Why We Still Don’t Have a True COVID Death Count
A year after the beginning of the pandemic, the U.S. still does not have a full accounting of how many people have died of COVID-19.
The Centers for Disease Control and Prevention calculates that more than 463,000 Americans have died from COVID-19. But senior administration officials say that number is likely higher—and they’re working to identify how many more coronavirus deaths may have gone unreported and uncounted in the U.S. in the last 12 months.
Understanding the full impact of the virus on various portions of the population is more difficult without a clear picture of mortality during the pandemic, officials said. The Biden administration has placed a particular emphasis on comprehending better how COVID-19 has affected Americans in underserved and minority populations.
Officials are analyzing COVID-19 mortality data and aim to publish a CDC report on death during the pandemic in the next several months. That report may not be able to fully account for all of the COVID-19 deaths from last year because the CDC is still gathering mortality data from states from 2020. Officials say they are about four to five weeks behind, primarily because data reporting from state health agencies slowed through the holiday season.
The federal government and state health departments across the country are also still working to analyze hundreds of thousands of “excess deaths” from over the past year—or the difference between the number of observed deaths in 2020 versus 2019. Epidemiologists and statisticians working with the federal government say many of those deaths are likely COVID-19 deaths or deaths linked to the virus in some other way. Officials and health experts say there are also likely deaths that went unaccounted for over the last year, either because of death certificate accounting errors or because they were perhaps misreported as, for example, pneumonia deaths.
The COVID-19 Death Undercount Is Scarier Than You Think
“Mortality data from death certificates is notoriously poor in the United States. There are huge delays and a lot of misinformation,” said Jeffrey Klausner, clinical professor of preventive medicine at the University of Southern California. “The U.S. has a very weak public health system. Ideally, vitality records [birth and death] would be accurate, carefully reviewed, and used to track population-level health. Sadly, that’s not the case.”
The push to gather more accurate COVID-19 mortality data comes as the Biden administration is ramping up vaccine supply and as states are inoculating residents at an increasingly quick tempo. Senior officials working with the White House have asked the CDC to conduct a study on how many Americans have died since receiving the vaccine and the circumstances that led to their deaths. Officials are not concerned the vaccine is linked to reported deaths but want to gather more data about the vaccine, its efficacy, and what happened to the recipients after they got the shot.
“We’re not testing people right before we’re vaccinating. Some chunk of those people are going to get sick and die,” said Ashish Jha, dean of Brown University’s School of Public Health. “Understanding that is going to be really important because the thing we care most about is prevention of death and hospitalizations. We can’t sort it out unless we have very clear tracking of infections, vaccination, and how many went on to die.”
Several states have taken steps to account for COVID-19 deaths more accurately in their health agency reporting systems. In New York, for example, the state health department announced in August that it would require nursing homes and hospitals to test for COVID-19 and the flu whenever a patient “has a known exposure or symptoms consistent with either disease.” It also required those health-care facilities to perform the testing on patients who were suspected of having died from either disease within 48 hours of their death if the individual had not previously received a test in the 14 days prior.
Multiple doctors in New York City who spoke to The Daily Beast for this story said they have recently been asked by their hospitals to go to the morgue to swab COVID-19 patients after their deaths—if those patients did not receive a test in the two weeks before they passed away. Doctors said they are confused by the requests; the patients have already been listed as COVID-19 victims on their death certificates. The doctors added that some of the patients tested positive when they arrived at the hospital but were treated in intensive care units for several weeks.
“We’ve been getting more calls from the health department asking us if some of the deaths we’ve recorded are actually COVID-19 deaths or something else,” one of those doctors, who requested anonymity to speak more freely about the matter, said. “And now we have to go and test people after they die. We’ve never had to do that until recently.”
Officials in the New York state health department said the postmortem swab regulations are supposed to help the state better account for COVID-19 deaths by weeding out those patients who, for example, had a disease like cancer and died but also tested positive for the virus. They are also trying to ensure they identify whether pneumonia deaths should be counted as COVID-19 deaths.
In Ohio, officials said Wednesday that they were adding an additional 4,000 deaths to the state’s COVID-19 death counts, many of which went uncounted during the holiday season.
The increased focus on mortality in the age of COVID-19 comes after the Trump White House repeatedly attempted to downplay the death count in the country. In one task force meeting in May, White House coronavirus task force member Dr. Deborah Birx said she thought the CDC was inflating mortality statistics and blamed the agency for the way it went about collecting virus data, according to a report by The Washington Post. That same month, White House officials attempted to pressure the CDC to change the way it calculated COVID-19 deaths, as first reported by The Daily Beast.
