Read our COVID-19 research and news.

A staffer at a Wuhan, China, hospital holds up a sample from a suspected COVID-19 patient.

STR/AFP/Getty Images

Labs scramble to spot hidden coronavirus infections

The seeming precision of the global tallies of cases and deaths caused by the novel coronavirus now spreading from Wuhan, China belies an alarming fact. The world is in the dark about the epidemic’s real scale and speed, because existing tests have limited powers—and testing is far too spotty. “We are underestimating how common this infection is,” cautions Jeremy Farrar, head of the Wellcome Trust.

Within days of Chinese researchers releasing the sequence of the virus on 11 January, scientists developed tests capable of detecting genetic sequences that distinguish the new agent from other coronaviruses circulating in humans. By 28 January, China’s National Medical Products Administration had approved diagnostic test kits from five companies. It was an astonishing pace for the response to a pathogen never seen before—and yet it was only a beginning.

Today, there aren’t nearly enough test kits available to keep up with the skyrocketing case numbers, and some parts of the world may lack enough trained laboratory staff to apply them. And because the genetic tests look for snippets of viral genetic material in nose and throat swabs or fluid collected from the lung, they only work when somebody has an active infection. Scientists are still scrambling to detect antibodies against the virus in the blood, which could help find people who had an infection and recovered. 

Hubei province, which includes Wuhan, accounts for 75% of the more than 43,000 confirmed cases of COVID-19, as the World Health Organization (WHO) named the new disease on 11 February. (A study group of the International Committee on Taxonomy of Viruses christened the novel virus severe acute respiratory syndrome coronavirus 2, or SARS-CoV-2, the same day.)

But many news stories have reported shortages of diagnostics in Hubei. “They’re overwhelmed,” says epidemiologist Ian Lipkin of Columbia University, who recently returned from China and is in self-imposed quarantine at home. Testing in Hubei has focused on people sick enough to seek medical care, so tens of thousands of milder cases may not have been picked up. Outside Hubei, testing is even patchier. “What’s the full picture in the other parts of China?” asks Keiji Fukuda, an epidemiologist at the University of Hong Kong who previously led outbreak responses at WHO.

Similar questions loom elsewhere. No cases have been confirmed in Africa, but there has been little testing. Initially, only two African labs were capable of detecting the virus, says John Nkengasong, who heads the African Centres for Disease Control and Prevention: “If this virus had shown up in Africa in December, or early January, it would have been devastating.” The continent is better prepared since a workshop in Dakar, Senegal, last week where lab workers from 15 African countries were taught how to use one of the new viral tests, which are based on the polymerase chain reaction assay, Nkengasong says. (Another workshop will follow next week.) Given that the virus has spread so widely, however, Farrar says he would be “very surprised” if it isn’t already in Africa.

Even in the United States, test kits are in short supply. Regulations require that the U.S. Centers for Disease Control and Prevention (CDC) supply all tests, but that agency only began to do so on 5 February and has shipped a mere 200 kits so far, each able to do at most 800 tests. U.S. officials still don’t test most people flying in from China but focus on those who have symptoms of the disease. “We’re not able to do the surveillance that we would want to do,” says Wendi Kuhnert-Tallman, who heads CDC’s laboratory task force for the virus.

Many labs, including Lipkin’s, are racing to develop antibody tests, which will do little to diagnose acute cases—it can take weeks for that immune response to kick in—but could help clarify mystifying questions about SARS-CoV-2’s spread.

Such tests use a surface protein of the virus—or, in Lipkin’s case, an array of peptides—to capture antibodies specific to the virus in the blood. But a new test has to be validated using blood from infected people. CDC prefers to wait for 3 weeks after a person becomes ill to let antibody levels build, Kuhnert-Tallman says. So far, “We have one single case in the U.S. that has reached the 21-day mark.” A team led by Marion Koopmans of Erasmus Medical Center in Rotterdam, the Netherlands, expects to launch studies of its first version of an antibody test next week. It could be several more weeks before a company develops antibody kits and can churn them out by the thousands.

Antibody tests might help pinpoint where and when this outbreak began, and which animal was the original source of the virus: Researchers could search for evidence of infection in stored samples of human blood or in animals that might be a natural reservoir of the virus. But the “most useful application is to screen different age groups of humans,” Koopmans says, to determine how many people become infected with few or no symptoms. If indeed scientists discover many mild cases, the rates of severe disease (estimated at about 20%) and death (2%) among infected people will plummet—which would finally be a bit of good news.