Can people who recover from a bout with the new coronavirus become infected again — and again?

The Japanese government reported this past week that a woman in Osaka had tested positive for the coronavirus for a second time, weeks after recovering from the infection and being discharged from a hospital.

Combined with reports from China of similar cases, the case in Japan has raised some uncomfortable questions. Reinfections are common among people who have recovered from coronaviruses that cause the common cold.

But those pathogens are very different from the new coronavirus, and experts said it’s unlikely that these are cases of people getting infected a second time.

“I’m not saying that reinfection can’t occur, will never occur, but in that short time it’s unlikely,” said Florian Krammer, a virologist at the Icahn School of Medicine at Mount Sinai in New York.

Even the mildest of infections should leave at least short-term immunity against the virus in the recovering patient, he said.

More likely, the “reinfected” patients still harbored low levels of the virus when they were discharged from the hospital, and testing failed to pick it up.

Even if there were occasional cases of reinfection, they do not seem to be occurring in numbers large enough to be a priority at this point in the outbreak.

A report published Thursday in JAMA supports the idea that people may test positive for the virus long after they seem to have recovered.

In four medical professionals exposed to the virus in Wuhan, China, the epicenter of the epidemic, a test that detects the viral genetic material remained positive five to 13 days after they were asymptomatic.

This does not necessarily mean that they were still able to infect others, however.

The PCR diagnostic test is highly sensitive and can amplify genetic material from even a single viral molecule. As such, the test could merely be picking up fragments of the virus.

PCR tests may detect remnants of the measles virus months after people who had the disease stop shedding infectious virus, Krammer said.

The other possibilities are that the negative test was done poorly or the samples were stored at a temperature at which the virus deteriorates. The throat swab may also simply miss the virus that is hiding elsewhere in the body.

“A virus test is positive if the virus was there on the swab in sufficient quantities at the time you swabbed the person,” said Marc Lipsitch, an epidemiologist at the Harvard T.H. Chan School of Public Health. “A negative test is not a definitive that there is no more virus in that person.”

Lipsitch offered an analogy: a jam jar with mold on top. Scraping off the surface might give the impression that the jam is now mold-free, but in fact the jar may still contain mold that continues to grow.

The Japanese woman initially had mild symptoms of coronavirus infection and tested positive in late January. She was released from the hospital Feb. 1. She tested positive again Wednesday after coming in for a sore throat and chest pain.

“That certainly sounds like it could be an actual resurgence of the virus in infectious form,” Lipsitch said. But, he added, “Single anecdotes are really hard to interpret.”

One worrisome possibility is that the coronavirus follows what is known as a biphasic infection: The virus persists and causes a different set of symptoms than observed in the initial bout.

In patients infected with Ebola, the virus may persist for months in the testes or eyes even after recovery — and can infect others and keep the epidemic going.

The recovered person, too, can develop other symptoms, including insomnia and neurological problems, said Angela Rasmussen, a virologist at Columbia University.

“We don’t know if that’s the case with this coronavirus,” Rasmussen said. “We don’t know anything about this virus.”

Coronaviruses are on the whole poorly understood, she said. Before the SARS epidemic, coronaviruses were not known to cause serious illnesses.

Some scientists have said that people infected with the new coronavirus produce antibodies that will protect them in the future. And a single-patient report suggests that the immunity may last at least seven days.

But this finding is neither surprising nor reassuring, said Dr. Stanley Perlman, a coronavirus expert at the University of Iowa. “The issue is whether you’ll see it in seven months or in a year,” he said. “That’s what you care about.”

The new coronavirus closely resembles the ones that cause SARS and, to a lesser extent, MERS. There are no reports of reinfections with the SARS virus, Perlman said, and only one that he has heard of in a patient recovering from MERS.

Perlman’s research with MERS has shown that the strength of the immune response depends on the severity of the infection, but that even in those with severe disease — which should produce the strongest immune responses — the immunity seemed to wane within a year.

How long immunity lasts will also be a key question to resolve when designing a vaccine for the new coronavirus, particularly if the virus becomes a seasonal threat like influenza.

“What is the nature of immunity to this virus after infection?” Lipsitch said. “That’s a research question that’s urgent.”