Team Trump Pushes CDC to Revise Down Its COVID Death Counts
In July, the Trump White House requested that the CDC draw up a report that showed the U.S. mortality rate was lower than the mortality rate in Europe, according to two individuals with direct knowledge of the matter. Trump wanted to use that report in a speech that he was set to give in August. After internal conversations, political appointees at the Department of Health and Human Services, including Secretary Alex Azar, had to inform the president that the data did not show the U.S. was doing better than Europe in terms of reported COVID-19 deaths and that no such report could be drawn up through the interagency. It’s unclear who drafted the inaccurate mortality presentation that Trump did end up unveiling in several press briefings and in an interview with Axios’ Jonathan Swan.
The CDC analyzes COVID-19 deaths through two parallel tracking systems. It relies on the data it receives from local departments of health and through information it gathers from states through a death certificate digital coding system. The death count is normally higher from states’ health systems than the death certificate system data, but officials say they have the ability to reconcile those differences.
The CDC and state health departments use excess death studies to better understand the impact of the virus.
For example, one study by the New York City Department of Health and Mental Hygiene published in May said there were thousands of “excess deaths” in the city from March 11 to May 2. About 18,879 of those deaths were explicitly tied to the coronavirus. There were also an additional 5,200 that were not identified as either laboratory-confirmed or probable COVID-19-associated deaths but could have been tied to the virus in some other way.
According to the CDC, the U.S. began to exceed its normal projected death toll in the third week of February 2020, when the nation attributed just five fatalities to COVID-19. In March, as the body count from the novel coronavirus mounted, the country saw a simultaneous surge in deaths due to pneumonia—even apart from those definitely ascribed to the new disease. For instance, in the week concluding March 28, the number of pneumonia deaths shot up to 6,181 from 4,547 for the previous seven-day span. Of these, 1,434 were confirmed "pneumonia and COVID-19 deaths."
The following week—in which the number of deaths blamed on COVID-19 roughly tripled—the country recorded 9,931 pneumonia mortalities, of which the CDC counted 4,768 cases in which COVID-19 had been a verified factor, leaving 5,163 cases with an unconfirmed cause. By comparison, in January, the total number of pneumonia cases hovered in the low 4,000s and in its final days had dipped into the high 3,000s.
New York City, home to the worst early cluster, reported 83 pneumonia-induced deaths the week that ended Jan. 4, 2020. The week that ended April 4, it reported 1,782, 1,368 of them “pneumonia and COVID-19.”
The gap between the number of pneumonia deaths that were not definitively pinned to COVID-19 and those that were narrowed as the pandemic raged on and testing spread and improved.
But the gap grew again as cold weather set in, peaking the final week of December, when the nation suffered 58,334 pneumonia deaths, 44,989 classified as “pneumonia and COVID-19.”
All the numbers began drastically dropping after the holidays, as all COVID-19-caused deaths fell from 18,584 the week that ended Jan. 9, 2021, to 2,528 the one that concluded Feb. 6. The number of pneumonia casualties plummeted in that time period from 11,955 to 1,687, and deaths attributed to “pneumonia and COVID-19” collapsed from 9,702 to 1,260. This decrease occurred simultaneously in multiple states The Daily Beast reviewed, including New York, California, Florida, and Texas.
“I would expect those pneumonia deaths were probably COVID-related deaths,” said Klausner, noting in particular the lack of comprehensive testing in hotspots such as New York at the start of the pandemic. “A lot of people were dying of probably COVID-related pneumonia, and the doctors were probably just writing ‘pneumonia’ and not knowing if it was COVID-related.”
“Cause of death is often scribbled down late at night by a medical intern who has never met the deceased,” the veteran infectious disease expert added. “The reality is it’s a very sloppy process in terms of the cause of death. You just want to put down something as cause of death that might be acceptable.”
Klausner warned against making prematurely optimistic inferences from the decline in COVID-19 deaths in the past few weeks and argued a true trend will only be visible a month out. For now, Klausner argued that the public can best understand the scope of COVID-19’s destruction by examining the excess death numbers, or by comparing the mortality totals from 2019 to 2020.
The first year saw 2,854,838 fatalities in the United States. According to the latest CDC tally, last year witnessed more than 3.3 million.
Got a tip? Send it to The Daily Beast here
Get our top stories in your inbox every day. Sign up now!
Daily Beast Membership: Beast Inside goes deeper on the stories that matter to you. Learn more.
