A potentially “game-changing” spit test for coronavirus is set to be trialled by the government on 5,000 police and army staff amid growing concern about the accuracy of invasive nasal swabs. The two-minute test requires someone to spit in a tube, and is thought to be as accurate, if not more so, than the throat and nose swab that detects if someone has Covid-19. Prof Paul Elliott, the Imperial College London scientist who is leading a major government programme on home testing, told the Guardian saliva tests would be trialled on 5,000 key workers in the next fortnight. He said he was “very, very interested” in the potential of the tests because they are much easier to use. A swab must penetrate deep into the mucous membrane behind the nose and in the mouth, often triggering the gag reflex. Elliott said that “clinical experiences” suggested that as many as 30% of nasal swab tests result in a false negative, where people are wrongly told they do not have the virus. The government trial will initially examine saliva antibody tests – which detect if someone has previously had Covid-19 – before being expanded to antigen tests showing current infections if it proves successful. The Guardian has also learned that the government is in advanced discussions with one of the biggest saliva test companies in the US about getting approval for its product to be used in the UK. The California-based firm, Curative, which is run by an award-winning UK scientist, has told officials it can provide 100,000 tests a week immediately, with the potential for capacity to be massively expanded if it gets the go-ahead. The company is providing tests for the US air force and said it had carried out more than 300,000 tests in five states including Texas and California. Early studies in the US, including one by Yale University, have found the saliva test to be as or more accurate than the nasal swab test. Philip Beales, a professor at the University College London Institute of Child Health, who has been helping to coordinate the efforts of UK testing firms, described saliva tests as a “game-changer” and said they could “really could get us out of this epidemiological nightmare”. Beales said his team had found that saliva samples were “superior” to nasopharyngeal swabs for detecting Covid-19 because they remove almost all possibility of “operator failure”. The government has refused to publish its research into the false negative rate of the nasopharyngeal test, despite tens of thousands people a day carrying out the difficult procedure. A number of the government’s regional drive-through testing facilities, which are run by private contractors, require people to carry out the swab on themselves with the aid of a nine-page, 20-step instruction manual. Matt Hancock, the health secretary, told the Commons this week that the efficacy of self-administered tests was “not significantly different” from tests administered by specially trained NHS staff. His claim has been met with scepticism by some in the scientific community, while health bosses in Greater Manchester have asked to see the research. “We’ve no idea how many [tests there are], are they statistically significant? I’d love to see those data,” said Beales, director of the Centre for Translational Genomics. The fallibility of the nasal swabs has led some GPs to advise patients to disregard negative results if they still have symptoms. Dr Richard Vautrey, the chair of the British Medical Association’s general practitioners committee, said he had seen patients with coronavirus symptoms but repeated negative tests. “No test is perfect, particularly when tests are self-administered and this is a particularly difficult test to do [so] there will be some false negatives,” he said. “Going on the symptoms is as good a guide as any as to what condition a patient has.” Martin Hibberd, professor of emerging infectious disease at the London School of Hygiene and Tropical Medicine, said false negatives “could be devastating if a patient was returned to a care home and passed the disease on”. Dr Philippa Whitford, the SNP’s health spokeswoman and a former breast surgeon, called for the government to publish its research. “A false-negative rate of a quarter or up to 30% is massive. There’s a danger that key staff are being told to go through the drive-through system [and self-testing] and so you could be sending positive staff who are virus carriers back into hospitals and care homes.”