With a president more concerned about downplaying the risks of COVID-19 than protecting citizens from it, and with swaths of the national media operating at a level that is outright dysfunctional, it's difficult at best for us to assess how worried we should be about the emerging coronavirus pandemic.
After the director of the World Health Organization recently put the estimated mortality rate of the virus at 3.4 percent, President Trump predictably quarreled with the assessment. In an interview with Sean Hannity of Fox News, Trump called the figure "a false number," and proclaimed that the actual fatality rate was "a fraction of 1 percent" — all based on what he called a "hunch." Meanwhile, NBC's Chuck Todd said recently on his daily MSNBC show that he had reassured his teenage son that "only 2 percent of those who get this coronavirus die."
Todd's chatty dismissal failed to note that a 2 percent mortality rate could well translate into the worldwide deaths of tens of millions of people. If Harvard epidemiologist Marc Lipsitch is correct in his prediction that COVID-19 is likely to infect between 40 and 70 percent of the world's population this year alone, a 2 percent mortality rate would result in between 62 million and 108 million deaths. After all, two percent was the mortality rate of the 1918 Spanish flu, which killed tens of millions of people — long before the era of mass transit, globalized trade, and widespread travel by planes or cruise ships.
But even if we could trust all the information we are receiving, we still wouldn't really know how deadly this disease is. It's just too soon. The mortality rate of the pathogen — how many of the people who contract it will go on to die — is still far from clear, despite the fact that seemingly official numbers are being released on a near-daily basis. Those numbers aren't "wrong." They more-or-less accurately tell us how many people whom we know have been diagnosed have gone on to die. But that only tells us so much.
By some estimates, the mortality rate of COVID-19 could even exceed that of the WHO's estimate. At first glance, the deaths of 4,262 people out of 118,100 confirmed global cases (3.6 percent) would seem to support the health organization's dire assessment. But as an interactive map developed at Johns Hopkins University makes clear, only 64,391 of all the patients who have contracted the disease are believed to have yet recovered, making it too soon assess the mortality rate for the 45 percent of global citizens who have contracted the disease but neither died nor recovered. If one divides the global death rate by the 68,653 patients who have either died or recovered, that crude assessment suggests a death rate approximating 6.2 percent.
While the numbers coming out of organizations such as the WHO and the Centers for Disease Control are meaningful to professionals who know how to interpret them, Dr. Arthur Reingold, division head of epidemiology and biostatistics at U.C. Berkeley's School of Public Health, cautions that "extrapolating from them can be dangerous" for everyday people. "We're still learning," he said.
In the early stages of an epidemic or pandemic, epidemiologists simply can't accurately measure mortality rates, Reingold said. For one thing, we don't really know how many people have it. For another, we don't know how many of the people who already have it will get over it, and how many will go on to die.
"The problem with making these calculations is: who's in the numerator and who's in the denominator," Reingold said. So far, that's unknown.
Most often in the case of new pandemics — but not always — the initial reported mortality rates start to head down as the diseases progress and more people get tested. In the United States, the CDC has slow-rolled and otherwise mismanaged the introduction of widespread testing to the point where estimates of total U.S. cases aren't necessarily trustworthy. For instance, according to The New York Times, as of Monday only 45 of the 2,400 passengers of the Grand Princess cruise liner docked at the Port of Oakland had yet been tested for the virus, despite the presence of at 21 people already known to be infected.
Most mainstream media outlets have been more responsible than Todd, but even there, and especially on TV, they often aren't providing the proper context for how we're supposed to interpret the numbers. "Often, such figures are presented with a certainty that belies what they can't tell us," wrote Jon Allsop in a critique of media coverage for the Columbia Journalism Review. For all the vaunting of "data journalism" over the past several years, a dire, fast-changing situation like this tends to reveal its limitations. "People trust data," science journalist Alexis Madrigal wrote in his examination of mortality rates for The Atlantic. "Numbers seem real. Charts have charismatic power. People believe what can be quantified. But data do not always accurately reflect the state of the world."
That certainly seems to be the case here, for the moment. "The situation is fluid," Reingold said. Besides the fact that we don't know the true numerator or denominator behind the mortality rate, he notes that the virus itself is changing. It might mutate to become less lethal — or much more so. The best course is to follow the directions of public-health officials — social distancing, frequent handwashing, etc. — and not worry too much about how bad the virus is until we learn more. "Give everyone time to either recover or die," Reingold said." Then we'll know."