If 100 people become infected with the new coronavirus from China, how many will die?
For weeks, it seemed that the answer was 2. Perhaps a little more, perhaps a little less.
The calculation was made by comparing the total number of people with confirmed cases of COVID-19 to the number of people who died of it. As both of those numbers grew, the ratio was bound to shift.
The 2% figure seemed stable on Feb. 24, when a massive study of nearly 45,000 Chinese patients whose infections had been confirmed with laboratory tests reported a case fatality rate of 2.3%.
Sign up for the latest news, best stories and what they mean for you, plus answers to your questions.
Later that week, on Feb. 28, a study of nearly 1,100 Chinese patients suggested a lower death rate, of 1.4%.
Four days later, on March 3, the World Health Organization said the global mortality rate was 3.4%.
How could it have changed so much in such a short period of time?
“It’s hard to say what the case-fatality rates are until the dust settles,” said Dr. George Rutherford, an epidemiologist and infectious disease expert at UC San Francisco.
“It’s not a statistic to be looking at kind of on an ongoing basis, even though I do it just as much as everybody else does,” he admitted.
That includes members of Congress. On Wednesday, they asked Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, what was going on. He urged them not to get so fixated on the fatality rate, emphasizing that scientists still have a lot to learn about it.
The biggest uncertainty surrounds the number of people who have been infected with the coronavirus. The official count has surpassed 100,000, but the true number is probably much higher.
Scientists suspected that many patients who experienced only mild illnesses never saw a doctor and therefore never got an official lab test to see if they were infected. As a result, they weren’t factored into the case-fatality rate.
Neither were people who wanted to be tested but weren’t due to a lack of resources. In a situation where test kits are limited, it’s better to save them for 80-year-olds with poor lung function and skip the healthier 40-year-olds with mild illness.
Counting problems aside, the case-fatality rate can change over the course of an outbreak.
For instance, the authors of the study of nearly 45,000 Chinese patients noted that the case-fatality rate in Hubei, the province at the center of the outbreak, was 2.9%. Elsewhere in China, it was 0.4%.
It’s possible that as China catches up on counting deaths, the fatality rate outside of Hubei will rise.
But it’s also possible that the difference is real because public health authorities outside of Hubei had a chance to get ready for the coronavirus, “so people aren’t dying at the same rates as they did before,” Rutherford said.
At UCSF, for instance, doctors are prepared to treat the coronavirus in a containment facility set up in the parking lot for the emergency room, he said.
Other countries are reporting different case-fatality rates.
With 6,767 confirmed cases as of late Friday, South Korea has more coronavirus patients than any country besides China. That high number may in part be a function of the country’s sweeping testing program that involved tens of thousands of people. When balanced against the country’s 44 deaths, the case-fatality ratio is less than 0.7%.
That figure may offer a better sense of COVID-19’s true fatality rate, said Dr. Jeremy Faust, an emergency physician at Brigham and Women’s Hospital in Boston and an instructor at Harvard Medical School.
“By testing so many people, they were actually able to find more cases,” Faust said. “It means they noticed the disease was everywhere, and it doesn’t kill that many people.
“The more testing you do, the more accurate your numbers become,” he added.
Rutherford predicted that by the time the dust settles, the overall case-fatality rate for COVID-19 will fall to somewhere between 1% to 2%.
Fauci agreed that the death rate could end up being as low as 1%. But even if that’s the case, he said, it would still be 10 times worse than the average death rate for the seasonal flu.