We Still Don’t Know How the Coronavirus Is Killing Us

Omar Rodriguez organizes bodies in the Gerard J. Neufeld funeral home in Elmhurst on April 22. Photo: Spencer Platt/Getty Images

We’re committed to keeping our readers informed. 
We’ve removed our paywall from essential coronavirus news stories. Become a subscriber to support our journalists. Subscribe now.

Over the last few weeks, the country has managed to stabilize the spread of the coronavirus sufficiently enough to begin debating when and in what ways to “reopen,” and to normalize, against all moral logic, the horrifying and ongoing death toll — thousands of Americans dying each day, in multiples of 9/11 every week now with the virus seemingly “under control.” The death rate is no longer accelerating, but holding steady, which is apparently the point at which an onrushing terror can begin fading into background noise. Meanwhile, the disease itself appears to be shape-shifting before our eyes.

In an acute column published April 13, the New York Times’ Charlie Warzel listed 48 basic questions that remain unanswered about the coronavirus and what must be done to protect ourselves against it, from how deadly it is to how many people caught it and shrugged it off to how long immunity to the disease lasts after infection (if any time at all). “Despite the relentless, heroic work of doctors and scientists around the world,” he wrote, “there’s so much we don’t know.” The 48 questions he listed, he was careful to point out, did not represent a comprehensive list. And those are just the coronavirus’s “known unknowns.”

In the two weeks since, we’ve gotten some clarifying information on at least a handful of Warzel’s queries. In early trials, more patients taking the Trump-hyped hydroxychloroquine died than those who didn’t, and the FDA has now issued a statement warning coronavirus patients and their doctors from using the drug. The World Health Organization got so worried about the much-touted antiviral remdesivir, which received a jolt of publicity (and stock appreciation) a few weeks ago on rumors of positive results, the organization leaked an unpublished, preliminary survey showing no benefit to COVID-19 patients. Globally, studies have consistently found exposure levels to the virus in most populations in the low single digits — meaning dozens of times more people have gotten the coronavirus than have been diagnosed with it, though still just a tiny fraction of the number needed to achieve herd immunity. In particular hot spots, the exposure has been significantly more widespread — one survey in New York City found that 21 percent of residents may have COVID-19 antibodies already, making the city not just the deadliest community in the deadliest country in a world during the deadliest pandemic since AIDS, but also the most infected (and, by corollary, the farthest along to herd immunity). A study in Chelsea, Massachusetts, found an even higher and therefore more encouraging figure: 32 percent of those tested were found to have antibodies, which would mean, at least in that area, the disease was only a fraction as severe as it might’ve seemed at first glance, and that the community as a whole could be as much as halfway along to herd immunity. In most of the rest of the country, the picture of exposure we now have is much more dire, with much more infection almost inevitably to come.

But there is one big question that didn’t even make it onto Warzel’s list that has only gotten more mysterious in the weeks since: How is COVID-19 actually killing us?

We are now almost six months into this pandemic, which began in November in Wuhan, with 50,000 Americans dead and 200,000 more around the world. If each of those deaths is a data point, together they represent a quite large body of evidence from which to form a clear picture of the pandemic threat. Early in the epidemic, the coronavirus was seen as a variant of a familiar family of disease, not a mysterious ailment, however infectious and concerning. But while uncertainties at the population level confuse and frustrate public-health officials, unsure when and in what form to shift gears out of lockdowns, the disease has proved just as mercurial at the clinical level, with doctors revising their understanding of COVID-19’s basic pattern and weaponry — indeed often revising that understanding in different directions at once. The clinical shape of the disease, long presumed to be a relatively predictable respiratory infection, is getting less clear by the week. Lately, it seems, by the day. As Carl Zimmer, probably the country’s most respected science journalist, asked virologists in a tweet last week, “is there any other virus out there that is this weird in terms of its range of symptoms?”

You probably have a sense of the range of common symptoms, and a sense that the range isn’t that weird: fever, dry cough, and shortness of breath have been, since the beginning of the outbreak, the familiar, oft-repeated group of tell-tale signs. But while the CDC does list fever as the top symptom of COVID-19, so confidently that for weeks patients were turned away from testing sites if they didn’t have an elevated temperature, according to the Journal of the American Medical Association, as many as 70 percent of patients sick enough to be admitted to New York State’s largest hospital system did not have a fever.

