Late last year, the federal government’s chief of mortality statistics received news of a tantalizing discovery:
Someone had died of COVID-19 in January 2020, according to a death certificate, a revelation that would have accelerated the timeline of the spread of the virus in the United States in several weeks.
In the end, that death was not what it seemed.
The person who certified it had referred to June 2020, not January.
But that blink on the radar screen of Robert Anderson, the chief of mortality statistics for a branch of the Centers for Disease Control and Prevention, helped launch a silent, year-long campaign at the agency to check and recheck the first suspicious deaths the country related to COVID in the uncertain days of early 2020.
Now, at least four possible COVID-19 deaths from January 2020 have survived Anderson’s investigation.
Spread across four states, they have become part of a sparse collection of clues about the early spread of the virus beyond China – some of them reliable, some not so much – that have begun to garner more attention as scientists and intelligence officials try to unravel how it began the pandemic.
According to some scientists, the chances that the four new fatalities from the CDC – from Kansas, California, Alabama and Wisconsin – are actually a consequence of COVID-19 are slim.
This year, a doctor or other official certifier reclassified them as related to COVID.
But it’s not clear if they did this based solely on the person’s symptoms or with the help of more useful blood or tissue samples.
The first death, which occurred on Jan. 9, 2020 in Kansas, was reclassified this spring based solely on the person’s symptoms, according to the state health department, requiring the doctor to assess the extent to which the disease of the patient matches COVID-19 symptoms.
The most common symptoms of the virus, such as ffever, shortness of breath and even loss of taste or smell coincide with those of other respiratory diseases.
In a sign of how difficult it can be to recategorize longstanding deaths, CDC records include a fifth COVID-related death from January 2020, in Oklahoma.
But after state officials investigated, the medical examiner removed COVID-19 from the death certificate, the Oklahoma health department said Wednesday, meaning it will likely drop from CDC records soon as well.
It is unclear whether any of the suspected cases had traveled to China.
“My guess is that they are probably not all real, maybe not even some of them,” said Michael Worobey, an evolutionary biologist at the University of Arizona.
It’s highly unlikely that any of the people caught the virus in the United States, he said, but some may have recently been to China.
“If any of them are real, they would be linked cases to travel, and that’s conceivable, “he said.
Worobey’s research, which includes careful analysis of viral genomes and simulations of epidemics, has indicated that the virus was unlikely to spread outside of China before mid-December 2019, raising doubts about the deaths of people who they did not travel to the United States the following month, he said.
It usually takes several weeks between someone being infected and dying, and any case is unlikely to be fatal.
“Extraordinary claims require extraordinary proof,” said Worobey.
For Anderson, the reliability of the January 2020 deaths is not just a public health issue, but also a personal concern.
Anderson, a demographer by training who landed at the CDC in 1996 after dealing with a treacherous academic job market, said that certain deaths had an undue impact, statistically speaking, and that includes COVID deaths early in the pandemic.
When states report COVID-19 death figures, the CDC often takes them at face value, as they do with the hundreds of thousands of deaths from heart disease or cancer in any given year.
The agency recorded nearly 3 million recorded deaths in 2019, and a mistake here or there doesn’t change the overall picture of the country’s mortality, Anderson said.
Not so a possible death from COVID at the dawn of a pandemic.
At the time, the evidence was scant.
until The Mercury News California recently reported possible cases from January 2020, the first suspected COVID-related death did not occur until February 6, 2020.
When Anderson receives notification of COVID-related deaths beginning in the first two months of 2020, he calls state health officials, who in turn request the check of the doctor or coroner who signed the death certificate.
In January of this year, for example, the CDC received a flood of reports of people who had died from COVID-19 in January 2020.
Or so they said.
Subsequent checks revealed that most doctors had simply forgotten to start writing 2021 next to their signatures.
In another case this summer, Anderson confirmed that a medical certifier had wanted to reclassify a January 2020 death as COVID-related, only for that certifier to back down once the state health department intervened.
For Anderson, this is a single death from 2007 that the agency says was caused by diphtheria, a serious bacterial infection that is virtually unheard of in the United States.
Later, the CDC determined that the death had not been caused by diphtheria at all; the fault was a simple coding error.
“It was a bit embarrassing to have it in our data file,” Anderson said.
“Even if it is a single death, in that context, it has a great impact, because it is very visible.”
Ultimately, however, by not having access to patient samples or medical records, the CDC has to trust the people who sign COVID-19 death certificates, he said.
“I can’t assure that they are all accurate,” Anderson said of the January 2020 deaths.
“But it seems to me unlikely that the certifiers have whimsically changed the death certificate.”
Most deaths from COVID-19 are straightforward to certify, said Marcus Nashelsky, a professor of pathology at the University of Iowa who helped the CDC write guidelines on how to attribute such deaths.
For example, early in the pandemic, when nursing home patients were not always tested, he said, a known outbreak at the nursing home, in combinación With the characteristic symptoms and signs of the virus, it could be enough for a doctor to declare a COVID-related death.
However, death certificates have become documents much discussed during the pandemic.
In some cases, unbelieving families have asked for any mention of COVID-19 to be removed from the death certificate, said James Gill, Connecticut’s chief medical examiner and president of the National Association of Coroners.
In others, families have urged that COVID-19 be added to a death certificate, apparently in an effort to be eligible for funeral funding under a federal assistance program, he said.
“It’s a very emotional thing for some families, whether they want it on the death certificate or not,” Gill said.
“It shouldn’t be like that. It’s a public health issue.”
In the spring of 2020, Gill said, she became concerned Connecticut was overlooking deaths from COVID-19, especially among nursing home residents whose complicated medical histories can sometimes darken causes of death.
In a few hundred cases, he said, the coroner’s office performed deep nasal swabs on the bodies at funeral homes.
The team found several cases.
As China refuses to share more information about its own early cases, the World Health Organization recently said it was helping researchers dig into late 2019 case reports outside of China.
In Italy, researchers have reported coronavirus antibodies in blood samples from September 2019, as well as signs of the virus in a skin sample from a patient from November of that year.
Some scientists, however, have questioned both findings.
“The further back you can go, the more informative it can be – if you have true, confirmed cases,” said Marion Koopmans, a Dutch virus expert whose lab retested Italian blood samples and was unable to confirm the first cases.
“To declare a much earlier introduction of the pandemic virus in a region, you have to have a high level of certainty“.
An analysis of blood tests from the United States released this summer suggested that the virus may have been circulating in Illinois since December 24, 2019, although scientists have said those methods are fallible as well.
Keri Althoff, a public health researcher at the Johns Hopkins Bloomberg School of Public Health and lead author of that study, said small clusters of cases could have developed without igniting a full epidemic.
“It is not entirely known where COVID was sown in the United States,” he said, “but it is not likely that it was a single seed.”
Alyssa Lukpat contributed information and Susan Beachy contributed research.
c.2021 The New York Times Company