Just how many people actually have Covid-19? How long will it be before we can safely begin to ease social distancing? And is this a one-off crisis or are we now facing the threat of repeated waves of coronavirus pandemics on an annual basis?

These are all questions to which scientists around the globe are racing to answer through serological testing – detecting tell-tale antibodies in the blood to identify the real number of people in a population who have ever come in contact with the virus. Over the coming months, the results will determine everything from how long society’s shutdown needs to be, to evalsuating the effectiveness of the new vaccines on the horizon.

Right now, the NHS tests for Sars-CoV-2 – the virus that causes Covid-19 – through a diagnostic technique called polymerase chain reaction (PCR) which detects the virus’ genetic material in oral or nasal swabs. It’s highly effective, but it only returns a positive result when the virus is still present in the body. Serological testing will tell us how many people crossed paths with the coronavirus weeks or even months ago – sometimes without knowing – a figure which epidemiologists modelling the spread of Covid-19, and governments need to know to make accurate public health decisions.

“You’ll have heard [the UK’s chief scientific advisor] Patrick Vallance talk about herd immunity as a useful outcome if enough people have had the infection,” says Andrew Freedman, consultant in infectious diseases at Cardiff University. “That’s something you can determine if you do serologic testing on enough representative people across the country. If you find that 60 per cent or more of the population have got antibodies to the virus, so they’ve already had the infection, this would tell you that herd immunity might be successful and it’ll stop spreading. So then the government could stop the social distancing and isolation precautions at the moment.”

Since the start of February, dozens of prospective serological test kits for Covid-19 have been developed all around the globe, and the number is still rising. Details on one of the latest proposed tests – from Icahn School of Medicine in New York City – were published just two days ago.

Most of these take the form of an enzyme-linked immunosorbent assay (ELISA), a commonly used biochemical test for detecting existing diseases like HIV and Lyme. It works by mixing an individual’s blood sample with a solution containing proteins from the virus. If antibodies to Covid-19 are present in the person’s blood, they will recognise and bind to these proteins, triggering a colour change.

In countries like Singapore, such serological tests are already being deployed on a large scale, a national surveillance program which is likely to yield the first population data on just how widespread Covid-19 has been within a particular nation.

But there is one major issue currently facing governments and healthcare authorities looking to launch similar surveillance programs. While there are an abundance of potential serological tests, such is the speed with which they have been developed and made available, no one has had the chance to check whether they really do what they say. If a particular test is faulty – for example it also yields positive results for patients who have come into contact with one of the six other strains of coronavirus known to affect humans – then the results will be at best useless, and worst, potentially lead governments to make fatally flawed decisions, such as assuming herd immunity when it isn’t actually present.

Right now at the Mayo Clinic in Rochester, Minnesota, clinical microbiology director Elitza Theel and her team are one of the many centres undergoing the painstaking process of trying to work out which tests actually work, and which do not.

“It is amazing how many serological assays are coming out of the woodwork,” she says. “As well as making sure that they don’t mistakenly test positive for other diseases, we’re also ensuring that they do actually recognise Covid-19. One of the challenges and delays has been just getting the kits in because of the transportation bans. There are not a lot of flights happening. We’re currently looking at one assay from the US, two from Europe and two from China. There’s a need for this, so once we identify one we think is suitable, we’ll begin offering the testing.”

Theel is particularly concerned about the emergence of serological tests, which are being falsely marketed by some unscrupulous companies. Last month virologists in Wuhan claimed they had developed a serological test which could diagnose Covid-19 in 15 minutes. Theel says such claims are misleading, because it takes between 8 and 11 days after infection for individuals to develop antibodies to the virus.

“More and more serological assays are coming out from manufacturers saying these can be used at the point of care,” she says. “But with that time delay, you don’t want to use these assays to diagnose acutely ill patients who are just starting to show symptoms. [...]. This is more of a retrospective method. We need to make sure that the intended use is clearly stated to clinicians.”

But once suitably accurate serological tests have been identified, scientists in the US and UK believe that one of the first steps should be to use them to screen individuals currently performing vital roles within society, such as healthcare workers, to check their levels of immunity.“Those people could then safely perform essential functions within the community without fearing re-infection,” says Marc Lipsitch, professor of epidemiology at Harvard School of Public Health. “That’s the first and maybe the most important thing.”

After being deployed on a wider scale, they can not only give scientists an idea of the extent to which herd immunity exists, but also how long the antibodies providing that immunity actually last for. This is a crucial question which will determine much of the public health decisions taken over the coming months.

Comparisons between Covid-19, and OC43 and HKU1 – two of the four coronaviruses that regularly circulate among humans and cause seasonal colds – suggest that immunity could be relatively short-lived. Humans infected with OC43, for example, stay immune for less than a year. On the other hand, immunity against the SARS virus from 2003 lasts for far longer.

Over the next year, tens of thousands of people across the UK are likely to be required to undergo repeated serological testing over the course of many months to try and work out whether we face further outbreaks of Covid-19 or whether the threat will reduce and the virus will go dormant for decades. This in turn will influence how soon people are safe to stop isolating and return to work.

Lipsitch believes that around 1,000 tests a week, distributed around the country will be needed to provide an accurate assessment. “That would give you some picture,” he says. “To get a full idea of the accumulation of immunity, it could have to be in the millions. But if this is what’s necessary to keep societies functioning then it suddenly becomes very worthwhile.”

In the meantime, as the wait for new drugs for Covid-19 goes on, serological testing is also likely to be used to identify people who have recovered from the virus, who may be asked to donate their blood as a form of emergency treatment for elderly or vulnerable people that are acutely ill. This is a treatment idea for virulent infections which dates back to the Spanish influenza, more than a century ago.

“Serology can be used to identify people with high levels of neutralising antibodies that can kill the virus,” says Theel. “Their plasma would be tested, collected, screened for HIV and hepatitis, and then administered to ill patients. The idea being these antibodies could quickly activate and kill the virus in those sick people.”

😓 How did coronavirus start and what happens next?

❓ Does alcohol kill coronavirus? The biggest myths, busted

🎮 World of Warcraft perfectly predicted our coronavirus panic

✈️ Flight data shows the huge scale of Covid-19

👉 Follow WIRED on Twitter, Instagram, Facebook and LinkedIn

This article was originally published by WIRED UK