In a significant twist that could reshape our understanding of the early days of the coronavirus pandemic, death records now indicate the first COVID-related deaths in California and across the country occurred in January 2020, weeks earlier than originally thought and before officials knew the virus was circulating here.

A half dozen death certificates from that month in six different states — California, Alabama, Georgia, Kansas, Oklahoma and Wisconsin — have been quietly amended to list COVID-19 as a contributing factor, suggesting the virus’s deadly path quickly reached far beyond coastal regions that were the country’s early known hotspots.

Up until now, the Feb. 6, 2020, death of San Jose’s Patricia Dowd had been considered the country’s first coronavirus fatality, although where and how she was infected remains unknown.

Even less is known about what are now believed to be the country’s earliest victims of the pandemic. The Bay Area News Group discovered evidence of them in provisional coronavirus death counts of the Centers for Disease Control and Prevention’s (CDC) National Center for Health Statistics (NCHS) — widely considered the definitive source for death data in the United States — and confirmed the information through interviews with state and federal public health officials. But amid privacy concerns and fierce debate over pandemic policies, the names, precise locations and circumstances behind these deaths have not been publicly revealed. That is frustrating to some experts.

“We need to sit back and really assess what was this thing, when it started, how did we handle it, did we create more of a problem than we needed to, could we have handled things differently?” said Matthew Memoli, director of the clinical studies unit at the Laboratory of Infectious Diseases at the National Institutes of Health in Bethesda, Maryland. “There’s a lot to think about here.”

For instance, the far-flung nature of the deaths — in the West, Midwest and South — might suggest that restrictions on travel and social interactions should have been used in more places much earlier — and that such rapid response could be a more critical tool in the next pandemic. In January, when the United States announced it would begin limiting travel from China and other international hotspots, the virus may already have been speeding across state borders.

While California, Georgia, Alabama and Oklahoma acknowledged or didn’t dispute that a death certificate in their states from January 2020 had been changed to include COVID, none of the states would provide further details to reporters from this news organization, citing privacy laws. The Wisconsin Department of Health Services now lists the probable COVID death of a 50-59-year-old woman on Jan. 22, 2020, in its data. Kansas did not respond to a request for comment.

This new data appears to be the result of months-long efforts by so-called certifiers — the coroners, medical examiners and doctors across the country tasked with explaining when and why people die — to take a closer look at deaths that occurred in the months before the outbreak.

The Bay Area News Group first reported in April that the CDC was investigating why multiple COVID-related fatalities before Dowd’s death began appearing earlier this year in state and federal records. At the time, most of those were explained away as incorrect dates and other data glitches.

But this past week, the federal agency told this news organization that it had worked with state officials to contact the certifiers in five cases — while waiting to hear back from a sixth — and confirmed that death certificates from January 2020 have now been intentionally revised to include COVID.

A screenshot of the CDC/NCHS weekly COVID-19 death counts, which show six deaths happening in January, 2020.
The National Center for Health Statistics describes its provisional death counts for COVID as “the most complete and accurate picture of lives lost to COVID-19.” 

“Certifiers are reluctant to amend death certificates unless there’s a good reason to do so,” said Robert Anderson, chief of the Mortality Statistics Branch at the NCHS.

But what led a coroner or doctor to make such a significant and possibly historic change in these cases is unclear. Anderson, whose team collects death data from every state, said his agency isn’t provided that level of detail.

When someone dies for unknown reasons, death certificates can be updated months and even years after a person has died to reflect new information. During the pandemic, a coroner who originally attributed an early 2020 death broadly to a respiratory virus might later learn the person had traveled to China, where the virus originated, or had contact with someone who had, and reach the conclusion that they had COVID. Coronavirus testing was not common in early 2020, but if health workers took blood at the time, a medical examiner could later test the sample for the virus or antibodies, or do a PCR test for the virus on a tissue sample if an autopsy was performed.

It’s likely, said John Swartzberg, an infectious disease expert and professor emeritus at UC Berkeley, that these early cases were initially written off as colds or flus.

Dowd’s mysterious death wasn’t originally attributed to COVID. It took more than two months for tests to reveal the otherwise healthy 57-year-old San Jose woman had been infected with the COVID-19 virus when she died.

The existence of January 2020 deaths would dramatically revise the timeline of COVID’s arrival in the United States. China first announced a mysterious viral pneumonia in late December 2019, and reported the first death from the illness on Jan. 9, 2020. The U.S. originally recorded its first case in mid-January when a traveler tested positive after returning from Wuhan, China. The first deaths reported in the United States, in late February, were also tied to travel.

In its current death count, which reflects the six newly-discovered fatalities, the NCHS now lists the country’s first COVID death during the week of Jan. 5-11 — the first full week of 2020. The agency is in the final stages of preparing its 2020 annual mortality report, a review and analysis of all deaths in the United States last year.

Swartzberg thinks — and the new death data suggests — it’s entirely possible that COVID was present in the United States as early as December or even November. The time from infection to death from COVID is typically around three weeks.

“I would certainly guess the virus was introduced on multiple occasions before it was identified as a problem,” Swartzberg said, noting that states like Alabama and Oklahoma don’t generally see a lot of travel to and from China.

Memoli of the National Institutes of Health agrees.

“I always thought it had to have been here in the U.S. well before we identified it as a big problem,” Memoli said.

His team is studying thousands of people across the country and their research suggests that by July 2020, there were about five unidentified cases for every known case and possibly more. That information may also suggest, Memoli said, the virus was spreading earlier than previously thought. So too does a CDC analysis of thousands of blood samples from nine states. The samples, collected by the American Red Cross in December 2019 and January 2020, found evidence of antibodies to COVID-19 in all nine states, but there is no record of whether any of those people got sick or died.

Some states, including California, are choosing not to recognize the amended death certificates in their official COVID death counts, however.

The California Department of Public Health acknowledged in a statement to this news organization that a death certificate from January 2020 now lists COVID as a significant condition contributing to death. “However,” the department said, “there was no laboratory confirmation of COVID for this individual and as such is not a confirmed COVID death.”

Anderson, whose agency is tabulating the country’s official death count, sees it differently.

“The death certificate is the permanent record for the decedent,” he said. “If COVID is on the death certificate, then it is a COVID death. You can’t just ignore it.”

The system for tracking mortality data, he said, relies on the medical expertise of death certifiers, since the CDC compiles records on approximately 3 million deaths in the U.S. every year. In their months-long review before finalizing the data for publication, the CDC follows up on anything out of the ordinary, like these early COVID deaths, to make sure there were no unintentional errors made by the certifiers or during data entry.

But experts say a deeper analysis is needed. For instance, it might be possible, depending on what evidence remains, to examine the viral characteristics in the six early deaths and determine the origin of the strains and whether they were related.

The CDC wouldn’t comment on whether it was conducting such an investigation.

Memoli, for one, is convinced a closer look could inform how the United States responds to future public health crises.

“There are things about the 1918 flu pandemic we still don’t completely understand and we try to dig up information from that time to try to better understand it and prepare for the future,” he said. “We have a lot of work to do in trying to understand.”


Help us tell the stories of the country’s earliest COVID-related deaths. If you have any information about these or other cases now linked to the coronavirus in January 2020, please email Harriet Blair Rowan at hrowan@bayareanewsgroup.com or Emily DeRuy at ederuy@bayareanewsgroup.com