Daily COVID-19 Deaths in the U.S. Have Fallen Dramatically Since April

The downward trend continued after states began lifting their lockdowns.

Amid concerns about post-lockdown increases in COVID-19 infections, there is one striking piece of good news: Daily deaths from the disease in the United States have fallen dramatically since April, and that downward trend has continued into June. Even taking into account the average lag between laboratory confirmation and death—about two weeks, according to the U.S. Centers for Disease Control and Prevention (CDC)—the loosening of restrictions on movement and economic activity that began at the end of April so far has not led to the surge in COVID-19 deaths that many lockdown supporters predicted.

The seven-day rolling average of daily deaths, which peaked at 2,210 on April 18, had fallen to 605 as of yesterday—a 73 percent drop. The downward trend has continued for more than a month since mid-May, when the impact of post-lockdown infections should have started to show up in fatality figures. In Texas, for example, the seven-day average fell from 58 on April 30, when the statewide lockdown was lifted, to 20 on June 13 before climbing to 30 as of yesterday.

Some states, including Texas, have seen notable increases in confirmed cases and hospitalizations since late May. Those increases, which cannot be fully explained by expanded virus testing, may be related to Memorial Day gatherings and the mass protests against police brutality triggered by George Floyd's death. The spike in cases that states such as Texas have seen can be expected to result in more deaths during the next couple of weeks than otherwise would have occurred. But if epidemiologists are correct in thinking that superspreading events on and after Memorial Day explain recent surges in infections—which makes senses given the timing—the resulting rise in daily deaths should be temporary.

The nationwide crude case fatality rate (CFR) has fallen along with daily deaths, from slightly more than 6 percent on May 16 to about 5.3 percent today. Since the denominator includes only confirmed cases, you would expect the crude CFR to drop as testing is expanded to include more people with mild symptoms. If all infections were detected, according to the CDC's "best estimate," the overall fatality rate among everyone infected by the virus in the United States would be less than 0.3 percent.

You would also expect the crude CFR to fall if newly infected people are less likely to have serious preexisting medical conditions than people infected earlier in the epidemic.* That seems plausible as younger, healthier people move about more now that they are allowed to do so, returning to work and visiting businesses more, especially if efforts to protect Americans who are most vulnerable to COVID-19 are successful. (As of May 22, nursing homes accounted for more than two-fifths of COVID-19 deaths in the United States.) More effective treatment of COVID-19 may also account for some of the drop in the crude CFR.

Youyang Gu's epidemiological model, which has a good track record of predicting COVID-19 fatalities, currently projects that daily deaths in the United States will continue to decline until early July, then rise through mid-August, exceeding the current level by late July, before declining again through September, dropping below the current level by the middle of that month. The model shows the nationwide death toll, currently about 120,000, rising to about 143,000 by the end of July, about 163,000 by the end of August, and about 180,000 by the end of September.

I would not call that good news, but it is not nearly as bad as the estimate of 200,000 deaths by June 1 that The New York Times trumpeted in early May. The Times claimed that wildly wrong projection "underscore[s] a sobering reality" and "confirm[s] the primary fear of public health experts: that a reopening of the economy will put the nation back where it was in mid-March, when cases were rising so rapidly in some parts of the country that patients were dying on gurneys in hospital hallways." The estimate embraced by the Times projected 3,000 daily deaths by now, five times the current level.

*Addendum: Data from several states confirm that the average age of people who test positive for the COVID-19 virus is falling. In Florida, the median age of newly identified patients fell from 65 in early March to 35 last week. Texas Gov. Greg Abbott notes that residents younger than 30 account for a growing share of confirmed cases in that state. In Arizona, 20-to-44-year-olds, who represent 37 percent of the population, now account for 48 percent of confirmed cases.