Over the past few months, Boston’s Brigham and Women’s Hospital has been compiling and revising, in real time, treatment guidelines for COVID-19 which have become a trusted clearinghouse of best-practices information for doctors throughout the country. According to those guidelines, as few as 44 percent of coronavirus patients presented with a fever (though, in their meta-analysis, the uncertainty is quite high, with a range of 44 to 94 percent). Cough is more common, according to Brigham and Women’s, with between 68 percent and 83 percent of patients presenting with some cough — though that means as many as three in ten sick enough to be hospitalized won’t be coughing. As for shortness of breath, the Brigham and Women’s estimate runs as low as 11 percent. The high end is only 40 percent, which would still mean that more patients hospitalized for COVID-19 do not have shortness of breath than do. At the low end of that range, shortness of breath would be roughly as common among COVID-19 patients as confusion (9 percent), headache (8 to 14 percent), and nausea and diarrhea (3 to 17 percent). That the ranges are so wide themselves tells you that the disease is presenting in very different ways in different hospitals and different populations of different patients — leading, for instance, some doctors and scientists to theorize the virus might be attacking the immune system like HIV does, with many others finding the disease is triggering something like the opposite response, an overwhelming overreaction of the immune system called a “cytokine storm.”

The most bedeviling confusion has arisen around the relationship of the disease to breathing, lung function, and oxygenation levels in the blood — typically, for a respiratory illness, a quite predictable relationship. But for weeks now, front-line doctors have been expressing confusion that so many coronavirus patients were registering lethally low blood-oxygenation levels while still appearing, by almost any vernacular measure, pretty okay. It’s one reason they’ve begun rethinking the initial clinical focus on ventilators, which are generally recommended when patients oxygenation falls below a certain level, but seemed, after a few weeks, of unclear benefit to COVID-19 patients, who may have done better, doctors began to suggest, on lesser or different forms of oxygen support. For a while, ventilators were seen so much as the essential tool in treating life-threatening coronavirus that shortages (and the president’s unwillingness to invoke the Defense Production Act to manufacture them quickly) became a scandal. But by one measure 88 percent of New York patients put on ventilators, for whom an outcome as known, had died. In China, the figure was 86 percent.

On April 20 in the New York Times, an ER doctor named Richard Levitan who had been volunteering at Bellevue proposed that the phenomenon of seemingly stable patients registering lethally low oxygen levels might be explained by “silent hypoxia” — the air sacs in the lung collapsing, not getting stiff or heavy with fluid, as is the case with the pneumonias doctors had been using as models in their treatment of COVID-19. But whether this explanation is universal, limited to the patients at Bellevue, or somewhere in between is not yet entirely clear. A couple of days later, in a pre-print paper others questioned, scientists reported finding that the ability of the disease to mutate has been “vastly underestimated” — investigating the disease as it appeared in just 11 patients, they said they found 30 mutations. “The most aggressive strains could generate 270 times as much viral load as the weakest type,” the South China Morning-Post reported. “These strains also killed the cells the fastest.”

That same day, the Washington Post reported on another theory gaining traction among American doctors treating the disease — that one key could be the way COVID-19 affects the blood of patients, producing much more clotting. “Autopsies have shown that some people’s lungs are filled with hundreds of microclots,” the Post reported. “Errant blood clots of a larger size can break off and travel to the brain or heart, causing a stroke or a heart attack.”

But the bigger-picture perspective the newspaper offered is perhaps more eye-opening and to the point:

One month ago, as the country went into lockdown to prepare for the first wave of coronavirus cases, many doctors felt confident that they knew what they were dealing with. Based on early reports, covid-19 appeared to be a standard variety respiratory virus, albeit a very contagious and lethal one with no vaccine and no treatment. But they’ve since become increasingly convinced that covid-19 attacks not only the lungs, but also the kidneys, heart, intestines, liver and brain.

