A growing number of recovered COVID-19 patients are relapsing in South Korea, raising new questions and concerns among scientists and health authorities after the country successfully flattened the curve.

By Friday, Korean health authorities had identified 163 patients who tested positive again after a full recovery. The number more than doubled in about a week, up from 74 cases on April 9. Those patients — just over 2% of the country’s 7,829 recovered patients — are now back in isolation.

According to Korea Centers for Disease Control and Prevention data on Friday, the age and regional distribution of relapse cases are largely in line with that of the total infections.

To find out reasons for relapse, South Korean health authorities are running a range of tests and vetting various scenarios. The World Health Organization said last week that it is investigating the issue. While a fuller analysis will take at least a few weeks, early findings suggest there can be more than one cause.

Top KCDC officials said in recent briefings that the most likely possibility is reactivation of remaining viruses in patients’ systems. If a patient had not developed sufficient immunity against the virus or if a patient’s immune system weakens after recovery, the previously undetectable level of virus concentration could rebound. Or the novel coronavirus may be capable of staying dormant before reactivating.

Another possibility is that tests are picking up dead virus particles that are no longer infectious or transmissible. KCDC director-general Jeong Eun-kyeong said Friday that viruses collected from six relapse cases could not be cultivated in isolation, signifying that they are either dead or too small in number.

But some relapsed patients may have living viruses that make them sick. As of Friday, at least 61 developed symptoms, albeit mild.

A live virus is probably also transmissible, according to Jeong, but no secondary transmission by relapsed cases has been reported.

Reinfection through another virus carrier is a less probable scenario, considering that patients are retesting positive not long after they are released from treatment. Jeong said on Friday that relapse cases are detected an average of 13.5 days after recovery. The longest reported interval, however, is 35 days.

KCDC has also mentioned errors in testing or sample collecting as potential causes.

Short of definitive answers, authorities are for now advising recovered patients to stay home for an additional two weeks and to monitor for symptoms.