A new study of COVID-19 cases in Wuhan, China, estimates that the death rate among people who were infected and developed symptoms was 1.4 percent. That is far lower than the crude case fatality rate (CFR) produced by dividing total deaths into total confirmed cases (4.5 percent) and far lower than the global CFR initially calculated by the World Health Organization (3.4 percent). The study, reported yesterday in Nature Medicine, suggests that the overall CFR—including people who are infected but do not develop symptoms—will prove to be much lower in the United States than many people feared.
During congressional testimony last week, Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, estimated that the overall CFR in the United States would be about 1 percent. The worst-case scenario sketched by the U.S. Centers for Disease Control and Prevention (CDC) imagines 214 million infections and 1.7 million deaths, which translates into a CFR of about 0.8 percent, similar to the crude CFR in South Korea, which has a much more robust testing program than the U.S. has managed. But all of these estimates exclude people infected by the coronavirus who have no symptoms or symptoms so mild that they never register in the official numbers.
John Ioannidis, an epidemiologist and biostatistician at Stanford University, notes that the CFR among passengers and crew on the Diamond Princess cruise ship—"the one situation where an entire, closed population was tested"—was 1 percent, but "but this was a largely elderly population, in which the death rate from Covid-19 is much higher." Given the age structure of the general U.S. population, he says, "the death rate among people infected with Covid-19 would be 0.125%." But since that estimate is based on a small sample of just 700 people, Ioannidis suggests that "the real death rate could stretch from five times lower (0.025%) to five times higher (0.625%)." Taking into account delayed deaths, he says, "reasonable estimates for the case fatality ratio in the general U.S. population vary from 0.05% to 1%."
That's a very wide range, highlighting the weak empirical basis for aggressive, economically ruinous COVID-19 control measures. If the CFR is as low as 0.05 percent, which is half the estimated CFR for the seasonal flu, the number of deaths in the CDC's worst-case scenario would be 107,000, meaning the projection is off by a factor of 16. If a third of the population is ultimately infected, rather than the 65 percent assumed in the CDC's worst-case scenario, the number of deaths plummets from 1.7 million to about 55,000. It is impossible to assess the cost-effectiveness of mass interventions such as statewide or nationwide "shelter in place" orders in the face of such uncertainty.
The Nature Medicine study confirms that COVID-19 death rates vary widely by age. The estimated fatality rate in symptomatic cases was 2.6 percent among patients 60 or older, compared to 0.5 percent for patients in their 30s, 40s, or 50s, and 0.3 percent among patients younger than 30. To put it another way, patients in the oldest age group were about five times as likely to die as those in the middle age group and more than eight times as likely to die as those in the youngest age group.
The researchers caution that the symptomatic CFR for Wuhan cannot necessarily be extrapolated to other parts of China, let alone other countries. "Given that Wuhan is no longer the only (albeit the first) location with sustained local spread, it would be important to assess and take into account the experience from elsewhere, both domestically in mainland China and overseas," they say. "These secondary epicenters, having learned from the early phase of the Wuhan epidemic, might have had a systematically different epidemiology and response that could impact the parameters estimated here."
The study's main symptomatic CFR calculation assumes that half of the people infected by the COVID-19 virus develop symptoms. If that's true, the number of infections is at least twice as high as the number of known cases, and that's assuming everyone with symptoms gets counted, which is certainly not the case, since people with mild symptoms may never seek medical attention or testing. But that factor alone means that the overall CFR is dramatically lower than it appears to be based on reported cases and deaths.
"One largely unknown factor at present is the number of asymptomatic, undiagnosed infections," the researchers note. "Estimates of both the observed and unobserved infections are essential for informing the development and evaluation of public health strategies, which need to be traded off against economic, social and personal freedom costs. For example, drastic social distancing and mobility restrictions, such as school closures and travel advisories/bans, should only be considered if an accurate estimation of case fatality risk warrants these interventions, which seriously disrupt social and economic stability."