What could be worse than getting the pneumonia-like illness now known as Covid-19? Getting it twice.

That’s what Japanese government officials say may have happened to a female tour bus guide in Osaka. The woman was first diagnosed with Covid-19 in late January, according to a statement released by Osaka’s prefectural government Wednesday. She was discharged shortly after, once her symptoms had improved. A subsequent test came back negative for the virus. Three weeks later she returned with a sore throat and chest pain and tested again. For a second time, she tested positive for Covid-19.

Read all of our coronavirus coverage here.

News reports detailing the case raised the possibility that people may not be developing immunity to the new coronavirus, even after they’ve recovered. But several infectious disease specialists say there’s not enough data to support that conclusion. Another possibility is that the virus subsided and flared up again. (Some viruses tend to do that.) Or the test was simply wrong.

“The question is really: How good is the proof?” says Donald Burke, an international health expert at the University of Pittsburgh. “And we’re a long way away from anything solid. What we need is the sequence of the virus.”

The only way to differentiate between a relapse–meaning the same coronavirus seemed to go away and came back–and reinfection, in which a second strain of coronavirus swooped in just as the first cleared out, is a full viral sequence. The tests currently available for diagnosing Covid-19 are based on a method called RT-PCR, which picks up some pieces of the virus’s genetic code, but not everything. It’s designed to grab only the chunks that are unique to the new coronavirus, but stable enough that they won’t disappear if it mutates. Thanks to some specialized fluorescent dye, the more viral bits there are, the brighter the genetic material glows, creating a pattern of light that signals for the presence of the virus.

In theory, if doctors completely sequenced the virus in a patient’s nose or mouth swab at the time of each positive diagnosis, they could compare each genetic letter using genome-reading software and determine if the person had the same strain of coronavirus or a new one. Without that kind of data, says Burke, there’s no way of knowing for sure.

But others think the more likely scenario is that the virus can just linger in some people’s bodies longer than expected. “I suspect this is in fact a continuation of the original infection,” says Susan Kline, an infectious disease physician and epidemiologist at the University of Minnesota.

For one thing, she says, just not enough time has elapsed for reinfection. In the case of the other coronaviruses people catch, like the ones that cause the common cold, people tend to develop immunity following an infection. But it doesn’t last forever. The body produces antibodies that are protective against subsequent exposures, and then over time that immune response wanes, says Kline. That process usually takes longer than three weeks. “It’s too soon,” says Kline. “This entire outbreak has only been going on for two months. I would be very surprised if people are getting reinfected in that time span.”

So why then, did the Japanese woman test negative? Kline says there are a few potential explanations.

One is that the test wasn’t sensitive enough to pick up traces of the virus. The RT-PCR tests currently in use for diagnosing Covid-19 require a fair amount of genetic material to work well. If the coronavirus is actively making more copies of itself, there will be lots of RNA for the test to detect. But if the viral load goes down, the test can give false negatives. That can happen because the patient’s immune system is mounting a strong response. Or it could be that a treatment is slowing down the coronavirus’s self-replication spree. Another possibility is that the sample was just bad–the swab didn’t pick up much virus in the first place.