In what appears to be a first, disease trackers in Singapore have used an experimental antibody test for COVID-19 to confirm that a suspected patient was infected with the coronavirus. The patient was one of two people who together formed a missing link between two clusters of cases that each occurred in a Singaporean church.

Researchers around the world are racing to develop antibody tests, also called serological tests, that can confirm whether someone was infected even after their immune system has cleared the virus that causes COVID-19. The group that developed the test, at Duke-NUS Medical School in Singapore, is among the front-runners, although its assay has to be validated before it is taken into production and deployed widely.

Current tests for SARS-CoV-2, the virus that causes COVID-19, look for genetic material of the virus, for instance in saliva or nasal, oral, or anal swabs, using the polymerase chain reaction (PCR). They have one huge drawback: They only give a positive result when the virus is still present. The tests can’t identify people who went through an infection, recovered, and cleared the virus from their bodies.

The new antibody test helped contact tracers at Singapore’s health ministry who have been trying to find the source of a cluster of 23 COVID-19 cases at the Grace Assembly of God church, which has two large houses of worship in the city-state. Health ministry contact tracers had identified the primary case, a 28-year-old man who fell ill on 29 January. But they could not determine how he became infected.

The other cluster of cases started during a service at the Life Church and Missions, a small, independent evangelical congregation, and was apparently “seeded” by visitors from Wuhan, China, on 19 January. Contact tracers found a possible link: The index case at Grace Assembly of God had attended a Lunar New Year celebration on 25 January along with a couple who had attended a church service with the Wuhan visitors.

The Life Church couple both developed symptoms and sought medical advice around 25 January, but they were not diagnosed as COVID-19 patients because of their mild symptoms. After discovering the Lunar New Year celebration link, investigators sent the couple to the National Center for Infectious Diseases for tests on 18 February. Because some time had already elapsed since they had recovered, specialists thought they had likely cleared the virus and it would be too late to confirm an infection using the PCR test. So, they tested the couple using both PCR and antibody tests.

Unexpectedly, the husband tested positive by PCR; he was hospitalized in isolation the next day. His wife’s PCR test was negative, but the antibody test results, available several days later, showed she had antibodies, as did her husband. “We believe this is the first time in the world where these particular tests have been used in this context” of contact tracing, virologist Danielle Anderson of Duke-NUS said during a Tuesday briefing.

Researchers are eagerly awaiting serological testing, not just to follow the virus’s path. “It will allow us to trace in a much more population-based way who has had the infection,” says Nigel McMillan, an infectious disease specialist at Griffith University, Gold Coast. For instance, “Many cases seem to be spread from asymptomatic patients who we can’t identify easily.” Serological tests are “essential to better understand the epidemiology of COVID-19,” says Keiji Fukuda, an epidemiologist at the University of Hong Kong.

The test was developed by a team led by Linfa Wang, an emerging disease specialist at Duke-NUS. In blood samples from recovered patients, the team identified antibodies targeting the spike protein that proved able to block the virus from killing cells in laboratory tests. In parallel, they created synthetic viral proteins that can detect those antibodies in a blood sample without having to use the live virus.

Serological assays need to be carefully validated to be sure they react reliably, but only to antibodies against the novel virus, says Bart Haagmans of Erasmus Medical Center, whose group is working on serological tests itself in collaboration with Utrecht University. Indeed, one concern was that the similarity between the viruses that cause severe acute respiratory syndrome and COVID-19 could lead to cross-reactivity, Wang says. But the newly developed test can distinguish the two viruses with high accuracy and confidence, he says.

Elsewhere, development of serological tests is advancing rapidly. A group at the Wuhan Institute of Virology in China used an in-house test to show that serologic testing can be used to confirm infections, a paper published online on 17 February in Emerging Microbes & Infections shows. The technology has been commercialized, team member Peng Zhou wrote in an email, but it’s not clear how widely it has been used. EUROIMMUN, a diagnostics manufacturer, announced on 21 February that it has developed tests to detect COVID-19 virus antibodies, but its product has yet to be approved.

Singapore has not had the explosive spike in case numbers seen in many other countries, possibly because of its aggressive contact tracing effort and legal authority to order people into quarantine. As of 26 February, the city had a total of 93 confirmed cases; 2848 close contacts have been quarantined for 2 weeks.