless altogether. Nearly 15 percent of America gets their Wi-Fi from McDonald’s, Starbucks, or their library – all of which are presently unavailable to them – and don’t have data access on their phones. An enormous amount of research, by the way, shows that in normal times McDonald’s serves as the most functional community center in many impoverished areas of the country, both urban and rural.
About 52 million Americans live in a three generation home, where kids, working age parents, and elderly relatives live together; the average size of such a home is just under 1,500 square feet, and across dozens of indicators, every sign shows that people living in such an arrangement tend to be significantly less happy than other Americans even in normal times. Presumably this reality is worsened when locked together without any break.
Nearly 50 million Americans don’t have access to laundry facilities in their building, to say nothing of their house or apartment, so they must travel out to a laundromat to clean their clothes. One in six kids under the age of 10 has a diagnosed mental or developmental disorder, a 750 percent rise since 2005 fueled in no small part by the explosion of drug use by their parents and grandparents. 15 percent of American homes don’t have AC (though some don’t need it), and about 96 million Americans live in a house with no laptop or tablet. It is hard to imagine these sorts of situations staying locked down long.
Furthermore, there is growing evidence that rural America has strong resentment for urban America in this pandemic. It is perhaps an ugly reality, and I don’t mean to paint with too broad a brush, but wealthy New Yorkers decamping for their vacation homes in rural areas have brought covid-19 with them, sucked resources from the rural communities, and – rightly or wrongly – driven the perception that dirty, arrogant city folk are responsible for this economic suffering. Two out of every three people surveyed in Florida, for instance, believes the disease was transported to Florida by New Yorkers, despite fairly substantial evidence that the clusters in Florida were seeded in early February by cruise travelers.
There are two parts to thwarting a breakdown in social cohesion. One part involves economic policy, and the other part involves health policy.
At 11am, Mike Pence will host a teleconference with governors, then he’ll have lunch with President Trump. After their lunch, Pence will oversee a coronavirus task force meeting, and at 5pm, the White House will hold a briefing on the pandemic. Congress is out of town until April 20, but if you’ve been paying attention in the last 24 hours, you know that politics underwent a seismic shift. Previously, Congress had been talking abstractly about a big fourth spending package, to include things like infrastructure spending and aid to state governments, that would come in the summer. This talk is on ice as policymakers have realized the extent of the economic carnage being unleashed; we’re now on to “Stimulus 3.5.”
Stimulus 3.5 looks a lot like Stimulus 3.0. It’ll probably have more money for hospitals, a further extension of the “enhanced” $600 per week unemployment benefit beyond four months, at least one (but maybe two) more direct payments to Americans over the $1,200 coming later this month, more small business lending (which will probably be exhausted this week), more forgivable paycheck protection grants with a longer time horizon (doubled to 16 weeks of coverage), a bailout program for medium businesses (with 500 to 1,000 employees), election assistance to allow the primary season to be finished remotely and to enhance security around the general cycle, hazard pay bumps for frontline responders, and more. The next bill, which has buy-in from Pelosi, McConnell, and Trump, will clock in at no less than $1.5 trillion and will probably be done in early May.
The deteriorating fiscal picture at the state level is especially important here. Most state and local budget years begin on July 1, and many jurisdictions require balanced budgets. States run with low cash flow in March because they get a big bump in cash from income taxes in April and from sales taxes in the late spring and early summer. Right now, with tax filing deadlines moved to July, unemployment skyrocketing, and retail sales cratering, many states are just flatly out of cash, and they won’t have cash before the end of their fiscal years on June 30 absent intervention from the federal government. One radical-but-very-realistic policy scenario would involve a sort of debt jubilee in which the US federal government buys up state level municipal debt and forgives those loans, effectively shifting the burden of that debt to the national debt from the states and stabilizing the municipal market in the process. Anecdotally, New York’s brand-new budget is probably already $10 billion in deficit despite being just six days old and despite a constitutional balanced budget requirement.
On the public health side, there’s a lot of friction between the federal government and state governments. Governors are empowered with substantial authority to impose quarantine and commandeer public health resources; the federal government is not. One shouldn’t expect a GOP White House to openly criticize GOP governors, but there’s palpable tension between the Trump administration and certain governors -- like Larry Hogan in MD, Charlie Baker in MA, and Greg Abbott in TX. Republican Brian Kemp, for instance, came out three days ago – 74 days after the first US case – to explain the state had held off on a shelter in place order because he didn’t know asymptomatic people could transmit the virus, blaming the CDC for making that message confusing.