That is a dizzying list. But it is not even comprehensive. In a fantastic survey published April 17 (“How does coronavirus kill? Clinicians trace a ferocious rampage through the body, from brain to toes,” by Meredith Wadman, Jennifer Couzin-Frankel, Jocelyn Kaiser, and Catherine Matacic), Science magazine took a thorough, detailed tour of the ever-evolving state of understanding of the disease. “Despite the more than 1,000 papers now spilling into journals and onto preprint servers every week,” Science concluded, “a clear picture is elusive, as the virus acts like no pathogen humanity has ever seen.”

In a single illuminating chart, Science lists the following organs as being vulnerable to COVID-19: brain, eyes, nose, lungs, heart, blood vessels, livers, kidneys, intestines. That is to say, nearly every organ:

And the disparate impacts were significant ones: Heart damage was discovered in 20 percent of patients hospitalized in Wuhan, where 44 percent of those in ICU exhibited arrhythmias; 38 percent of Dutch ICU patients had irregular blood clotting; 27 percent of Wuhan patients had kidney failure, with many more showing signs of kidney damage; half of Chinese patients showed signs of liver damage; and, depending on the study, between 20 percent and 50 percent of patients had diarrhea.

On April 15, the Washington Post reported that, in New York and Wuhan, between 14 and 30 percent of ICU patients had lost kidney function, requiring dialysis. New York hospitals were treating so much kidney failure “they need more personnel who can perform dialysis and have issued an urgent call for volunteers from other parts of the country. They also are running dangerously short of the sterile fluids used to deliver that therapy.” The result, the Post said, was rationed care: patients needing 24-hour support getting considerably less. On Saturday, the paper reported that “[y]oung and middle-aged people, barely sick with COVID-19, are dying from strokes.” Many of the patients described didn’t even know they were sick:

The patient’s chart appeared unremarkable at first glance. He took no medications and had no history of chronic conditions. He had been feeling fine, hanging out at home during the lockdown like the rest of the country, when suddenly, he had trouble talking and moving the right side of his body. Imaging showed a large blockage on the left side of his head. Oxley gasped when he got to the patient’s age and covid-19 status: 44, positive.

The man was among several recent stroke patients in their 30s to 40s who were all infected with the coronavirus. The median age for that type of severe stroke is 74.

But the patient’s age wasn’t the only abnormality of the case:

As Oxley, an interventional neurologist, began the procedure to remove the clot, he observed something he had never seen before. On the monitors, the brain typically shows up as a tangle of black squiggles — “like a can of spaghetti,” he said — that provide a map of blood vessels. A clot shows up as a blank spot. As he used a needlelike device to pull out the clot, he saw new clots forming in real-time around it.

“This is crazy,” he remembers telling his boss.

These strokes, several doctors who spoke to the Post theorized, could explain the high number of patients dying at home — four times the usual rate in New York, many or most of them, perhaps, dying quite suddenly. According to the Brigham and Women’s guidelines, only 53 percent of COVID-19 patients have died from respiratory failure alone.

It’s not unheard of, of course, for a disease to express itself in complicated or hard-to-parse ways, attacking or undermining the functioning of a variety of organs. And it’s common, as researchers and doctors scramble to map the shape of a new disease, for their understanding to evolve quite quickly. But the degree to which doctors and scientists are, still, feeling their way, as though blindfolded, toward a true picture of the disease cautions against any sense that things have stabilized, given that our knowledge of the disease hasn’t even stabilized. Perhaps more importantly, it’s a reminder that the coronavirus pandemic is not just a public-health crisis but a scientific one as well. And that as deep as it may feel we are into the coronavirus, with tens of thousands dead and literally billions in precautionary lockdown, we are still in the very early stages, when each new finding seems as likely to cloud or complicate our understanding of the coronavirus as it is to clarify it. Instead, confidence gives way to uncertainty.

In the space of a few months, we’ve gone from thinking there was no “asymptomatic transmission” to believing it accounts for perhaps half or more of all cases, from thinking the young were invulnerable to thinking they were just somewhat less vulnerable, from believing masks were unnecessary to requiring their use at all times outside the house, from panicking about ventilator shortages to deploying pregnancy massage pillows instead. Six months since patient zero, we still have no drugs proven to even help treat the disease. Almost certainly, we are past the “Rare Cancer Seen in 41 Homosexuals” stage of this pandemic. But how far past?

We Still Don’t Know How the Coronavirus Is Killing Us