Singapore, renowned for its careful testing, contact tracing and isolation of patients, experienced only 10 deaths out of 4,427 cases through Thursday. That yields a strikingly low case fatality rate of 0.2 percent, about twice the rate of seasonal influenza.
In the United States, the case fatality rate has steadily ticked upward, from about 1.35 percent in late March to over 4 percent on Wednesday, according to figures compiled by the Centers for Disease Control and Prevention. The rate saw another spike to nearly 5 percent Thursday because of a large wave of “probable” deaths reported in New York City.
This does not mean that the disease itself is getting deadlier, though. Covid-19 typically takes weeks to become severe enough to kill a patient, and the rising rate may reflect the disease’s gradual progression, combined with discoveries of additional deaths.
In Michigan, whose official case fatality rate has surged to 7.2 percent, public health experts and elected officials say they need more tests and better data.
“Honestly, we talk about this every day,” said Detroit’s EMS medical director, Robert Dunne. “That’s something we’re all wondering. What’s the actual case fatality rate?”
The testing shortfalls and other weaknesses in public health surveillance have also sparked concerns that there are jurisdictions missing large numbers of covid-19 deaths. In some instances, officials have increased their numbers after redefining what counts as a coronavirus-related death. On Friday, China acknowledged that many people in Wuhan died at home from undiagnosed covid-19 in the early days of the outbreak, and the government raised the Wuhan death toll by 50 percent, from 2,579 to 3,869.
A case fatality rate is the number of deaths divided by the number of confirmed cases. But in this global crisis, both the numerator and the denominator are fuzzy.
“You need to do more testing,” said Teena Chopra, professor of medicine at Wayne State University’s division of infectious diseases. Without testing, she said, public health experts are forced “to live in an unknown world, and an unknown environment.”
In the early days of the outbreak in China, scientists around the world realized that this was a remarkably contagious virus that could lead to a pandemic. But they were cautious about drawing a parallel with the 1918 influenza pandemic.
The two pandemics were caused by completely different viruses, and their deadliness depends on many factors, including the vulnerability of the population. The 1918 pandemic occurred when viral diseases were not well understood, medical interventions remained primitive, and many nations were mired in a terrible, preoccupying war that led to a censoring of news about the contagion.
No one knows exactly how many people died of influenza in 1918; estimates range from 15 million to 100 million globally. Historians estimate that the virus killed about 675,000 people in the United States across three waves of the pandemic. In contrast to this current pandemic, older people in 1918 seemed to carry at least partial immunity to the influenza virus, probably from exposure to pandemic flu earlier in their lives. The median age of a victim in 1918 was just 28, according to Cecile Viboud, an epidemiologist at the Fogarty International Center, part of the National Institutes of Health.
Even with these sharp distinctions, the lethality and contagiousness of the novel coronavirus, along with its ability to spread and disrupt economies planetwide, has now made 1918 the inescapable comparison point for infectious disease experts.
“If, in fact, the case fatality rate is higher than the 1918 flu, then this one has the potential to kill even more people,” said Donald Forthal, an immunologist at the University of California at Irvine. “There’s been nothing like this in my generation. Or my parents’ generation. It was the generation before that that lived through 1918.”
A safe and effective vaccine could be at least a year away. That leaves “mitigation,” such as social distancing, the only currently available tool for fighting the pandemic. A newly published paper in the journal Science argues that some mitigation will probably be necessary until 2022.
Patients often see symptoms come in waves, and sometimes a patient who seems on the road to recovery will take a drastic turn for the worse. There is concern about long-term effects even among those who recover. Covid-19 is categorized as a respiratory disease, but doctors have found that it can affect many organs, including the heart, liver and kidneys.
The “virulence” of the virus — its ability to cause illness — has been steadily coming into focus. The disease is far more likely to cause severe outcomes in older people, with the oldest cohorts the most vulnerable. That said, in every age group — even 85-plus — most people who contract the disease will recover.
Preliminary research indicates that the virus is not mutating significantly as it spreads, and so there is no evidence that some countries are dealing with a more virulent strain of SARS-CoV-2.
What is sharply different from country to country is the demographic profile. Some European countries with relatively elderly populations — Belgium, France, Italy, the Netherlands, Spain and the United Kingdom — are reporting official figures that equate to fatality rates over 10 percent. There are also European countries that are doing far better, such as Norway and Germany, where the figure is in the 2 to 3 percent range.
“Think about Italy and Germany,” said Carlos del Rio, an epidemiologist at Emory University. “They’re pretty close to each other . . . [but] one of the things that is clearly different is the median age of patients in Italy is 63 or 64 years; the median age of patients in Germany is 47. The mortality is much lower [in Germany] because they avoided having the older population affected.”
The other major factor in mortality is chronic disease. Most people hospitalized with severe cases of covid-19 have chronic health conditions such as diabetes, lung disease and heart disease. Where there is a high percentage of noncommunicable diseases like high blood pressure, the coronavirus will also be more deadly.
