New autopsy results show two Californians died of coronavirus in early and mid-February — up to three weeks before the previously known first US death from the virus.

These deaths now stand as the country’s earliest two attributed to the novel coronavirus, a development that appears to shift the understanding of how early the virus was spreading in the country, health experts told CNN Wednesday.

Two deaths in Northern California’s Santa Clara County happened February 6 and 17, the county said Tuesday in a news release.

The previously understood first coronavirus death happened on February 29 in Kirkland, Washington.

The two in California had no known travel histories to China or anywhere else that would have exposed them to the virus, Dr. Sara Cody, the county’s chief medical officer, told The New York Times. They are presumed to have caught the virus through community spread, she told the Times.

“That is a very significant finding,” Dr. Ashish K. Jha, director of the Harvard Global Health Institute, told CNN’s “New Day” on Wednesday.

“Somebody who died on February 6, they probably contracted that virus early to mid-January. It takes at least two to three weeks from the time you contract the virus and you die from it.”

If they did not contract coronavirus through travel abroad, that also is significant, Jha said.

“That means there was community spread happening in California as early as mid-January, if not earlier than that,” Jha said.

“We really need to now go back, look at a lot more cases from January — even December — and try to sort out when did we first really encounter this virus in the United States,” Jha said.

CDC confirmed Tuesday that tissue samples were positive

The Santa Clara County Medical Examiner-Coroner performed autopsies on two people who died in their homes February 6 and 17 and sent samples to the US Centers for Disease Control and Prevention, the county said Tuesday.

The CDC confirmed Tuesday that the tissue samples tested positive for coronavirus, the county said.

A third death in early March was also confirmed to be virus-related, the release said.

At the time of the deaths, the county said, testing for the virus was very limited — generally restricted only to people with a known travel history and seeking treatment for certain symptoms, and available only through the CDC.

Further details on the victims were not provided.

The county said that as more deaths in the county are investigated, it’s likely there will be more that are tied to the virus.

Dr. Colleen Kraft, associate chief medical officer of Atlanta’s Emory University Hospital, agreed with Jha about the announcement’s significance.

“That also means that a lot more people have had this, probably asymptomatically or with mild illness, than we thought before,” Kraft said Wednesday.

Studies: Far more people may have been infected in LA, Santa Clara counties

News about the two February deaths adds to evidence suggesting not only that coronavirus may have been spreading in the US earlier but also the US case and death tallies may be significant undercounts.

Two studies involving antibody blood tests in California — one in Los Angeles County and the other in Santa Clara County — so far have suggested that the number of people already infected is dozens of times higher than officially reported.

Both studies are ongoing and have yet to be peer reviewed. Preliminary results were announced in the last few days.

Still, the studies estimate that only small fractions of the county populations have antibodies.

The Los Angeles County study estimated between 2.8% and 5.6% of the population — 221,000 to 442,000 people — had Covid-19 antibodies, based on drive-through testing conducted on April 10 and 11. That would have been 28 to 55 times the number of cases that county officials recorded around that time.

In Santa Clara County, the study estimated 2.49% to 4.16% of people there had been infected with Covid-19 by April 1. That represents between 48,000 and 81,000 people — and 50 to 85 times the cases that county officials recorded by that date.

Similar efforts to estimate local antibody prevalence have launched in places like Miami-Dade County, Florida; San Miguel County, Colorado; and New York City.

Experts say these studies could help improve projections and disease modeling, and give a more realistic sense of how deadly the virus really is.