The crisis has also elevated certain Democratic public officials, from Gretchen Whitmer in MI to J.B. Pritzker in IL to Andrew Cuomo in NY, each of whom could now serve as a great surrogate for Democrats in the 2020 cycle. Jay Inslee, once popular, was skittish in his response for weeks, and that cost people their lives in Washington. London Breed, the mayor of San Francisco, made a huge name for herself on the national stage by implementing the first major shelter-in-place order and organizing the city’s hospitals to respond; she is surely now on Biden’s short list for VP.
But the meat-and-potatoes of a public health response centers on therapeutic medications and vaccines. They’re the future with this disease, and at the moment, there’s a lot of promise but no fulfillment. Hydroxychloroquine is the most controversial of the drugs, and political fighting both inside and outside the White House has sown confusion and discord. On Saturday, a heated Situation Room showdown between Peter Navarro, an economist, and Anthony Fauci, the nation’s top infectious disease doctor, centered on the drug, with Navarro arguing the drug had “clear therapeutic efficacy” while Fauci cautioned against drawing conclusions that were overly broad.
One FDA official told me yesterday that nearly every half hour for the last month, she’s received a question rooted in a new chloroquine anecdote. In my conversation with Scott Gottlieb yesterday, he was sharply negative on the drug and believes it’s so popular largely because it’s available. More than 190 million doses of chloroquine were administered to corona patients in the second half of March, despite little rigorous clinical data demonstrating efficacy.
Rather than speculating on the anecdotes, we ought to be rigorous about what we know. Here’s everything that we know.
First, a poll of about 6,200 doctors in 30 countries conducted across three days in late March showed that a third of doctors had prescribed some chloroquine formulation, which was by far the most prescribed drug. 27 percent of doctors had prescribed nothing, however, and while 37 percent of doctors reported that they believed chloroquine was effective at reducing symptoms, 32 percent of doctors report that it had no effect at all. These results imply that doctors are seeing little difference between prescribing the drug and not prescribing anything, despite regularly prescribing chloroquine.
Second, data on the drug is incomplete. There are 62 treatment trials underway, including 40 that test hydroxychloroquine and 22 that test regular chloroquine phosphate. Of these studies, 18 are randomized, double blind studies – the gold standard in medical research – while 37 are randomized, open-label trials, which should give some useful data but limit the extent of conclusions that can be drawn. Seven are single-arm, open-label studies, which are of only minor use. Five of the studies are being conducted in the US, with the largest and most important being run in New Jersey; results from that study should be available at the end of the third week of April. 19 studies are in the EU and 20 are in China; four randomized, double blind Chinese studies will have data before the end of April. There are also 12 studies ongoing on whether chloroquine is a useful prophylaxis. 6 are based in the US, and 6 are elsewhere; 12 are testing hydroxychloroquine and two are also testing chloroquine phosphate. The first and best data is expected from the University of Minnesota between April and August.
There are six studies that form the basis of the thesis around chloroquine. None are peer reviewed, though five have been submitted for pre-print, meaning they are published and are under peer-review in the interim. One – the Korea Study – has had the data retracted as the researchers reexamine their conclusions. I’ve included a chart here for your reference on the five studies, but the results are not especially convincing. There are serious disagreements about what dosage of the drug is appropriate, and no conclusion on whether there’s any impact on the viral load, though there’s anecdotal evidence (ahem, not data) that viral shedding (reproduction and release of new strands of virus) is lessened by the drug. There seems to be some baseline consensus that clinical symptoms are lessened by administering the drug, allowing the body to fight off infection more easily, but only when the drug is administered as a relatively early intervention (ie, inside the first 6 days).
There is some evidence from three studies that chloroquine, which distributes rapidly to the lungs, can meaningfully improve pneumonia symptoms. Each of the six studies has fairly serious criticisms of its data, though each of the six also has been subject to fairly severe ad hominem attacks. There is no study that explains the mechanism by which chloroquine works on the system to help in the fight against covid-19. Chloroquine has previously been proposed as a therapeutic for SARS, Ebola, and chikungunya fever; in all three cases, initial clinical data were later proven to be outliers and the drug was not approved to treat these diseases.
