The chief medical officer at the practice where the tests were ordered is betting that the positive result is the right one.

“A false negative is problematic because it tells the patient they don’t have the virus,” said the doctor, Craig Deligdish, an oncologist at Omni Healthcare, a Melbourne, Fla., medical group. Meanwhile, that patient can unwittingly spread it, he said.

Health experts say they now believe nearly one in three patients who are infected are nevertheless getting a negative test result. They caution that only limited data is available, and their estimates are based on their own experience in the absence of hard science.

That picture is troubling, many doctors say, as it casts doubt on the reliability of a wave of new tests developed by manufacturers, lab companies and the U.S. Centers for Disease Control and Prevention. Most of these are operating with minimal regulatory oversight and little time to do robust studies amid a desperate call for wider testing.

“The whole testing field is in flux,” said Bill Miller, a physician and epidemiologist at the Ohio State University. “The thing that is different this time is most of these tests are going through a really rapid validation process. As a result we can’t be completely confident in how they will perform.”

The unknown rates of false-negative results could leave patients with unjustified confidence that they can return to business as usual and could show that the nation’s view into the spread of the disease has a blind spot, Dr. Miller said.

All lab tests have some margin of error, measured in terms of sensitivity— the percent of positive cases a test accurately detects—and specificity, the percent of negative cases it accurately detects.

The sensitivity of the currently available coronavirus tests seems to be lower than other similar tests, said Mike Lozano, a Tampa, Fla.,-based executive at Envision Healthcare Corp., a medical group that contracts to provide care at hospitals. Dr. Lozano said he estimates that the sensitivity of the tests is in the neighborhood of 70%, meaning nearly one in three positive patients walks away with a reassuring negative result.

Chris Smalley, a primary-care doctor at Louisville, Ky.,-based Norton Healthcare, said the 70% estimate was gaining traction with doctors finding out how well the tests work in the real world, absent better data. He said several of his patients with negative tests went on to develop highly suspicious coronavirus symptoms and require long hospital stays.

Dr. Smalley said a negative result is more likely to be accurate in places like Louisville where the prevalence is low, but could be virtually useless in New York, where it is high.

A February study of about 1,000 patients in Wuhan, China, who were hospitalized with suspected coronavirus there, where the pandemic began, found that about 60% tested positive using lab tests similar to those available in the U.S. But, almost 90% showed tell-tale signs of the virus in CT scans of their chests, the article, published in the journal Radiology, found, suggesting many patients in the group were testing negative despite active coronavirus infections.

The technology that underpins the new coronavirus tests so far available, a genetic method called polymerase chain reaction, typically detects over 90% of positive cases accurately, said Tom Taylor, a professor and former CDC statistician. In less dire circumstances, he said, the CDC would typically do a study to verify the accuracy of such tests, which might take a year.

The U.S. Food and Drug Administration initially required testing companies to submit applications for new tests, a move the agency said was required by law. Critics said that deterred private companies from moving quickly to develop badly needed tests.

As demand surged in late February, the agency relaxed the requirements, saying labs could start testing without approval and submit their internal studies later. Published details of those studies are often limited.

In a statement Monday, the FDA said it is balancing the need for more testing with its prerogative to ensure tests are accurate. Right now, the FDA’s priority remains getting more tests into the field, an agency spokeswoman said.

A Dallas-area emergency-room doctor, Peter Brokish, said several lab results he had sent to Laboratory Corp. of America Holdings came back negative only to have the patients return to the emergency department where he works with worsened symptoms, including fevers and coughs. After being admitted to a hospital, they tested positive.

“You kind of figure, yeah, that’s a false negative,” Dr. Brokish said.

Linda Girgis, a New Jersey primary-care doctor, saw the same pattern with a test run for a health-care worker by Quest Diagnostics Inc., a finding that made her fear that frontline workers might report for duty thinking they weren’t contagious.

“Our materials for providers, including our test reports, make clear that while these tests are designed to minimize false negatives and false positives, such results can occur,” said Wendy Bost, a Quest spokeswoman.

A Labcorp spokesman said that while a positive result is a definitive sign of coronavirus infection, “a negative result does not definitively rule out infection.” He said test accuracy can be affected by how a specimen is collected, and how long an individual has been infected before testing.

In New York City, “a negative is not clearing anybody who is symptomatic,” said Dara Kass, an emergency-room doctor at NewYork-Presbyterian, a hospital system. Dr. Kass, who is working from home after she tested positive for coronavirus in mid-March, said all patients with symptoms are being treated as coronavirus-positive, because the disease is so widespread in the city.

A continuing scarcity of testing materials is probably obscuring the true rate of incorrect results, doctors say. Few say they are willing to waste specimen collection kits double-checking a patient who already tested negative.

They say incorrect results could be caused by differences in how samples are collected and transported, as well as underlying problems with testing platforms.

Swabs taken from the front of patients’ noses, for instance, tend to have fewer copies of the virus mixed into the mucus than those taken from the nasopharynx, the intersection of the nasal cavity and throat. Swabbing the nasopharynx, however, requires inserting a long Q-tip through a patient’s nostril and deep into the head.

On Tuesday, in an effort to sort out what went wrong with the Florida baby’s tests, Omni’s Dr. Deligdish sent a third, freshly taken specimen from the patient to the lab that performed the initial positive test, Diatherix, a commercial lab in Alabama. That specimen tested negative late Wednesday.

Still, after Diatherix double-checked its earlier positive result late last week, Dr. Deligdish concluded that it was likely the accurate one, and he doubts the earlier negative result produced by the Florida public-health laboratory. Dr. Deligdish said he suspects the infant has now overcome the virus.

“It’s really hard to have a false positive” on the type of tests used to detect coronavirus, he said.

The Florida Department of Health didn’t comment. Diatherix declined to comment on the patient result.

A Diatherix spokeswoman said the company submitted an application to do testing to the FDA on March 23 and the FDA is reviewing it. So far, she said, the company and a sister lab also owned by Luxembourg-based parent Eurofins Scientific SE have run around 44,000 tests in the U.S. as of Wednesday.

Ambiguous results have left patients grappling with the consequences.

On March 17, Michele Hickle, of Lubbock, Texas, grew ill with a fever and later a rasping cough, she said. Her husband, a physician and businessman, suggested she effectively isolate herself in their master suite. He moved a minifridge and microwave into a hallway in the suite, she said.

That Friday, she received a test for coronavirus from a drive-through testing site, she said. Her specimen was tested at Texas’ state public-health lab and she received the results the next day: negative.

“At that point, we welcomed her back into the family,” said Scott Hickle, her stepson.

Her symptoms worsened. She said she woke up many times throughout the nights, drenched in sweat. The following Tuesday she went to an urgent-care center in town. Her children’s doctor happened to be on call. She reported her symptoms, including vital signs her husband had helped her record in recent days, and the doctor took x-rays and diagnosed pneumonia. Despite the negative test result a few days earlier, the doctor suspected coronavirus, she said.

Yet another sample was taken, this time sent to Quest, Randy Hickle, her husband, said. Quest took nearly a week to report results, on Monday, Dr. Hickle said. This time, it was positive.

A spokeswoman for the Texas Department of State Health Services said the test might not detect the virus in the early stage of infection and that patients could become infected between tests.

Michele Hickle now fears she could have spread the virus to her family. And indeed, she said, her youngest daughter, Grace, has had some symptoms: aches and fatigue.

The family is now awaiting a test result for her.