This is a critical problem for people of color and those living in low-income communities, who have higher rates of these long-term conditions and often have less access to health care.
The fatality rate for African Americans in Detroit, for example, has been significantly higher than for whites in the region.
“We haven’t seen [the infection rate] to be materially different, but the fatality rate is running two to three times higher among blacks than whites,” Detroit Mayor Mike Duggan said.
Duggan — who ran a major hospital system in Detroit before becoming mayor — said the higher fatality rate for blacks is the product of an African American population that suffers from a series of ailments that leave them more vulnerable to covid-19.
“We saw it every day. African Americans have three times the rate of chronic kidney disease that Caucasians have, and 25 percent higher heart disease. They’ve got higher rates of diabetes, hypertension and asthma,” Duggan said. “I fully expect that when people are hit hard and they are on a ventilator to breathe and their body needs to fight the infection, that people who already have compromised hearts or kidneys or lungs are that much more in jeopardy.”
Also critical is the nature, and robustness, of the national health system. For instance, Japan, where the current case fatality rate is 1.6 percent, and Singapore are reporting extremely high rates of hospitalization for coronavirus patients, at 80 percent and higher, figures that are unheard of in the United States. But this probably helps improve treatment and also reduces disease spread by isolating patients. The result is fewer deaths.
Policies clearly count. Several of the countries with low fatality rates — Germany, South Korea, Norway — have very high rates of coronavirus testing. This gave them a better look at the disease within their borders.
These disparities can be seen within the United States. John Balmes, a professor at the University of California at San Francisco School of Medicine who is working with covid-19 patients at Zuckerberg San Francisco General Hospital, noted that the hospital nearly tripled the capacity of its intensive care unit by adding doctors, nurses and technicians while the city adopted social distancing measures shortly ahead of New York.
“We were ready for a surge that never happened,” Balmes said. “They’re every bit as good as we are in intensive care in New York, but the system was overwhelmed. We did physical distancing just a few days earlier than New York, but it was a few days to the good.”
New York state, which has now developed high levels of coronavirus testing, is still reporting an official fatality rate of 5.4 percent.
Early in the outbreak, China’s case fatality rate was 2.3 percent. That was an alarming figure, suggesting that a global pandemic could kill millions of people. Then on March 3 the head of the World Health Organization, Tedros Adhanom Ghebreyesus, said in a news conference that the global case fatality rate was 3.4 percent. That was treated as a revelation about the innate deadliness of the disease, but in fact was simply the WHO’s crude mortality ratio for confirmed covid-19 cases up to that point in time.
President Trump said he believed the 3.4 percent was “a false number,” calling that a “hunch” and saying he’d spoken to many people about it. His view echoed that of top U.S. health officials, who had recently told members of Congress that the fatality rate might be between 0.1 percent and 1 percent.
But more than a month later, the WHO number has gone even higher: On April 16, the WHO showed a global fatality rate of 6.6 percent among confirmed cases.
Any case fatality rate is, in general, an inaccurate number. And yet in the middle of an epidemic, it is hard to do better. With this coronavirus in particular, the inadequate testing means many infections have been missed. Someone who is asymptomatic or has only a mild case is unlikely to seek a test. That makes a virus appear more deadly than it really is.
The size of this asymptomatic cohort could be huge, as some preliminary scientific studies have suggested in recent weeks. A new study from researchers at Stanford, not yet peer-reviewed, looked for coronavirus antibodies in a sample population in Santa Clara County, Calif., and concluded that the actual infection rate in the county by early April was 50 to 85 times greater than the rate of confirmed cases.
“The story of this virus is turning out to be more about its contagiousness and less about its case fatality rate,” said David Rubin, director of PolicyLab at Children’s Hospital of Philadelphia and a University of Pennsylvania professor of pediatrics. “It’s less fatal than we thought, but it’s more contagious.”
Where extensive testing has been done, estimates for the case fatality rate are often below 1 percent, The Post has found, suggesting these countries are getting closer to a rate that takes into account all infections. In Iceland, which has tested over 10 percent of the population, vastly more than other countries, the fatality rate is just 0.5 percent.
Any search for a global case fatality rate would have to mesh such numbers with those from countries with sky-high current fatality rates, typically above 3 percent. Harvard epidemiologist Marc Lipsitch has written that he and most experts suspect the fatality rate is about 1 to 2 percent for symptomatic cases. A 1 percent fatality rate is 10 times the average fatality rate for seasonal flu.
“It’s probably about an order of magnitude higher for covid-19,” said Viboud, the NIH epidemiologist. “It’s more severe in terms of mortality than the pandemics we’ve seen since 1918.”
Greg Jaffe contributed to this report.
Editor’s note: This story has been updated to remove an estimate of the mortality rate from the 1918 flu pandemic, because the figure, although from an academic study, was not consistent with other research studies.