Chloroquine is relatively safe; it has been on the market since 1934 and costs about $0.04 per dose to produce. It is approved for the treatment and prevention of malaria, lupus, and rheumatoid arthritis. It takes nearly 1.5 months to totally clear your system, and there are some serious risks to taking it. First, since it is metabolized through the liver and kidneys, people with impaired liver or kidney function, including those over age 65, should not take it. It can cause serious digestive issues, including cramping, nausea, and vomiting, and may cause bleeding problems, shortness of breath, blurred vision, itchiness, deafness, or irreversible heart failure. There are serious interactions between it and other common drugs, like cyclosporine, mefloquine, and various antacids, and it can be extremely harmful to long-term vision. That said, billions of doses of the drug have been administered with relatively minor problems over its long history.
So, should chloroquine be used? Probably, for prophylaxis for medical workers and for severe cases on a compassionate use basis. But it’s not a magic bullet, and there’s little to suggest right now that we’ll suddenly discover that it’s the “right” treatment for the disease. Doctors need to weigh the harm they can do with the drug against the theoretical benefits, which may not be all that great.
On the vaccine front, there’s no easy path forward. We’ve never developed an mRNA vaccine before, and we’ve never developed a vaccine for any coronavirus, so there’s just not a lot to work from. Trevor Bedford, a computational biologist at the Fred Hutchinson Center in Washington, has demonstrated two pivotal facts. First, the spike protein that helps the virus bind to other healthy cells is very stable, so it’s a good candidate to target as a means of breaking the chain of spread. Second, the virus isn’t mutating quickly, and it’s essentially true that the same virus is infecting people in China, Italy, New York, and California. These are two breakthrough discoveries that put a handful of experimental drugs in play, including the five clinical trials underway from Johnson and Johnson, GlaxoSmithKline, Merck, Regeneron, and Moderna.
I have erroneously heard many comparisons to the H1N1 outbreak in 2009 throughout this crisis. There are two super important vaccine-related facts that made that outbreak different than the covid-19 outbreak. First, many older Americans – about 40 percent of the population – had been exposed to the virus in prior outbreaks, and a number of these people had pre-existing immunity when the ’09 outbreak began. Second, we got super lucky on timing. The first US case was identified on April 15, 2009; six days later, the CDC has released a test for the disease nationally. Vaccine manufacturers had just finished manufacturing the seasonal flu vaccine, and H1N1 was not virulent to birds, so fertilization of chicken eggs was quickly scalable using the existing flu vaccine infrastructure, which in turn allowed us to begin harvesting antigens from the eggs inside three months. A vaccine was approved in September, another two in October, and four more in November. By December, more than half the country was inoculated; if you were an adult then, you got two flu shots that year – a trivalent “normal” flu shot and a monovalent H1N1 shot. In this outbreak, we have neither of those two advantages, and so the timeline will be much more like 18 months to two years.
A vaccine comes with stacked risk and stacked uncertainty. There’s a theoretical concern among some microbiologists (that I candidly don't understand) that says an mRNA vaccine would actually increase your risk of getting a serious coronavirus. Setting this aside, it’s simply true that we’re using a novel platform to build this vaccine, which means we don’t fully understand every aspect of such a treatment. The nightmare scenario for public health would be rolling out a vaccine with a massive unintended consequence, prompting push back against both this vaccine specifically and vaccines generally and undermining public health more broadly.
In the meantime, we can start getting people back to work without a vaccine as we develop targeted surveillance. There are privacy concerns here, of course, but the idea would be to identify ways we can protect each other without locking everyone down. GPS tracking probably isn’t a great idea, but the technology exists to implement some tracking of patients who exhibit severe symptoms. We do this on a small scale now with the website FluNearYou.org, where users voluntarily input their symptoms on a regular basis to help track the spread of flu across the country. As with most successful digital programs, FluNearYou works because it gives people something useful that they want (local information on the flu) in exchange for volunteering their personal data. Presumably something similar could exist for covid-19.
The CDC would like you to wear a mask if you go outside. I was in the grocery store at 5am this morning, and there were about 60 other customers there, all of them over the age of 65. I was not wearing a mask; every single other person was. Don’t be me. Wear a face covering, whether a disposable mask or a fabric mask that you can regularly clean. Cover your mouth and nose to protect yourself and others. Here’s a useful guide on how to put on, wear, and take off masks properly: https://bit.ly/39Gdrgs.
Hospitals are essentially closed to anything but corona treatments and severe emergencies; elective emergencies have been put off. This is going to be a problem for a long time for two reasons. First, hospitals are spending tons of money retooling for coronavirus, and it’s going to be expensive to go back to normal service. Second, hospitals and their staff are going to have to be disinfected after cases trail off so that these facilities don’t become epicenters of spread like they did in Italy. Hospitalizations will probably continue to increase nationally into mid-May, which means elective procedures probably won’t be back in America into July or later.
I’ve already written 4,100 words here, so I’ll say very little about the economy, except to make three comments. First, it seems like the recession has a long way to go in terms of getting worse. In every recession since 1950, corporate earnings didn’t recover for at least 3 years after the recession start. We don’t know if this recession will be long, but we do know it will be deeper than most recessions. Economists keep talking about a V-shaped recovery, which implies the economy rebounds vigorously after hitting the bottom; this has never happened in the West in the post-war period. It is more likely – and more consistent with emerging data from, among other places, Asia – that businesses and consumers come out of the crisis with lasting fear, which in turn leads to a relatively protracted (18 month?) recession and sustained, elevated unemployment. The S&P is trading at 2,600 this morning; that’s 14 percent overvalued if you use the average recession effect since World War II to discount free cash flow relative to equity price (ie, the level should be something like 2,200).
Second, industries are going to be changed forever, and that implies they’re going to hire and keep fewer employees than they did before the crisis. I spoke yesterday with my old friend Jeff Shell, who’s the CEO of NBC Universal, and he (who has personally had covid-19) pointed out to me that NBC Universal had 64,000 furloughed employees and 3,000 employees coming into the office. Virtually everyone coming in is integral to either news broadcasts or keeping signals online. Comcast, their parent, serves 40 percent of the country their broadband, and the network hasn’t crashed yet despite not really having employees in the office. People aren’t going to keep paying for content – especially sports – that isn’t fresh, so revenue is going to crater. Right now, that’s a little OK because NBC, which spends $18 billion a year producing TV and movies, has cut 2020 spending to $1 billion. Eventually, however, people are going to demand new content, and recession is bad for everyone. Even in March, there was a surge in cable sign-ups, but 10 million of Comcast’s 30 million autopay customers turned off autopay, implying they’re concerned about their ability to keep paying for broadband. This company will employ fewer people and spend less money on content on the other side of this crisis, and it will take a long time to scale the business back up. I suspect that, as a CEO, Jeff’s view of the world is extremely common.
Finally, there seems to be evidence that the enhanced unemployment insurance is already both inflationary and a disincentive to work. Exhibit One in this regard is American trucking, where demand has gone up as demand for goods, especially food and cleaning products, has risen markedly and inventories have needed to be replenished. April trucking rates are the highest they’ve been since 1950, and capacity utilization is higher than it typically is in the early fall, when trucks are full of holiday-related goods. Despite this, truckers don’t want work, which is very understandable but also bad for the national economy. The Hub Group, JB Hunt, CH Robinson, and Old Dominion Freight Lines have all reported that more than a third of their drivers don’t want to drive for a variety of reasons, and all four companies – which comprise about half the US trucking industry – have reserved cash for raises for truck drivers who are unwilling to work.
Perhaps truckers should make more money; the average trucker in the US earns just over $42,000 per year. But right now, unemployment pays more than median trucking income in all 50 states, and the cost of higher trucking wages will be passed on to end-users in the form of higher prices for goods shipped by truck. If the average trucker needs a 23 percent wage hike to keep trucking (which would bring median wage to the level of unemployment), the average US family’s grocery bill would rise by about $7 per week or $364 per year.
Nevertheless, things are going better. This week will see many US deaths and continued straining of US social fabric, but it won’t break us. We are bending the curve and making progress towards fighting the specter of covid-19. Our heroic front line workers are doing the Lord’s work treating patients in major hot spots, and with a little luck and a little providence, we may avoid the more dire death scenarios and giant outbreaks in loads of other cities not currently under siege.
This nation has the heart and soul of a lion. And like the old lion with her cubs at her side, we are steeled against the onslaught of a hunter armed with a deadly and destructive weapon. Tragedy lurks on the horizon, but tragedy is not now – nor will it ever be – the end of our tale. The stars in their course proclaim the deliverance that is ours, and not so easily shall our heart and soul be tamed. Not so easily shall the clarion call of community be silenced; not so gently shall our determination, our resolve, and our sobriety be snuffed out.
Time is short, but this time is ours. We must do the most we can with it.
PS – there’s a supermoon on April 7-8, as the moon will be the closest in its elliptical orbit on Tuesday moonrise and Wednesday moonset. Due to lockdown, air pollution has dropped markedly in most of the Western world, so you should be able to get a great shot of it.